BY THE SAME AUTHOR
INDIAN CURRENCY AND FINANCE.
Pp. viii + 263. 1913.
78. 6d. (5th thousand. )
THE ECONOMIC CONSEQUENCES OP THE PEACE.
Pp. vii + 279. 1919.
8s. 6d. (84th thousand.)
A TREATISE ON PROBABILITY.
Pp. xi + 466. 1921.
188. (3rd thousand.)
A REVISION OF THE TREATY.
Pp. viii + 223. 1922.
7s. 6d. (13th thousand.)
A TRACT ON MONETARY REFORM.
Pp. viii + 209. 1923.
7s. 6d. (12th thousand.)
A TREATISE ON MONEY.
Vol. I. — Pp. xvii + 363. 1930.
Vol. II. — Pp. viii + 424. 1930.
15s. each. (8th thousand.)
ESSAYS IN PERSUASION.
Pp. xiii + 376. 1931.
10s. 6d. (9th thousand.)
Popular Edition, 1933.
5s.
ESSAYS IN BIOGRAPHY
Pp. x-f 318. 1933,
7s. 6d. (4th thousand.)
THE MEANS TO PROSPERITY.
Pp. 40. 1933.
Is. (14th thousand.)
(The figures of sales are for Great Britain and the United Statpj, exclusive of translations.)
A TKEATISE ON MONEY
VOLUME n
THE APPLIED THEORY OF MONEY.
MACMILLAN AND CO., Limited
LONDON • BOMBAY • CALCUTTA • MADRAS MELBOURNE
A TEEATISE ON MON^Y
BY
JOHN MAYNARD KEYNES
KClIiOW or kino's collioi, cahbridob »
IN TWO VOLUMES VOLUME II
THE APPLIED THEORY OF MONEY
MACMILLAN AND CO., LIMITED ST. MARTIN’S STREET, LONDON
1914
PRINTS!) IN GREAT BRITAIN BT R. & R. CLARK, UMITSD, KDINBGBOH
CONTENTS
VOLUME n
THE APPLIED THEORY OF MONEY BOOK V
MONETARY FACTORS AND THEIR FLUCTUATIONS CHAPTER 22
PAGE
The Applied Theory of Money ..... 3
CHAPTER 23
The Proportion of Savings-dbposits to Cash-deposits , . 7
1. England ....... 8
2. The United States ...... 14
3. Elsewhere ....... 18
CHAPTER 24
The Velocities of Circulation. ....
(i.) The Conception of “ Velocity ” as applied to Bank Money (ii.) The Velocity of the Income-deposits and of the Business deposits distinguished ....
(iii.) The Velocity of the Income-deposits .
(iv.) The Velocity of the Business-deposits .
1. Great Britain .....
2. The United States ......
(v.) The Variability of the Velocity of the Business-deposits (vi.) Factors determining True Velocities .
CHAPTER 25
The Ratio of Bank Money to Beseryb Money . . .49
(i.) The Stability of Reserve-ratios . . . . .53
1. England ....... 55
2. The United^States . . . . . .61
3. Elsewhere ....... 64
(ii.) The Interchangeability of Non-Reserve Bank Assets . 66
(iii.) How ought the Reserve-ratio to be fixed ? . • .68
xr iVv
20
20
22
25
30
31 35 39 43
vi
A TREATISE ON MONEY
i
CHAPTER 26
PAGE
The Activity of Business ...... 79
(i.) The Influence of Business Activity on the Velocity of the
Business-deposits ...... 79
(ii.) The Relation between the Bank Clearings and the Volume
of Trade ....... 82
(iii.) Statistical Summary ...... 88
BOOK VI
THE RATE OF INVESTMENT AND ITS FLUCTUATIONS
CHAPTER 27
Fluctuations in the Rate of Investment — I. Fixed Capital . 96
(i.) The Statistical Indications . . . . .97
(ii.) Theories of the Credit Cycle based on Fluctuations in Fixed
Capital Investment ...... 99
CHAPTER 28
Fluctuations in the Rate of Investment — II. Working Capital 102 (i.) The Statistical Indications ..... 103
(ii.) The Theory of Working Capital . . . .116
(iii.) Productive and Unproductive Consumption . . . 124
(iv.) The True Wages Fund ...... 127
CHAPTER 29
Fluctuations in the Rate of Investment — ^III. Liquid Capital 130 (i.) Mr. Hawtrey's Theory of Liquid Stocks . . .131
(ii.) The Obstacles to an Accumulation of Liquid Capital . 133
(iii.) The “ Carrying ” Coste 136
(iv.) The Equation relating Price Fluctuations to ‘‘ Carrying ”
Costs 140
(v.) The Theory of the “ Forward Market ” . . .142
(vi.) Conclusion . 146
CONTENTS
V
#
vu
CHAPTER 30
Histobioal Illustrations ......
(i.) Spanish Treasure .......
(ii.) The Depression of the Eighteen -Nioeties (iii.) The War Boom, 1914-18 .....
(iv.) The Post-war Boom, 1919-20 .....
(y.) Great Britain’s Return to the Gold Standard.
(vi.) British Home and Foreign Investment after the Return to O^old ........
(vii.) The Upited States, 1925-30 .....
(viii.) The “ Gibson Paradox ”. . . . . .
PAoa
148
152
164
170
176
181
184
190
198
BOOK VII
THE MANAGEMENT OF MONEY
CHAPTER 31
The Problem op the Management op Money . . .211
(i.) The Control of Prices through the Rate of Investment . 211
(ii.) The Dual Functions of Bankers .... 213
CHAPTER 32
Methods op National Management — I. The Control op the
Member Banks ....... 225
(i.) The British System ...... 227
(ii.) The Continental System ...... 233
(iii.) The United States Federal Reserve System . . . 234
(iv.) Will Member Banks borrow from the Central Bank at a Rate
above the Market ? . . . . . . 243
(v.) Open-market Policy further analysed . . . . 250
(vi.) Method of varying Member Bank Reserve-ratios . . 260
CHAPTER 33
Methods op National Management — II. The Regulation op the
Central Reserves ....... 262
(i.) Existing Methods of Regulating Note-Issues . . . 265
(ii.) The Right Principles of Regulation .... 272
• CHAPTER 34
» Problems of International Management— I. The Relations of Central Banks to one another . . • •
279
VIU
A TREATISE ON MONEY
i
V CHAPTER 35
PAGE
Problems op International Management—II. The Gold Standard ........ 289
(i.) Auri Sacra Fames ...... 289
. (ii.) The Case for the Gold Standard .... 293
CHAPTER 36
Problems op International Management— III. The Problem op
National Autonomy ....... 302
(i.) The Dilemma of an International System . . . 302
(ii.) Methods of Regulating the Rate of Foreign Lending . 306
(iii.) The Significance of the Gold Points .... 319
(iv.) Should Standards of Value be International ? . . 332
CHAPTER 37
Problems of National Management— III. The Control op the
Rate of iNVBstMENT . . . . .339
(i.) Can the Banking System Control the Price-level ? . . 339
(ii.) Short-term Ratos of Interest and Long-term Rates • • 352
(iii.) Can the Banking System control the Rate of Investment ?. 362
a. The Direct Influence of Changes in the Short-term
Rato of Interest ...... 363
b. The Fringe of Unsatisfied Borrowers . . . 364
c. The Position of Issue Houses and Underwriters . 367
d. Open-market Operations to the Point of Saturation . 369
e. International Complications .... 374
. (iv.) The Slump of 1930 . . . . . . 377
CHAPTER 38
Problems of Supernational Management . . . .388
(i.) The Dual Problem of Supemational Management . . 389
(ii.) Methods of Supemational Management . . . 395
(iii.) The Bank of International Settlements . . . 402
(iv.) Conclusion . . . ..... 405
Index .
. 409
BOOK V
MONETAEY FACTORS AND THEIR FLUCTUATIONS
CHAPTER 22
' THE APPLIED THEORY OP MONEY
We now pass from the Pure Theory of Money and a qualitative study of the characteristics of a System of Representative Money to the Applied Theory and a quantitative study of the facts as they exist in the leading Monetary Systems of to-day, chiefly in Great Britain and in the United States.
The plan of this volume is as follows.
In Book V. we shall treat of the Monetary Factors and their statistical fluctuations, such as the pro- portions in which the aggregate of Bank-money is divided between Savings-deposits and Cash-deposits, the Velocity of Circulation of Bank-money, and the causes which make this aggregate of Bank-money to be what it is. The somewhat detailed character of these chapters is essential to enable us to judge of the relative quantitative importance qf diflerent factors. For, when the aggregate of Bank-fboney has been determined, the statistics of the Savings-^posits are the most important indication as to how much of this aggregate is being employed in the Financial Circula- tion and how much is left for the Industrial Cir- culation ; and given the volume of the* Industrial Circulation, the Velocities of Circulation determine, broadly speaking, what level of outout and incomes this is capable of supporting.
In Book VI. we turn aside from what might be called “ influences on the ^ide of money ” to in-
3
BE. V
4 A TREATISE ON MONEY
fluences on the side of investment ”, examining the causes of fluctuations in the rate of investment and illustrating the argument of this and the preced^ Books by analysing what happened on several typical occasions in recent history.
In Book VII. we reach for the first time the norma- tive side of our subject — namely, the question of what ideal objectives ought to be aimed at % the Currency Authorities of National Monetary Systems and also of the world as a whole, of the obstacles in the way of the attainment of these objectives, of the dilemmas' which sometimes present themselves, and of the best means of solution.
• In Book V. J may appear to the reader to be reverting to the old-fashioned “ quantity of money ” approach to ' the problem of price-determination, in- asmuch as I shall be concentrating on the supply of monetary facilities, or rather on the amount of them available for the Industrial Circulation. It may be well, therefore, that I should say a precautionary word as to the relationship, as I conceive it, between the quantity of money and the price-level.
If habits and methods in the receipt’lf^d dispos^ ^ of incomes are assumed unchanged, if the level of iii-' comes and the volume of output are given, and if there is no change in the velocity of circidation of the business-deposits A,, the amount of monetary facilities needed in the Industrial Circulation is uniquely deter- mined. If less than this is avaikMe after providing for the requiremente of the Financial Circulation, it will be impossible to maintain the existmg volume of incomes. Furthermore, in equilibrium, when all the factors of production are employed and Savings are equal to Investment, not only does the volume of the Industrial Circulation determine the volume of in- comes, but it also determines the level of prices and, subject to correction for fluctuations in the volume of output and employment, the rate of earnings. That
OH. 22 THE APPLIED THEORY OF MONEY ^ 6
is to say, when the •price-level is in equilibrium •with the cost of production the (^uantily of money available for the Industrial Circulation does (if habits and methods are unchanged) rule the situation ; and the only modification which we need introduce into the tradi- tional formula is the addition of the words “ avail- able for the Industrial Circulation The significance of the changes in the accepted theory which we have developed in Books III. and IV. lie in their applica- tion to the modus operandi of price -determination when equilibrium is disturbed, by reason of in- equality between Savings and Investment, and during the transition from one position of equilibrium to another.
This change is, of course, formally compatible with the traditional Quantity Theory ; — ^indeed it must be, since the latter is an identity, a truism. But it is not brought out by the traditional theory in an instructive and intelligible form, and lies covered up, along idth other factors, under the omnibus conception “ velocity of circulation ”.
Let us write our Quantity Equation as follows :
M'.V' = n.O
^ere M' is the volume of the Industrial Circulation, O^the volume of output, and n the price-level of out- put ; then — so we have claimed — V' is a complex notion not identified with V, the Velocity of Cir- culation. It is compounded of two elements ; — one of them dependent on habits and methods of bank- ing, business and industry, which is of a similar character to the traditional velocities of circulation, and the other dependent on the balance between Saving and Investment, being greater than unity when investment is in excess, equal to unity when invest- ment and saving are> equal, and less than unity when saving is in excess. ^
Now Book V. of this Treatise consists, in the main,
VOL. II u
6 ^ A TREATISE ON MONEY bk. v
of a statistical study of the monetary elements on the left-hand side of this Equation, as ^tinguished from what we have called the investment elements ; and thei^ purely monetary elements are the same as, or similar to, those of which the traditional Quantity Equation takes account. 4
CHAPTER 23
THE PROPORTION OP SAVINGS-DEPOSITS TO CASH-DEPOSITS
The meanings of Savings-deposits and of Cash-deposits have been defined in Chapter 3, and their relationships to the Financial Circulation and the Industrial Cir- culation in Chapter 16. Since the two together make up the total deposits, it follows that a fluctuation in the proportion of the Savings-deposits to the total deposits is liable to react on the volume of the Cash- deposits, and in particular on that of the Income- deposits, unless it is deliberately counteracted by a corresponding fluctuation in the volume of the total deposits. In this chapter we shall consider, on the basis of the statistical evidence, the extent to which the proportion of the Savings-deposits fluctuates in actual experience, so as to be in a position to consider the magnitude of the reaction of these fluctuations on the general monetary situation.
We saw in Chapter 3 that Deposit Accoimts in England and Time Deposits in the United States roughly correspond to the Savings-deposits, whilst Current Accounts in England and Demand Deposits in the United States roughly correspond to the Cash- deposits. In the United States the law requires that the amounts of Time Deposits and Demand Deposits respectively shall be separately published, so that there is no difficulty in obtaining statistical data — provided we can regard fluctuations in Time Deposits as approximately representative of the Savings-
8 A A TREATISE ON MONEY bk. t
deposits. But in England it has been impossible hitherto to obtain any reliable indications, except by the courtesy of the banks themselves.
I ^ England. — ^As a result, however, of inquiries, which, though not exhaustive, represent a fair sample, I have obtained indications which are very interesting indeed and will serve to illustrate the practical, as well as the theoretical, importance under the present British banking system of transferences backwards and for- wards between the deposit and the current accounts. At the same time we must remember, what has been pointed out in Chapter 3, that in England the dividing line between deposit and current accounts is decidedly blurred. The bulk of the deposits are held at quite short notice — ^seven to fourteen days — and as a matter of actual practice are often repaid on demand, a few days’ interest being deducted in lieu of notice. The deposits definitely fixed for longer periods are said to be not above a quarter to a third of the total deposit accounts.
In pre-war days the normal percentage of fixed deposit to total deposits in England was generally assumed to have approached 50 per cent.^ During the war deposit accounts were (relatively to the increase of deposits generally) very greatly depleted, and by 19f9 they seem to have stood at not much more than one-third of the total deposits instead of one- half. On the basis of such information as I have been able to obtain from bankers,® the annual per- centages of deposit and current accounts to total deposits seem to have moved as follows :
^ Annual averages have been given by Lloyds Bank back to 1902. The percentages of fixed to total deposits were 41*8 in 1902» about 44 from 1903 to 1906, 46*4 in 1906, and about 48*6 from 1907 to 1914. This is voiy similar to post-war experience. The fixed deposits seem to have l)een depleted by the Boer War, to have recovered steadily thereafter towards 60 per cent, and to have had this recovery aha^ly stimulated by the collapse of “ bull maricets in 1906-7.
* The figures given are the average of indications supplied to me by three of the ** Big Five More recently two of these have published their figures.
OH. 23 THE PROPORTION OF SAVINGS-DEPOSITS, 9
PKRCBNTAaS OF DbFOSIT AcCOUHTS AND CTTBBBNT AcOODNTS TO Total Deposits in England
|
1913 |
Deposit Accounts. |
Current Accounts. |
|
Per Cent. 48 |
Per Cent. 52 |
|
|
1919 |
34 |
66 |
|
1920 |
38 |
62 |
|
1921 |
44 |
56 |
|
1922 |
44 |
56 |
|
1923 |
43 |
57 |
|
1924 |
44 |
56 |
|
1925 |
45 |
55 |
|
1926 |
46 |
54 |
|
1927 |
46 |
54 |
|
1928 |
47 |
53 |
|
1929 |
48 |
52 j |
Thus there has been a progressive restoration in the proportion of deposit accounts towards the pre- war figure almost uninterruptedly from 1919 to 1929.
Since the Midland Bank and Lloyds Bank now publish their percentage figures month by month — examples which will, I hope, be followed by the other banks — ^it will be useful to quote these figures alongside the more comprehensive estimate given above :
Pebcbntaoe of Deposit Accounts to Total Deposits .
|
Midland Bank. |
Lloyds Bank. |
|
|
Per Cent. |
Per Cent. |
|
|
1919 |
28-6 |
39-3 |
|
1920 |
33-8 |
43*3 |
|
1921 |
39-7 |
49*3 |
|
1922 |
400 |
60-3 |
|
1923 |
40*2 |
48*5 |
|
1924 |
41-5 |
490 |
|
1925 |
42-7 |
50-4 |
|
1926 |
43-7 |
61-4 |
|
1927 |
44*3 |
62-6 |
|
1928 |
• 44*7 |
63-6 |
|
1929 |
. 46-8 |
64-% |
|
1930 (6 mos.) |
48*3 |
66*6 |
10 A TEEATISE ON MONEY bk. t
Tl^e figures show that the changes in the relative propo]^6ns of deposit and current accounts have been sufficiently large to cause the movements in total deposits to be a very misleading guide to the move- ments in current accounts — as is clearly brought out by the following table :
|
Average ^tal Deposits of Nine Clearing Banks.^ 1924=100. |
Assumed Proportion of Current Accounts to Total Deposits. |
Calculated Total of Current Accounts. 192^100. |
|
|
1919 |
90* |
ni 66 |
106 |
|
1920 |
100* |
62 |
111 |
|
1921 |
108 |
66 |
108 |
|
1922 |
106 |
66 |
106 |
|
1923 |
100 |
67 |
102 |
|
1924 |
100 |
66 ? |
100 |
|
1925 |
99 |
66 * |
97 |
|
1926 |
100, |
64 |
96 |
|
1927 |
103 |
64 |
99 |
|
1928 |
106 » |
63 |
100 |
|
1929 |
108 |
52 |
100 |
^ * Estimated figures. The actual figures have never been published.
Thus, if these estimates are correct,^ whilst total deposits Ibiiay have been no higher in 1920 than in 1926, current abcounts were 16 per centjhigher. The continual transference frbm current to deposit account, brought about by the gradual recovery of ^•the deposit accounts after their war depletion to their normal pre-war proportion, was, in fact, operat- ing af a concealed measure of deflation, sufficient — assui^iDg that it afiected the Business-deposits and the Income-deposits eqmkUy — to explain a drop in the price-level of about 20 per cent without the
I
^ If they are not, I hope that the bankers who are in a position to know will oorrect them.
OH. 23 THE PKOPORTION OP SAVINGS-DBPOSITS, 11
help of , any change in the volume of thn* total deposits, '
These figures are particularly helpful towards explaining the magnitude of the fall of pricedevels between 1920 and 1923, which was quite out of pro- portion to the fall, if any, in the total de|(osits, and also their subsequent history between 1923 and 1926. Between 1923 ancT 1926 the total Bank Deposits, as published, were unchanged, wJJlst the Consumption Index was also almost unchaiiged. The Total of Current Accounts, however, as is ifibwn by the above table, fe4 from 102 to 96. Since the Volume of Output was almost ceHaiuly less in 1926 than in 1923, the decrease of Current Accounts during this period supplies an important missing part of the explanation of the course of monetary events.
The most striking contribution to the explanation of the course of monetary events which is supplied by the separation of the figures of Current Accounts from Total Deposits in England rehites, however, to the War Period. It will be remeiibered that after the. failure of the first War Loan in 1914, which had to be largely subscribed by the Bank of England and other Banks, there was an intensive campaign of propaganda to secure subscriptions to ^bsequent War Loans from the public. Hi was argued that subscriptions by the Banks were mflationary, whereas subsqriptions from the public were not ; and this argument was held to cover even the case where members of the public took money off deposit-account* to pay for their subscriptions to War Loahs — ^indeed the Banks made special “ patriotic ” arrangeioH^ to facilitate this course. I do not think that^ was noticed at the time by anyone that this prii^dure was capable of becoming a potent instrument of infiation. Actually such subscriptions had the same effect as if the Banks had subscribed directly them-
12 ^ A TREATISE ON MONEY bk. v
selves and had increased their current accounts by the full amount of these subscriptions. The transference of money held on deposit -account by the public, where it was not functioning as cash at all, to the current account of the Government by whom it was expended, thus swelling other current accounts, increased the quantity of active money quite out of proportion to any observable movement in the total deposits of the banks. Some of the deposit accounts thus disturbed were probably amongst the oldest standing and most reliable accounts of this kind which the banks held. If we were to suppose that no more than one-third of the pre-war deposit accounts were invested in War Loans under the influence of patriotic propaganda, then (supposing deposit accounts to have accounted for half the total deposits before the war) current accounts would be swollen by 33 per cent and would be sufficient to support price-levels this much higher than before. The character of the Government expenditure was, in fact, such that by far the greater part of it found its way rapidly into the Income-deposits, with the result that the transferences from Sayings-deposits had their full effect in diminishing the purchasing power of money. The course of monetary events in Great Britain in 1916 and 1916 is, indeed, almost a perfect illustration of the manner in which such transferences can affect prices.
It is probable, therefore, that thia unobserved factor of transferences backwards and fq;rwards between deposit and current accounts playe4 a con- siderable part both in the rise of British’ prices between 1914 and 1920 and in their fall between 1920 and 1926. We should, however, be able to of this more precisely as regards the past, and also to adjust our policy in the future, so as to take acgouht of this factor, if all the Banks would agree to publish separately the figures of their deposit and of their
OH. 23 THE PROPORTION OF SAVINGS-DEPOSITS , 18
current Accounts, both retrospectively and hence- forward.
There is, moreover, a reform of banking practice, as well as of banking publicity, which is indicated as desirable. The present practice of the British banks is to keep the same proportion of reserves against their deposit accounts as against their current accounts. In the United States, on the other hand, this is not so. The Member Banks of the Federal Reserve System are required by law to keep a reserve of only 3 per cent against their time deposits as compared with figures of from 7 to 13 per cent against their demand deposits. This greatly mitigates the inflationary or deflationary influence on the In- dustnal Circulation of fluctuations in that part of the Financial Circulation which is represented by the Savings-deposits. It would tend to a smoother work- ing of the British Banking System in its eflects on industry if the banks were to work to a very low reserve percentage against deposit accounts instead of to the same percentage as against current accounts.^
If, indeed, we could be assured that deposit accounts accurately corresponded to the Savings-deposits, it would be justifiable from some points of view to main- tain no reserves at aU against them — ^which would ensure that changes in the volume of the Savings- deposits need involve no changes in the aggregate of Cash-deposits. The practical argument against f^is course, and esven against allowing a lower reserve ratio iox deposit accounts than for current accounts, is the ];isk of encouraging banks to make private arrangements with, and concessions to, their cus- tomers, by which what were really Cash-deposits would masquerade as Savings-deposite, and so avoid
The Rqyal Commission on Indian Currency (1926) has recommended (§ 161) the American arrangement for the proposed new Central Bank for India. Indian Banks are to keep 10 per cent of their demand liabilities and 3 per cent of their time liabilities with the Central Bank. Moreover, this arrangement has been in force in South Africa since 1923.
14 ^ A TREATISE ON MONEY bk. v
the necessity of providing a reserve. It is stated (see p. 17 below) that this evasion has in fact occurred to a certein extent in the United States. If this objection is held to be a good one on practical grounds, much the same result could be reached by the Central Bank having regard to the proportion of the Member Banks’ deposits held on deposit- and on current-account respectively in deciding the appro- priate level at which the aggregate of the reserves of the Member Banks might be allowed to stand. At present in Great Britain it is probable that the Bank of England does not even know the proportion, and much more probable that it does not appreciate its relevance.
Granted that it is often desirable that the Central Bank should pay attention to the volume of the Financial Circulation with a view to avoiding Capital Inflations or Deflations, which are liable to react sooner or later on the volume of Investment, nevertheless any change is desirable which would make it easier for the Central Bank to consider and deal separately with the Industrial and the Financial Circulations.
2. The United States. — This brings us to a study of the same phenomenon in the United States — ^though for the reason given above it cannot be as disturbing a factor there as it is in Great Britain. Nevertheless it is signiflcant to notice that it was the growth of time deposits in the Federal Reserve System between 1925 and 1929 which made possible the great increase in the loans and investments of the Member Banks , without a corresponding increase in their reserve requirements and, at the same time, without an increase in commodity prices.
The statistics of the Federal Reserve System is a separate study in itself, upon which I am not qualified to embark. The following table, 'however, of Demand and !^e Deposits respectively will serve to illustrate the significance of fluctuations, in the United States
OH. 23 THE PEOPORTION OF SAVINGS-DEPOSITS, 16
equally with Great Britain, of the ratio between Savings-deposits and Cash-deposits :
Perobntags to Total Deposits ^
|
Time Deposits. |
Demand Deposits. |
|
|
1918 |
23 |
77 |
|
1919 |
24 |
76 |
|
1920 |
28 |
72 |
|
1921 |
32 |
68 |
|
1922 |
32 |
68 |
|
1923 |
35 |
65 |
|
1924 |
36 |
64 |
|
1925 |
37 |
63 |
|
1926 |
38 |
62 |
|
1927 |
40 |
60 |
|
1928 |
42 |
58 |
Thus the upward tendency of the ratio of time to demand deposits has been more or less the same in direction and in amount in the United States as in Great Britain. Owing to the much lower percentage of reserves required against Time De- posits, this movement has permitted a substan- tially greater growth of available Bank Credit on a given basis of Reserves than could have existed otherwise. Moreover, if there had not been so great a growth of Time Deposits, there could not have been so great an extension of Bank Credit without raising prices. This has been a healthy development. But if the same reserve practices had been observed as in Great Britain, this growth of Time Deposits would have exercised a severe deflationary influence. Even
^ These figures are based on the average of the amounts held on the ** call *’ dates (3 to 5 a yeav) ; see ThirteerUk Annwl Report of the Federal Reserve Roard (1926), p. 142. The Time Deposits include postal-savings deposits. Very slightly different figures would result from taking ** net demand deposits ” instead of ** demand deposits
16 ^ A treatise on money bk. t
as it is (with the requirement of a 3 per cent reserve), it has beep an influence towards lower price-levels on a given basis of reserves than those reserves would have supported otherwise.
The different effects of the British and American banking practices respectively as regards the per- centage reserves held against Time Deposits can be illustrated by the following calculation. Taking the average reserves held in America against Demand Deposits at 11 per cent and against Time Deposits at 3 per cent, and the reserves held in Great Britain against Total Deposits at 11 per cent ; then if 10 per cent of the total deposits are transferred from de- mand account to time account, so that time deposits increase from 30 per cent to 40 per cent of total deposits, the effect on price-levels in the United States is to bring about, other influences apart, a tendency towards a fall of 4 per cent, whilst a similar transference in Great Britain would produce a tendency for price-levels to fall 14 per cent.
In fact the great growth of Time Deposits in the tJnited States, which doubled between 1920 and 1929, allowed the B^inancial Circulation to increase during those years to an extent which, with British reserve usages, would have involved eitj^er a severe fall of prices resulting from the contraction of the Industrial Circulation or a proportionate increase in reserves. The British system in its anxiety to prevent the possi- bility of inflationary activities on the part of the banks as purveyors of money for industrial purposes allows them no elasticity in the performance of their other function, namely, of meet^ the ,far more fluctu- ating requirements of the Financial Circulation. The American system, on the other hand, puts only a slight break on the ei^ansion of the Financial Circulation.
For preserving industrial stability the American system Actions with the greater wisdom and effective- ness. It can be objected on the other side that this
OH. 23 THE PROPORTION OF SAVINGS-DEPOSITS # 17
system may be capable, for this very reason, of allowing Capital Inflation to proceed to lengths which would be impossible on the British system. But if the maintenance of industrial stability and the optimum output is to be — as, on the whole, I think it ought to be — our main objective, then the American system is, nevertheless, preferable.
I should add that the American statistics must be read, quantitatively speaking, with caution. The increase in Time Deposits, according to the Report of the Federal Reserve Board (1926), p. 8, “ to some extent represents merely a transfer from demand to time deposits caused by the lower percentages of reserves required against time deposits and the greater efforts made by member banks to encourage savings accounts This is balanced, however, by the stringent character of the provision which reckons all funds held at less than thirty days’ notice in the demand category — an excellent provision in itself, since a system is to be welcomed which encourages as strict a segregation as possible between savings deposits and cash deposits.
Further, the rate of growth of Time Deposits in the United States since the war is partly due to the greater efforts now made by the Member Banks, under the stimulus of the low reserve proportion required to be held against them, to secure this business at the expense of the Mutual Savings Banks. The following table, given by Mr. W. R. Burgess {The Reserve Banks and the Money Market, p. 38), shows the result of combining the two classes of Savin gs- deposits :
' This is confirmed by Professor Parker Willis, “ Great Changes in American Banking ’* (The, Banker^ May 1927, p. 385), where he attributes the growth of time deposits to excessive competition between the banks, which leads them, not only to offer high rates of interest, but also to encourage customers to shift over into the time category funds which would normally have been held on demand.
18
A TR^TI^ ON MONEY
BK. V
(In millions of dollars)
|
Year. |
Savings Deposits. |
Per cent of Savings |
|||
|
Mutual Savings Banks. |
Time Deposits in Commercial Banks. |
Total. |
Individual Deposits. |
and Time Deposits to Total. |
|
|
1911 |
3,469 |
4,504 |
7,963 |
16,604 |
61 |
|
1914 |
3,910 |
4,802 |
8,712 |
18,891 |
46 |
|
1916 |
4,102 |
6,367 |
9,469 |
22,066 |
43 |
|
1918 |
4,382 |
7,163 |
11,636 |
24,618 |
47 |
|
1920 |
6,058 |
10,266 |
16,314 |
32,361 |
47 |
|
1922 |
6,818 |
11,761 |
17,679 |
36,336 |
48 |
|
1924 |
6,693 |
14,496 |
21,189 |
41,064 |
61 |
|
1926 |
7,626 |
17,171 |
24,696 |
47,472 |
62 |
This table is chiefly interesting, perhaps, for the remarkable stability of social practices in changing circumstances which it shows ; and also for the remarkable proximity between the corresponding American and British figures, — for if, to make them more nearly comparable with the above, we were to include in the case of Great Britain the Post Office Savings Bank deposits, the British pre-war and post-war percentages corresponding to the last column above would both be about 54,^ as compared with 61 and 52 for the United States.
3. Elsewhere. — ^It is evident that the normal per- centage of Savings-deposits to total bank-deposits will vary widely from country to country in accordance with the var3n[ng customs and traditions of the bank- ing public. I do not venture to quote figures for European countries with any confidence, since bank- ing statistics are apt to mean different things in different countries. Some comprehensive figures are available, however, for Germany. • On March 31, 1928,
^ The inclusion of the Trustee Savings Banks would raise the figure by 1 or 2 per cent.
OH. 23 THE PROPORTION OP SAtlNGS-DEPOSITS . 19
83 credit banks (including the 6 principal Berlin banks) 22 State and provincial banks, and 17 Giro-zentralen, having altogether deposits of £688,000,000 (exclusive of cheques in transit), held 39 per cent of these at call within seven days, 50 per cent at more than seven days’ and less than three months’ notice, and 11 per cent at more than three months’ notice. In the case of Australia the fixed deposits at interest, many of them fixed for one or two years, were estimated in 1927 at 60 per cent of the total deposits of £285,000,000.
CHAPTER 24
THE VELOCITIES OP CIRCULATION
(i.) The Conception of “Velocity” as applied TO Bank Money
The expression “ velocity (or rapidity) of circulation ” first came into use before the development of the cheque system, when the currency was mainly com- posed of coins and bank-notes. The “ velocity ” measured the average frequency with which a coin (or a bank-note) changed hands, and thus indicated the “ efficiency ” of the currency for the transaction of business.^ This was a definite and unambiguous idea. But it was necessary for its clarity that it should be applied only to the coins and notes which were being actually used as money, and not to hoards. For, otherwise, an increase (or decrease) in the amount of the hoards would appear as causing a decrease (or increase) in the velocity of the money, whereas what they were really causing was a decrease (or increase) in the supply, or quantity, of effective money. Thus it has been usual to limit the “ velocity of circula- tion ”, so far as practicable, to the effective money
^ There is also, however, a very ancient tradition in favour of regarding ** velocity *’ as connected with the ratio between a country's annual income and its stock of cash. I return to this in the next section of, this chapter. For a most interesting historical summary oi the development of the con- ception of velocity, bearing on this and other matters, see Holtrop’s ** Theories of the Velocity of Circulation of Money in Earlier Economic Literature ” (Economic Journal HiHory Supplemcni, Jan. 1929).
20
CH. 24 the velocities op CmCDLATION , 21
or money in active circulation, and not to stultify the conception by watering down the velocity of the money in circulation by including money which was not in circulation at all, but was being used as a “ store of value ” and therefore had no velocity ; changes in the amount of hoards being allowed for by regarding these as involving changes in the supply or quantity of circulating money raider than as changes in its velocity. For example, in estimating the velocity of circulation of money in India, it has been the prac- tice to exclude hoarded rupees as far as possible ; and even in the days of open mints, it was not usual to include in the stock of circulating money the hoarded ingots and ornaments held as a store of value by the nobles and the people of the country. A reduction of these hoards, as a result of famine for example, is more conveniently described as increasing the quantity of circulating money than as increasing the velocity of circulation.
When we extend the conception to Bank-Money, an analogous question arises whether we should take the total deposits or the cash deposits as representing the volume of money. It has been not uncommon to treat the whole of the deposits as belonging to the active circulation for this purpose,^ defining the velocity of circulation, when applied to Bank-Money, as being measured by the ratio of the total volume of cheque transactions per unit of time to the total volume of the Bank Deposits. In England, where we have no separate statistics of the deposit-accounts and the current-accounts, this is tantamount to treating “ hoards ” as cash for the purpose of com- puting the velocity. For we are treating as current money what is really a “ store of value ” — with the result that variations in the amount of this stofe
•
^ E,g, Professor Pigon takes this course {Industrial Fluctuations, chap. XT.), but not Professor Inring Fisher, who limits the application of ** velocity ** to the demand deposits.
VOL. II
G
22
A TREATISE ON MONEY
BK. V
appeax in the misleading guise of variations in the velocity of circulation. For reasons which we have made clear in Chapter 23 this inconvenient termino- logy is liable to stand in the way of drawing the right conclusions.
To avoid this difiElculiy, I propose to employ two terms, namely Velocity (V) and Efficiency (E) ; ^ of which the latter represents the ratio of the Bank Clearings to the Total Deposits. This leaves us free to use the expression “ velocity of circulation ” to denote unambiguously the velocity or rate of turn- over of what is truly serving the purposes of cash, namely the Cash Deposits. It follows that E = Yw, where w is the proportion of the Cash-deposits to the total deposits. It is not inappropriate to call the expression E the EJidency or Cash-Efficiency of Bank- Money. For the greater this fraction becomes, the greater is the volume of cash-turnover which corre- sponds to a given volume of Bank-Money ; — a growth of savings-deposits decreasing, and a falling away increasing, the “ efficiency ” of Bank-Money for cash purposes.
(ii.) The Velocity of the Income-deposits and of . THE Business-deposits distinguished
Even more important, however, than the dis- tinction between the Velocity (V) of the Cash-deposits and the Efficiency (E) of the total deposits, is the distinction between Vi, the Velocity of the Income- deposits, and Vj, the Velocity of the Business-deposits. The expression V is an average of two quite different things, and in a sense is not a true velocity at all. V is capable of changing, even though there is no change in either Vj or V,, as a result of a change in the propor-
•
^ In an earlier chapter 1 have used the letter £ to stand for “ Earn- ings*’. I hope that this duplication of the symbol win not lead to confusion.
,23
OH. 24 THE VELOCITIES OF CIRCULATION
tions of the Cash-deposits which represent Income- deposits and Business-deposits respectively ; just as the velocity of taransport of London passengers by tram and train might increase without there being any change in the velocities of trams or trains, because of an increase in the proportion of passengers travelling by trains.
Thus if, as before, Mj, M,, Ms and M repre- sent the Income-deposits, the Business-deposits, the Savings-deposits and the total deposits, the velocities of Ml, Mj and Ms are Vi, Vs and zero, the average weighted velocity of Mi and Ms is V, and the average weighted velocity, or, as I have called it, the efficiency of Ml, Ms and Ms, i.e. of M, is E ; so that, if B is the total volume of cash-transactions or money-turnover, we have
B =MiVi +M,Vs = V(Mi +Ms) =E(Mi +Ms +M,) = EM.
It is obvious from this that E and V may vary, even though Vi and Vj are constant, as a result of variations in the ratios of Mi, Ms and Mg to M. We have, therefore, to distinguish between changes in the composites V and E due to changes in the true velocities Vi and Vs, and those due to changes
in and
M ’ M M
This distinction will enable us, I think, to clear up a very ancient confusion. From the earliest litera- ture on the subject, as Dr. Holtrop has shown, ^ monetary theorists have oscillated between an in- clination to regard Velocity (or Rapidity) as a relation between the national stock of money and the national income, and an inclination to regard it as a relation between the stock of money and the total volume of transactions. The earlier writers were mainly in- fluenced by the former conception, but during the nineteenth century the latter has gained ground, until
^ Op. eU.9 paatim. Also his De Ondoopaandheid van hai Odd.
24.
A TKEATISE ON MONEY
SK. r
it has now become somewhat himly established, especially in contemporary American literature, by the writings of Professor Irving Fisher. Nevertheless clear traces of the former are still to be found not only in John Stuart Mill, but also — as we shall see — ^in the writings of Professor Schumpeter and Professor Pigou.
The element of confusion is this. If we are to interest ourselves in the relationship between the average stock of money and the national income, we must mean by the former the average stock held by the members of the public as enjoyers of income, i.e. the income-deposits, and not the total stock of money including the business -deposits. The relationship between the total annual receipts of income-receivers and the average stock of money held by them is one thing, which we call the velocity of the income- deposits ; and the relationship between the total flow of transactions for all purposes and the average stock of money held for all purposes is another thing, which we call the velocity of the cash-deposits. But the relationship between the total annual receipts of income-receivers and the average stock of money held for all purposes is a hybrid conception having no particular significance. Yet it is something of this kind which turns up in economic literature over and over again. For example. Professor Pigou, in a recent discussion round this problem {Industrial FhuAuatwnSt chap, xv.), distinguishes three kinds of “ Velocity Velocity in his first sense (op. cU. p. ^ 152) — ^the one for which he has, I think, a preference — ^is measured by the ratio of the money mcome of the community to the total stock of money; his second by the ratio of the money value of the quan- tity of income goods turned owr jor cash to the total stock of money ; and the third by the money value of the quantify of aU hinds of thirds turr^ over for cash to the total stock of money. The third of these is, in my terminology, the velocily (V) of the Cash-
OH. 24 THE VELOCITIES OF CIRCULATION
*26
deposits, or, if Professor Pigou includes the Savings- deposits in “ money ”, it is the efficiency (E) of
M
Bank-money. But the first of them =
notation, and is, therefore, the product of two quite different things. It is as though he were to divide the passenger-miles travelled in an hour by passengers in trams by the aggregate number of passengers in trams and trains and to call the result a “ velocity ”. Professor Schumpeter, also, has employed the term in what is, I think, nearly the same way.^
(iii.) The Velocity of the Income-deposits
Since we do not at present possess for any country even an estimate of the amount of the Income- deposits separately from the Business-deposits, it is impracticable to calculate their velocity directly by comparing their amount with the National Income. It is possible, however, to arrive at some estimate of the approximate limits within which this velocity is likely to lie by general considerations based on what we know of the habits of the community.
The velocity of the Income-deposits is a function of the community’s habits in regard to the intervals between wages and salary payments, whether weekly, monthly or quarterly,* and so forth, as to whether
^ Whether the expression “ circuit velocity of money ” used by Messrs. Foster and Catchings (in Frofila) is intended in my sense or in that of Professors Pigou and Schumpeter, 1 am not quite sure.
* This point was quite clearly apprehended by Sir William Petty in his Verbum SapierUi (1664), quoted by Dr. Holtrop (op. cii.)^ in a passage where he is considering the adequacy of the money in circulation : ** The expense being 40 millions, i( the revolutions were in such short Circles, viz., weekly, as happens among poorer artisans and labourers, who receive and pay every Satu^y, then 40/52 parts of 1 million would answer these ends. But if the Circles be qtiarterly, according to our custom of paying rent and gathering taxes, then 10 millions were requisite. Wherefore supposing payments in general to be a mixed Circle between one week and thirteen, then add 10 millions to 40/52, the half of the which will be 5|, so as if we have 5} millions we have enough,”
26'
A TREATISE ON MONEY
BK. ▼
they disburse their income regularly or irregularly between income-dates, and as to what proportion of their incomes they carry forward from one income- date over the next. More precisely :
Let R •= the annual income in question,
a; = the number of times incomes are paid annually {e.g^ x = 52 if incomes are paid weekly).
Let us assume that incomes are spent at a regular level rate between one income-date and the next, and that fluctuations are concentrated on the amount carried forward at the end of each income-period, i.e. on the amount of mcome-deposits in hand just before an income-date (this assumption is really no more than a matter of arithmetical convenience) ; and let
the average amount so carried forward be — .
It follows from this that the average level of the
income-deposits is therefore, that
their velocity Vi -
^ ^ 2x + y
The numerical value of Vi on various hypotheses is easily calculated. If incomes are paid weekly, i.e.
X = 62, and — is three weeks’ income, Vi = 15 approx. ;
R
whilst if — is one week’s income, Vi = 36. If incomes
. ^ . R .
are paid monthly, i.e. x = 12, and ~ ^ fortnight’s
income, Vi = 12. If incomes are paid quarterly, i.e.
a: * 4, and — is one month’s income, Vi = 6 approx.
The reader is free to make any other hypothesis which seems to him to be platisible. It is evident that the shorter the intervals at which incomes are paid, the larger will be the sum carried forward on the
OH. 24 THE VELOCITIES OP CIECULATION • 27
average in proportion to the amount of each income- payment, because for given incomes the more numer- ous will be the disbursements which fall at longer in- tervals than income-payments — e.g. holiday expenses.
It would not be difficult by means of a sampling inquiry to ascertain what proportion of their annual incomes typical members of typical classes hold on the average in cash or on current account. Failing this, I venture to guess, including in the definition of income-deposits both notes and bank-balances, that in England at the present time the value of Vi may be somewhere in the neighbourhood of 12 per annum, being perhaps 17 for weekly wage-earners and about 10 for those paid monthly and quarterly, etc. This means that on the average a weekly wa^e- eamer would hold cash equal to about 3 weeks’ in- come, whilst the rest of the community would hold cash and current bank-balances equal to about 6 weeks’ income, the average for the community as a whole working out at nearly a month’s income.* These figures are for the balances held on the average of the year and are compatible with substantially higher balances at quarter-days, etc.
These figures are, of course, in the absence of statistical evidence, no better than guesses intended to indicate the probable order of magnitude of the quantities involved and to stimulate the production of better statistics in future. Nevertheless the figures given would tally reasonably well with the known facts. For if we take the iucomes of the weekly wage-earners without banking accounts at £1,700,000,000 per annum and those of others at £3,000,000,000 per annum,® it would follow from the
^ Perhaps this is much too high for weekly wage-earners and somewhat too low for others. If so, notes held for business purposes must be in excess of my estimatcR
* These two figures add up to more than the net income of the country, because it is the gross income (before deduction of the interest on the National Debt, etc.) which is relevant to the present calculation.
28
A TREATISE ON MONEY
BK. T
above that tbe average quantity of Notes held as income-eaE^ by the former class would be £100,000,000, and the average quantity of Notes and Income- deposits held by the latter would be £300,000,000, of which, let us say, £276,000,000 would be Bank- deposits and £26,000,000 Notes. Since the active Note Circulation (i.e. notes in circulation not held by the banks) is probably about £260,000,000, the above assumption would leave £126,000,000 of Notes held otherwise, of which (say) £100,000,000 might be Busi- ness-cash and £26,000,000 Savings-cash (i.e. hoards held by those with no bank-accounts) ; and since bank balances on current account are probably about £1,076,000,000 (on the average of the year), it would leave £800,000,000 to cover the Business-deposits and disguised Savings-deposits. All these figures seem quite plausible. For if we raise the velocily for the salaried classes to 16, this would diminish the Bank Income-deposits to £176,000,000 (which seems much too low) and raise the Business-deposits, etc., to £900,000,000. On the other hand, a reduction of the velocity for weekly wage-earners seems scarcely compatible with the Imown facts of the volume of note issue ; whilst any material reduction of the velocity for the salaried classes would raise the Bank Income-deposits to what would appear to be an improbably high figure in relation to the Business- deposits. If any amendment is necessary it is likely to be in the direction of increasing the estimated velocity for wage-earners. This would change the average distribution of notes in active circulation as between earners and business, but would not afiect the above general conclusions relating to bank-money.
In defining the Savings-deposits and the Income- deposits in Chapter 3, we pointed out that the line between the two is not quite precise. A man who has Savings-deposits, upon which he can fall back if necessary, may regard this as a reason for economis-
OH. 24 THE VELOCITIES OF CIRCULATION
29
ing in the amount of his income-deposits. Some statistics relating to the amount of the Savings- deposits will, therefore, be in place here. In 1926 the deposit-accounts in Great Britain amounted, according to our estimate below (p. 31), to about £860,000,000, not the whole of which, however, would be held for the account of private individuals. The Savings - deposits in the Post Office and Trustee Savings Banks, which broadly speaking are the corresponding thing for weekly wage-earners who have no bank-account, were in that year about £370,000,000 (including Ireland). Thus the total Savings-deposits of private individuals may have been somewhere about £1,000,000,000, which is equal to about a quarter of a year’s income. If, therefore, our estimates are correct, the Income-deposits, in- cluding Income-cash, are between one-tenth and one- twelfth of a year’s income, and the Savings-deposits about a quarter of a year’s income.
The fact of some important disbursements, against which income has to be accumulated, being made at longer intervals than those between the date at which the majority of wages and salaries are paid, is clearly responsible for reducing the velocity of circulation below what it would otherwise be. The most important of these are probably quarterly rents, half-yearly rates and income-tax, annual payments of insurance premiums, holidays and Christmas extravagance (according to Mr. Burgess, op. dt. p. 76, department stores in New York and other cities usually do about one-seventh of their year’s business in the month of December alone). The payment of salaries at such long intervals as quarterly, harvest receipts to agriculturalists at annual intervals, and the pa3nnent of interest and dividends usually (in Great Britain) at half-yearly intervals also operate in the direction of reducing the velocity of circulation. As has been pointed out already, the more nearly
A TEEATISE ON MONEY
SK. T
30 •
the receipts and disbursements of ordinary individuals s3mchronise in date the less will be the average cash- requirements in proportion to income, and the greater, therefore, the velocity of circulation. Thus the velo- city of circulation is to a great extent a function of social habits and practices.
For this reason one would expect the velocity of the Income-deposits to be relatively stable from one year to the next, although it might show a definite trend over a longer period due to a progressive change of custom. But tms conclusion is subject to one important qualification. Many individuals are unable or unwilling to adjust their expenditure quickly to a change in their incomes, especially downwards. Thus at a time when money-incomes are changing, the sums carried forward from one income-date over the next may also tend to change in the same direction. That is to say, the first brunt of falling incomes will be borne by income-balances, and also the first advantage of rising incomes. For example, if Un- employment and bad times were to reduce the working-class average holdings of cash from three and a half weeks’ income to two and a half weeks’, and the middle-class holdings from five weeks’ to four weeks’, the result would be to increase the corresponding velocities of circulation from 15 to 20 and from 10 to 13, and the average from 12 to 15 ; and similarly in good times the velocity of circulation might be diminished below the normal. But there is no reason to expect that the abnormal figures would be lasting.
(iv.) The Velocity of the Business-deposits
In the case of the Business-deposits we have no secure data on which to base our guesses, corresponding to the probable amount of balances in proportion to income held by individual consumers. On the other hand, if our guesses about the volume of the Income-
31
OH. 24 THE VELOCITIES OF CIRCULATION
deposits and of the Savings-deposits can be regarded as reliable, we can deduce the volume and velocity of the Business-deposits from the statistics available for the total deposits and the total clearings.
There are many pitfalls and a wide margin for error in our figures ; but here again it should be possible to make a fair shot at the order of magnitude of the quantities involved, which is all we require in order to make progress in our general argument.
1. Great Britain
We will begin by attempting to calculate the velocity of the Cash-deposits, i.e. the weighted average of Vi and V*, the velocities of the Income-deposits and the Business-deposits respectively. The following table applies to England and Wales : ^
(£ Million)
|
(1) Total Clearings.* |
(2) Total Deposits on Dec. 31. |
(3) Current accounts. |
(4) Crude Velocity (Ratio of 1 to3). |
||
|
%of Total. |
Amount. |
||||
|
1909 |
14,215 |
711 |
52 |
370 |
38 |
|
1913 |
17,336 |
836 |
62 |
436 |
40 |
|
1920 |
42,161 |
2012 |
62 |
1247 |
34 |
|
1921 |
36,717 |
2023 |
56 |
1133 |
32 |
|
1922 |
38,958 |
1885 |
56 |
1056 |
37 |
|
1923 |
38,429 |
1856 |
57 |
1058 |
36 |
|
1924 |
41,414 |
1843 |
56 |
1032 |
40 |
|
1925 |
42,302 |
1835 |
55 |
1009 |
42 |
|
1926 |
41,463 |
1878 |
54 |
1014 |
41 |
|
1927 |
43,261 |
1923 |
54 |
1038 |
42 |
|
1928 |
46,878 |
1982 |
53 |
1060 |
44 |
|
1929 |
46,496 |
1940 |
52 |
1009 |
46 |
* Including Provincial ClearingB.
^ The figures for Great Britain, including Scotland, would probably be about 10 per cent greater.
32 A TREATISE ON MONEY »k. t
These figures are of value for the purpose of indicating thfe" variability of I3ie velocity of the Cash -deposits, and therefore, assuming that the velocity of the Income-deposits is approximately constant, of the variability of the velocity of the Business -deposits. But before we can reach “an estimate of the absolute value of this velocity certain corrections are necessary. In the first place, the total clearings, as already explained, do not cover the total cheque-transactions, since they do not include internal clearings between a bank’s own customers or local clearings between banks where there is no official clearing-house. Now that the bulk of the business is done by so small a number of banks as five, this must involve a serious error. To reach the aggregate of the total cheque-transactions, we must probably increase the total clearings by at least 35 ^ per cent, making the total for England and Wales £68,725 million on the average of the years 1926-28. In the second place, the total Deposits, as published on Dec. 31 of each year, certainly overstate the average figure for the year, probably by as much as 6 per cent and perhaps by as much as 10 per cent. If, therefore, we raise column (1) in the above table by 35 per cent and reduce columns (2) and (3) by 6 per cent, the crude velocity given in column (4) will be in- creased by 43 per cent. Thus our best estimate of the average velocity of the Cash-deposits from 1924-29 is raised to about 60 per annum.
These estimates for the English banking system as a whole may be compared with some more exact figures with which I have been supplied by the courtesy of Barclays Bank. This Bank compiles annual aggregates of the total debits passed through its books, i.e. of the
^ Even in the United States, where there are innumerable banks, it is estimated that only about two-thirds of the cheques drawn pass through the clearings (vide J. S. Lawrence, “ Borrowed Reserves and Bank Ex- pansion QuarUrly Journal of Economics^ 1928, p. 614).
33
OH. 24 THE VELOCITIES OF CIRCULATION
total volume of cheque-transactions, and not merely of those passed through the clearings, and these aggregates can be divided by the actual average amounts held on current account. The results are as follows :
Velocity op Current Acoounts in England
|
Barclays Bank (actual figures). |
The Banks as a whole * (estimated). |
|
|
1924 |
49 |
57 |
|
1925 |
51 |
60 |
|
1926 |
55 |
59 |
|
1927 |
58 |
60 |
|
1928 |
58 |
63 |
* This column is the “ crude velocity, as calculated above, increased by 43 per cent. As these pages are being passed through the press, it is announced that the banks as a whole have decided to publish the figures of total debits.
It is evident that the velocity for any particular bank will be influenced by the extent to which it is the bank’s practice to require its customers to keep a TniniTnnm balance on current account as a means of remunerating the bank, in preference to other methods of remuneration ; and also, in a large degree, by the extent to which it participates in high-velocity business such as Stock Exchange and other financial transactions.
From the above and from our previous estimates relating to the Income-deposits we can deduce an estimate of the velocity of the Business-deposits. Our previous arguments indicate the following ap- proximate figures in round numbers for Great Britain (1926-28)
Total Cheque-tranaaotions . £Mn.64,600
Total Cuirent Accounts . . £Mn.l,076
Velocity of Bank Cash-deposits 60
* The figures for clearings and for Cash-depositB just given, which relate to England and Wales, are increased by 10 per cent to allow for Scotland.
34
A TREATISE ON MONEY
BK. ▼
Bank Inoome-depoaita . . .
Velocity of Ba^ Inoome-depoaita . Cheque-tranaaotiona againat Inoome-depoaita
Buaineaa-depoaita
Cheque-tranaaotiona against BuBinees-depoaits Velocity of Buameaa-depoaita
^Mn.276
111
£lfn.3000
£Mn.80a
£Mn.61,500
77
To sum up, tlie best guesses we can make on the basis of existing statistics are that in England tbe volumes of tbe Bank Income-deposits, Business- deposits and Savings-deposits are roughly in the pro- portions 1 to 3 to 4 (Ms =4Mi, Ms =3Mi), whilst their velocities are (Vi) 11, (Va) 70-80, and zero.* But these guesses at the normal proportions between the different types of deposit must not lead the reader to forget what is an essential part of our analysis, namely, that these proportions are within limits variable, with
the result that changes in E ( =
MV -t-MVN ' Ml +JW.2 +Ms/
* ^ * * ' may be due not only to changes
or m
Mi-i-M,
in Vi and V*, but to changes in the ratios between Mi, Ma and Ms. Indeed I lean to the view that the velocity of Bank-deposits held for a given purpose does not readily change to any great extent except over long periods, and that the observed sharp short-period ffuctuations in the efficiency (E) of the currency or in thf velocity (V) of the Cash-deposits taken as a whole may often be due mainly to v^ations in the propor- tions of the deposits held for different purposes. We shall return to this matter in Section (v.) of this chapter.’
^ This is consistent with our previous estimate of a velocity of 10 for income-cash, if we include the encashment ^ cheques to self for notes as cheque-transactions against income-deposits, but exclude such changes of one kind of money for another in our previous calculation.
* Adam Smith concluded in a well-known passage ( WeaUh of Nationa, Bk. II. chap, ii.) that the velocity of circulation of business-cash was less than that of income-cash. But the argun^nt by which he supports this, namely that the individual transaetions m smaller in the latter case and that ** small sums circulate much faster than large ones ”, is not impressive, and was probably invalid even in his own day.
36
OH. 24 THE VELOCITIES OF CIKCULATION
I should remind the reader that we have been deal- ing above with the velocity of “ curreiit accounts ” in the^nse of this term or^narily adopted by British banks; and that the velocity of the “ cash facilities ” (see Chapter 3) would therefore differ from this. For we have had to include in “ current accounts ” the fixed minimum deposits retained by customers by agreement with their banks to remunerate the latter, but on the other hand have made no allowance for unused over- draft facilities. Perhaps these two sources of error may be regarded as roughly cancelling one another out.
2. The United Stales
The corresponding figures for the United States rest on a more secure statistical basis than the above. The pioneer work, in this case as in so many others relating to monetary statistics, was done by Professor Irving Fisher.^ Subsequently Dr. Burgess ® made a careful statistical examination for the period Jan. 1919 to Feb. 1923. Over such a period as this the figures naturally showed, not only an important seasonal fluctuation, but a wide cyclical fluctuation (seasonal range between maximum and minimum about 20 per cent and cyclical range at least 30 per cent). But as regards the average velocity during this period. Dr. Burgess concluded that for the whole country it was between 26 and 35 times a year, and probably under father than over 30. This average figure; concealed, however, wide variations for different parts of the country ranging from 74 for New York and 46 for Chicago to 20 for Buffalo and Rochester and 10 for Syracuse.
More recently Mr. Snyder has made further
' Professor Fisher’s guesses have turned out too high. Since my guesses for England are not much bettei^ based than were his for the United States twenty -five years ago* they also inay sufFer from the same defect.
•“Velocity of Bank Deposits”, Journal of the American Statistical Association, June 1923.
36 A TREATISE ON MONEr »k.t
invest^tions, wiiich add three years to the period dealt with by Buigetes and cover 1919-1926. Mr. Snyder’s figiires, for which he claims that the error does not exceed 6 per cent, broadly donfirm Dr. Burgess’s. But they bring out even more clearly the wide diyetgences of the average for different parts of the country, as the following taHe ^ shows :
Cash-skposits in U.S.A
|
Velocity. |
/ |
Whole Country. |
|||
|
New York City. |
141 Cities including New York. |
Whole Country. |
Volume of cheque payments. Billions of dollars. |
Volume of Cash- deposits. &luons of dollars. |
|
|
1919 |
76-2 |
42*3 |
28-8 |
546*8 , |
18-99 |
|
mm |
74-1 |
41*9 |
27*9 |
687-7 |
21-08 |
|
mSSm |
68*3 |
38*5 |
24*7 |
19-63 |
|
|
76-8 |
40*5 |
26*1 |
633-9 |
20*47 |
|
|
79-1 |
41*4 |
670-3 |
22*11 |
||
|
ld24 |
79-6 |
40*9 |
600-1 |
23*53 |
|
|
1925 |
87-7 |
44*2 |
25*98 |
||
|
1926 Average |
• • |
* * |
27-2 |
26*«7 |
|
|
1019-26 |
77-1 |
41*4 |
26-3 |
e • • |
It will be noticed that the average figure for New York, which is dominated by the Business-c^eposits, namely 77- 1, is very near my figure, namely 77,* fdr
^ The figures for New York and the 141 Gij^es are taken from Snyder’s Buaineaa Cydes and MeaswremmUSf p. 294 ; those for the whole country from Wesley MitcheU’s Busing Cydea^ p. 126. 1 understand that the former have been computed for each year separately, whereas the latter are primarily based on a single omnputation made in 1922 for 240 cities holding more than four-fifths of the total deposits. Subsequently (Review of Economic Statiatica, Feb. 11128) Mr. Snyder has estimated the cheque- tradsaotions in 1927 at 766 bi^ioi^ and the average velocity for the whole country in that year at about 30.
* The extreme closeness of these two figures is a chance coincidence. I had forgotten what the American figure was, and did not look it up, so as to avoid being infiuenped, until after I had worked out my English figure.
37
CB. *4 THE VELOCITIES OF CIRCULATION
the velocity of the Business-deposits in Great Britain. Cheques drawn on New York dt^ banks make up, moreover, between 40 and 50 per cent of all cheques drawn in the United States.'
4ls an example of the variability of the velocity according to the t3q)e of deposit in question, Dr. Burgess shows ‘ that, in the case of the balances of tlie Treasury Department, the velocity is in the neighbourhood of 300 times a year.
Let us next make a very rough attempt (in the hope of stimulating something better) to estimate separately the Income-deposits and the Business; deposits in the United States in 1923. The coin and notes in active circulation amounted to about 3*7 billion dollars, of which we might perhaps assume that 2 billion dollars would be held on the average as income-cash (i.e. held by the public as current cash apart from their hoards). If the velocity were about the same as that assumed for England, namely 16, the total income expended in this way would be 30 billions.® The net income was 70 billions, so that allowing an increase for gross income and for sundry duplications in payments, this would leave (say) 66 billions of income-transactions to be transacted by . means of cheques.* If we were to assume the same velocity as in England, namely 10,
^ 8nyder, Review of Economic Statistics, Feb. 1928^ p. 41.
* Th^ Reserve Banks and the Money Market, p. 91.
* This agrees reasonably with Wesl^ J^tohell’s estimate of 26 for the velocity of the cash in aU uses, since wtjy payment from income-cash to business-cash wiU be balanced by an equal payment from business-cash to income-cash, so that the above would aecohnt altogether for 60 billions turnover out of Mitchell’s estimated total of 94. In any case 30 billions seems a fully high proportion to be expended in this way.
^ Since making the above quite independent guess, I find that this figure is confirmed by Wesley i^tchell, who puts what I call the income- transactions at about 10 per cent of the total, which in 1923 would be 57 billions. The whole of the passage In trhich he makes this estimate is interesting and deserves quotation (Wesley Mitchell, Business Cycles, p. 149): “ Retail sales account for not much more than 9*7th of the aggregate volume of payments, and the payment of money incomes to individuals for about xVth. 'Even the round fiow of money incomes to individuals and from
VOL. II D
38
A TRBATISl ON MONEY
BK. T
tliis would give a total of 6'5 billions of income-deposits out of the total of 22, which is nearly the same proportion as in England. Further, we should be left with 16*5 billions of Business-deposits and 515 biUions of cheques drawn against them, which yields a velocity of 31 ^ — ^much lower than the English figure of 77. It is, however, evident on a broad survey of the figures and without making detailed hypotheses that the average velocity of the business- deposits in England must be much greater than in the United States. Conceivably this might be due partly to the greater use of the overdraft in England, but more probably to the greater distances in the United States which means a much greater loss of monetary efiiciency through the delays of cheques in the post.* In tliis case the use of air-mails might revolutionise the average volume of balances required to transact a given volume of business.
If we were to confine our attention to the 141 cities, then assuming that one-quarter of the cash- deposits are income-deposits with a velocity of 10, the velocity of the business-deposits works out at 55 — which is much nearer to the English figure.
It is interesting to notice that our guesses for the total income -deposits (cash and bank -balances to- gether) for Great Britain is £400 millions against a net income of £4000 millions or 10 per cent, whilst the total for the United States (in 1923) is $7*5 billions against a net income of $70 billions or 11 per cent. The reasonable agreement of these figures is some
irttfridnalB seems to make only jth of the aggregate payments in average business years. While these rather precise ratios may be faulty, it seems certain that the payments arising from other business transactions are several times the volume of payments involved in receiving and spending personal incomes.^*
^ If we lowered the velocity of income-deposits to 6, we should on the above hypotheses raise that of the business-deposits to
s As soon as I have posted a cheque, 1 treat this as out of my balance; but you do not treat it as in yours until the cheque has been received by your bank.
39
OH. 24 THE VELOCITIES OF CIRCULATION
slight confirmation that the order of magnitude of. these guesses may be fairly accurate — ^which is all I claim for them.
(v.) The Variability op the Velocity OP THE Business-deposits
We have endeavoured above to estimate the order of magnitude of the velocity (V2) of the Business- deposits. But we have no direct evidence of its variahility, except as an inference from the observed variability of the Cash-deposits taken as a whole.
We must begin, therefore, by considering the variability of the velocity of the Cash-deposits as shown by the English statistics. The best we can do with the existing statistics is to take an index of the total Bank Clearings and divide them by an index of the volume of Current Accounts ^ (based on the esti- mated proportion of Current to Total Accounts given on page above) ; or alternatively to take an index of the Country and Provincial Clearings and — since no separate statistics are available of Provincial Current Accounts — divide them by the same figure. The first set of figures will give us an index of the variability of the velocity of total Business-deposits, but, being overshadowed by the financial transactions of the City, will not be a satisfactory index of the velocity of the industrial Business-deposits or Business-deposits A, as we have called them above (Vol. i. p. 244) ; whfist the second set, which is better calculated to indicate the velocity of Business-deposits A, will be vitiated by the assumption that the Provincial Current Accounts have varied in the same degree as the aggre- gate of Current Accounts. The result, for what it is worth, is as follows :
^ index, based on the monthly average of the nine clearing banks,
dinen slightly, but not materially, from the index, given on p. 31, based on the end-year figures of all the banks.
40
A TKEATISE ON MONEY
BK. T
England and Walks (1024= 100)
|
Total Bank Clearings. |
Country and Provincial Bank Clearings.* |
Computed Level of Cash- deposits. |
Velocity of Cash-deposits based on |
||
|
Total Clearings. |
Country and Pro- vincial Clearings. |
||||
|
1920 |
102 |
152 |
Ill |
91 |
137 |
|
1921 |
88 |
101 |
108 |
81-5 |
93 |
|
1922 |
93 |
96 |
106 |
87 |
91 |
|
1923 |
92 |
96 |
102 |
90 |
94 |
|
1924 |
100 |
100 |
100 |
100 |
100 |
|
1925 |
102 |
101 |
97 |
105 |
104 |
|
1926 |
101 |
93 |
96 |
106 |
97 |
|
1927 |
106 |
98 |
99 |
106 |
99 |
|
1928 |
111 |
98 |
100 |
111 |
98 |
|
1929 |
112 |
98 |
100 |
112 |
98 |
* The Provincial Clearings are based on five towns up to 1921 and eleven towns subsequently.
The velocities in this table, based on the Country and Provincial Clearings, probably supply as good an estimate as we can get of the variability of the velocity of the Industrial Circulation ; whilst those based on the Total Clearings indicate the effect of the varying proportions in which the latter are made up of trans- actions relating to the Financial Circulation and to the Industrial Circulation respectively.
The general direction of the changes shown by the velocity based on the Country and Provincial Clearings id a satisfactory corroboration of the conclusions of Books III. and IV. that we should e35)ect velocity to be higher when business profits are being enjoyed than when losses are being suffered. The extent to which the Country Cheque Clearings are relatively free from the influence of large financial transaqtions is shown by the average value of the cheques passing, as cal-
OH. 24 THE VELOCITIES OF CIRCULATION
41
ciliated on two selected test days in 1928, namely as follows :
|
January 3,* 1928. |
August 24,t 1928. |
|||
|
No. of Items. |
Average Value. |
No. of Items. |
Average Value. |
|
|
Town clearing |
322,000 |
£538 |
£911 |
|
|
Metropolitan clearing |
234,000 |
36 |
34 |
|
|
Country clearing . i |
488,000 |
31 |
25 |
*** Selected as a typically heavy day. f Selected as a typically light day.
This is confirmed, subject to a qualification, by Mr. Holland-Martin, the Secretary of the Clearing House, who said in his Report on the year 1928 :
I should like to warn you that the town figures day by day become more swollen by the growing habit, for easier accounting, of using cheques for each transaction, and not for the balances of transactions, so that they are little guide to trade movements. The figures of the Country Clearing and those of the Provincial Clearings are a more reliable guide, but these also tend to be upset if, as happened in one district at the end of last year, the figures were swollen by large cheques passing between firms as the result of sales and purchases of businesses.
Now comes the question — are the observed changes in the velocity of the Cash-deposits really a reflection of changes in the velocities of the Business-deposits ? Or are they a reflection of changes in the proportionate amounts of the Business-deposits and of the Income- deposits, or of different categories of the Business- deposits ? The statistics themselves cannot answer this question. But we can calculate, first, what changes in the proportionate amounts of the Business- deposits and of the Income-deposits would produce the observed results, given constancy in their respective
42
A TREATISE ON MONEY
BK. ▼
velocities ; and, secondly, what changes in the velocity of the Business-deposits are implied, assuming a con- stant proportion between the two iypes of deposits and a constant velocity for the Income-deposits.
For the first of these two calculations, let us assume that the normal relation between the Income-deposits and the Business-deposits is that the quantity of the latter is 3 times that of the former, and its velocity 7 times that of the former (these velocities being both constant at 77 and 11 respectively). Then, since MiVi+M,V, 11 (Mi + 7Mj) ^ t.- 1. -x
mI-hm! • k~Ta'.' “
follows that a downward fluctuation of V from 60 to
65 implies a change in from about 3 to 2, whilst
an upward fluctuation of V from 60 to 64 implies a
M
change in from 3 to 4 approximately.
Next, let us assume that the relative amounts of the Income-deposits and the Business-deposits are con- stant. Then, taking the Business-deposits (M,) at 3 times the Income-deposits (Mj) and the velocity of the
latter (Vi) as 11, since Vg = |(Mi +M2)V it
follows that Vj»=^ (4V-11). Consequently, if the velocity of the Cash-deposits (V) fluctuates from: 60 to 66, we can infer on these assumptions that the velocity of the Business-deposits (Vg) moves from 76 to 70 ; whilst if V moves from 60 to 64, Vg moves from 76 to 82.
Furthermore, just as the Cash-deposits can be analysed into Income deposits and Business-deposits with different velocities, so the Business-deposits in their tom can be analysed into Industrial-deposits and Fmancial-deposits havio^ different velocities, which in Vol. i. I have called Business-deposits A and Business-deposits B respectively, the former belonging
OH. 24 THE VELOCITIES OF CIRCULATION
43
to the Industrial Circulation and the latter to the Financial Circulation ; so that observed fluctuations in the velocity of the Business-deposits may be due to fluctuations in the proportionate amounts of the Financial-deposits and the Industrial-deposits with- out any change in the velocities appropriate to these different fypes of deposit. Since it is probable that the velocity of the Financial-deposits is much greater than that of the Industrial-deposits, this may be important.
The available statistics do not allow us to carry our inductive illustrations any further. But enough has been said to show that the observed facts do not necessarily require for their explanation a sub- stantial fluctuation in true velocities ; — for they can also be explained by fluctuations in the proportions of deposits employed in different uses. Thus the observed increase in the velocity of the Cash-deposits during booms and its observed decrease during slumps may be only partly due to changes in true velocities ; — part of the change may be regarded as due to an increase of transactions relating to Business- deposits B, and part as a corroboration of our previous expectation that the value of transactions relating to the Industrial Circulation will increase relatively to the Income-deposits durii^ profit-periods and decrease relatively during loss-periods.
(vi.) Factoes deteemining Teue Velocities
It is important, therefore, to distinguish between the “ average ” velocity of money in a variety of uses and the “ true ” velocity of money in a par- ticular use, — ^meaning by the latter the ratio of the volume of a particular iype of transactions to the quantity of money employed in them ; for fluctua- tions in “average” velocities may be due, not to fluctuations in “ true ” velocities, but to fluctuations in the relative importance of different types of trans-
44
A TREATISE ON MONEY
BK. V
actions. But we n^ust now consider a little further the factors capable of causurg fluctuations ip the tnie velocities.
In deciding what amount of balances to hold in proportion to the prospective volume of transactions (both measured in money), a depositor is influenced partly by what amount is requisite for the transaction of business with a given demee of convenience, and partly by the degree of sacrifice involved in locking up that amount of resources in that way. Thus the level of balances held for a given type of transaction, proportionately to the volume of transactions, i.e. the true velocity in that use, is determined by a balance at the margin between the conveniencies gained and the sacrifices involved. The degrees both of con- venience and of sacrifice vary. But the variability of the latter is probably much greater than that of the former. Experience shows that the business world can get on, when the sacrifice is great, with a volume of balances which is only a fraction of what they hold at a time when the sacrifice is so slight that they can afford to indulge theic convenience.
The considerations of convenience are chiefly governed by slowly-changing social and business habits. But there is also a less sluggish short-period influence — ^namely that, when business is brisk, the business world can carry through a greater volume of transactions by means of the same quantify of cash** facilities without any loss of convenience.
Nevertheless the observed correlation between brisker business and increased velocity, in so far as this is due to an increase in “ true ” velocities, is probably also to be explained by the fact that when business is brisk it means a greater sacrijice to lock up in the shape of cash a command over real resources which might be converted into working capital. We must turn, therefore, to possible causes of variability in the degree of sacrifice involved in holding balances
45
OH. 24 THE VELOCITIES OF CIRCULATION
as compared with using this command over resources in sojjiej'p'ther way. In particular, we have to consider tl^ influence of (a) changes in the demand for work- in|,cap!tal and in the rate of discount, and (6) expecta- tions as to the future course of prices.
Ithere .is an element of sacrifice under heading (a) because the holding of balances locks up, so far as the individual depositor is concerned, a command of resources which but for this could be used in other ways^-on fixed deposit, in investment, or in the expansion of business. When business is dull and investment imattractive, the cost of holding cash- deposits is merely the sacrifice of interest involved in holding them thus rather than as savings-deposits or bills or other liquid short-dated investments. If, at the same time, the rate of discount and the deposit rate of interest are low, considerations of convenience in determining the amount of balances proportionately to business done can be allowed to have their way. When, on the other hand, business is active and expanding, so that the whole of the resources of a firm or an individual are required as active work- ing capital, when its borrowing powers are already strained towards their limit, and when the interest payable and obtainable on loans is high, then there will be a powerful motive to restrict balances to as low a level as is in any way practicable, even if this involves taking some risk in the provision made agaiust contingencies. Thus more transactions will be carried through with the same real balances as before, or higher-priced transactions with the same money- balances as before, i.e. with diminished money- balances proportionately to the money-volume of business.
When demand for working capital is keen and is tending to outstrip the supply, the pressure of these motives to economise balances may have a very important influence on prices. Obviously this state
46
A TREATISE ON MONEY
of affairs may be associated with a simultaneous increase in the volume of transactions ; but on the other hand there is no necessary connection. This is part of the reason why those who seek to find an exact relationship between the variability of V and that of T, the volume of transactions, are able to discover a statistical verification of what they expect when they direct their inquiries to some periods, yet fail to do so at some other period. For example, in the boom of 1920 in the United States the increase in the volume of transactions was accompanied by a stringency in the supply of working capital, with the result that V and T appeared to move together. In the recovery, however, which followed the slump of 1921-22 the volume of transactions reacted back to its former high level, but this time without any accompanying stringency in the supply of loanable funds, with the result that the supposed correla- tion between V and T, which the earlier period had appeared to substantiate, no longer held good. In other words, it is stringent conditions of credit which tend to increase velocity.
Thus the amount of sacrifice under heading (a) depends on the degree of the competition from other opportunities of employing resources. The amount of sacrifice under heading (6), however, depends on whether there is an expectation or a fear of deprecia- tion in the real value of the balances — i.e. on thtf possibility of rising price-levels.
The downward movement of real balances resulting from this cause has, of course, reached its extreme limit in the European post - war experiences of distrust of the currency resulting in a so-called “ flight ” from the currency. But the same pheno- menon is present when depositors generally are inclined to make purchases in excess of their im- mediate or normal needs, because they fear that, if they put them off to the usual time, the price will have
47
OH. 24 THE VELOCITIES OF CIBCULATION
moved against them in the meantime. To a certain extent tms shades off into the increased demand for working capital discussed under (a) above ; but the two influences can be distinguished according as the additional demand for goods is due to brisker trade rendering a larger supply necessary or to the expectation of higher prices leading to the purchase of more goods than are genuinely required at the moment. Obviously the two phenomena may often tend to exist together.
I am not clear that this cause is an important one in its direct influence on prices through increasing (or decreasing) velocity, except in the extreme cases where the fear of the depreciation of money is acute. It only operates where people take delivery of more goods than they require for the time being for their normal operations and hoard the goods, so to speak, in specie. This rarely occurs on a large scale. For it must not be confused with speculation or anticipatory purchases or sales of goods on the forward markets.
In extreme cases, however, where the expectation of a rising price-level is universal, so that private individuals begin turning their cash balances into goods or securities as soon as possible after receiving them for fear lest they depreciate in value on their hands, the effect of this influence towards diminishing real balances (and increasing velocity) and thus rais- ing prices may be catastrophic. Recent examples of this are so fa.Tni1iar that the point scarcely needs to be expanded.^ As soon as alarm has been awakened, price-levels tend to rise faster than the rate of monetary inflation ; and when alarm has be- come universal they tend to rise much faster. When this stage has been reached, even the restriction of further issues of currency will be of little avail in retarding the rise of prices, unless and until it has the effect of restoring popular confidence. Measures of
* I have dealt with it myself in my Tract on Mondary Reform, uhap. ii.
48
A TEEATISE ON MONEY
BK. ▼
deflation, of lugh bank-rate, of exchange control, of price control, etc., have to be judged by reference primarily to their probable eflects in influencing the public to restore their real balances to a normal level, rather than by their reactions on the other monetary elements.
In a country such as France, where the efficiency of the currency is normally rather low — ^mainly on account of the readiness to hoard notes — ^the value of money is especially susceptible to sentiments of distrust, inasmuch as efficiency can be greatly increased without anyone being put to any real inconvenience.
CHAPTER 26
THE BATIO OF BANE MONEY TO RESERVE MONEY
The crude Quantity Theory of old days — though it protected itself with the words ath^r things being equal — was liable to suggest that the total quantity of money (M) was the main, if not the exclusive, deter- minant of the supply of cash-facilities. Experience, under war finance and during the post-war monetary inflations, of price fluctuations which did not closely correspond with the changes in the ratio of the total quantity of money to the volume of output, has led almost everyone to-day to lay a more equal stress on the relative importance of the other monetary factors. Nevertheless the total quantity of money remains, if not an overruling factor, at least in the long run a dominant one — and of exceptional practical signifi- cance because it is the most controllable factor.
Let us, therefore, now proceed to a determination of the causes governing the total quantity of money. We shall move by two stages in this and a later chapter, considering, first of all, how the quantity of Bank- money is related to the quantity of Reserve-money, and then (in Book VII. Chapter 32) what governs the quantity of Reserve-money.
We have seen in Chapter 2 that the aggregate volume of the deposits of the member banks of a modem banking system depends on the reserve-ratio (i.e. the proportion of reserves to deposits) which the member banks aim, at keeping, and the amount of
49
60
A TEEATISE ON MONEY
BK. y
these reserves (in the shape of cash and deposits at the Central Bank). For if any bank finds itself in the possession of an amount of reserves which exceeds this ratio, it proceeds to “ create ” additional deposits, as described in Chapter 2, by lending and investing more freely ; which has the effect of increasing the reserves of the other banks and so causing them to create additional deposits ; and so on, until the normal reserve-ratio has been restored for the system as a whole. Thus, although the first bank will not be able to create additional deposits which are the full multiple (measured by the inverse of the reserve- ratio) of the surplus reserve with which it finds itself, since, as soon as it begins to lend more freely, it will lose a part of this additional reserve to other banks, nevertheless the existence of the surplus reserve in the banking system as a whole will set up a reverberation which will end in the aggregate deposits being increased in the same proportion as the aggregate reserves have been increased (after deducting from the initial increment of the latter any loss of cash out of the reserves into the active cir- culation consequent on the higher level of deposits) — ^for imtil this situation is reached some bank will find itself with more than a normal reserve-ratio. Thus the additional reserves and the increase of lending and of deposits will, in the end, be shared in appropriate proportions by all the banks of the system, and it will be of no consequence which par- ticular bank started the ball rolling. Some baiters have been inclined to dispute these propositions. But their objection usually turns out to amount to the contention that any given bank finding itself with surplus reserves cannot then increase its own lending by ten times (or whatever multiple corresponds to the normal reserve-ratio) the amount of this surplus — ^which is, of course, true.
The precise quantitative effect of the injection of
0H.2S RATIO OF BANK MONEY TO RESERVES 61
additional reserve resources into a banking system partly depends on what increment of cash in active circulation corresponds to a given increment in the volume of bank-money, as weU as on the ratio of the reserves to the volume of bank-money. For example — ^to take the simplest possible numerical illustration — ^if cash in circulation increases by 10 per cent of the increment in the volume of bank-money and if bank- reserves are normally 10 per cent of bank-money, an injection of additional resources will lead ultimately to an increase in deposits of five times their amount, since these additional resources will be eventually half added to the cash in circulation and half left as reserves, which latter half will lead to an increase in deposits of ten times their amount.
In the United States attempts have been made to estimate the actual statistical ratios relevant to this calculation. Governor Strong in his evidence before the U.S. Congress Committee on stabilisation, estimated that the increment of cash in circulation is, in the United States, about 20 per cent of the incre- ment in demand-deposits and the reserves 10 per cent of the demand-deposits. If these figures are correct, an injection of additional resources (in cash or Central- Bank money) will lead to an ultimate increase in demand-deposits equal to 3’3 times the amoimt of the additional resources.^
But whilst the amount of income-cash in circulation might be expected to bear a fairly stable relation to the amount of the income-deposits, it is not at all to be expected that the total amount of cash in cir- culation will bear any stable relation to the total amount of cash-deposits — at any rate over short periods or without an appreciable time-lag. Nor is
^ Professor J. S. Lawrence, in an interesting article in the Quarterly Journal of Economics, 1928, “ Borrowed Reserves and Bank Expansion , pp. 693-626, arrives by a ooniplicated argument, which I have not able clearly to follow, at 4-97 as his “ coefficient of ultimate expansion ”.
82 A TREATISE ON MONEY bk. v
the trend of change necessarily the same. For example, from 1921 to 1929 the amount of cash in circulation in the United States remained almost stationary, whereas during the same period the cash-deposits increased by more than 40 per cent. Indeed, one of the causes of disequilibrium in any period of change is sometimes the fact that, when reserves are first increased, bank-money can expand m a greater degree than can be sustained in the long run after there has been time for a corresponding increase in money- incomes and in that part of the income-deposits which consist of cash. It is a tiresome — and indeed a dangerous — ^feature of existing systems that different amounts of reserves have to be kept against income- deposits in the shape of cash from those which have to be kept against cash-deposits in the shape of bank-money, so that a change-over between the one and the other possesses a practical significance which it ought not, perhaps, to have.
The quantity of Bank-Money, i.e. of Bank-deposits, depends, therefore, on the amount of the reserves of the member banks and on the reserve-ratios at which the member banks are aiming either in conformity with law or convention or to suit theirown convenience. In this chapter I propose to assume the amount of the member banks’ reserves as given and to consider the amount of deposits which they will create on this reserve basis. In Chapter 32 of Book VII. we shall proceed to the further problem of the factors which determine the amount of the member banks’ reserves, and in particular to the question how far, in different monetary systems, the member banks are able them- selves to influence the kipount of these reserve^, and how far the amount id determined for them either by ,the Central Baik by external circumstances over which they have |^i^ntrol.
OH. 25 RATIO OP BANK MONEY TO RESERVES 63
(i.) The Stability of Reserve-ratios
It has sometimes been suggested that the banks vary their reserve-ratios in accordance with the state of trade. For example, Professor Pigou {Industrial Fhtducaions, p. 269) says that “this proportion is not rigid but elastic. In booms the joint-stock banks, feeling optimistic like everybody else, may be willing to cut down their proportion, and in depressions, being pessimistic, may wish to expand it.” The only evidence which Professor Pigou quotes in support of this conclusion relates to more than a hundred years ago. As we shall see below, modem statistics do not bear this out, so far as concerns Great Britain and the United States. We shall find that the ratio varies with different t3q>es of bank, and that it has been changed, for various reasons, from time to time. But at any given time banks stick closely to their established ratio, and, as the figures to be given below will show, such fluctuations as there are exhibit no correlation with the state of trade. This result is what one would expect. To let the ratio fall below the figure which has been fixed upon as that which is recommended by considerations of pmdence and of reputation would be a sign of weak- ness or, at least, of weak-mindedness ; whilst to let it rise above would be to forgo quite unnecessarily a source of profit, since surplus reserves can always be employed in the purchase of bills or of investments.
Accordingly, the statistics (to be quoted below) show that, save in exceptional circumstances, all banks use their reserves up^dto *the^hUt ; that is to say, they seldom or neverm^^h^^i^ reserves in excess of what is their con^^tmial or legal pro- portion for the time being. why should*
they, so long (is a perfectly liquid lliset can be pur- chased whicdi yields a rate of intd||Bt 1 The problem
VOL. n ®
64
A TREATISE ON MONEY
BK. V
before a bank is not bow much to lend — ^the answer to that question, namely the appropriate multiple of its reserves, can be furnished by simple arithmetic — but what proportion of its loans can be safely Bmtde in the relatively less liquid forms. Thus, in practice, the Member Banks exercise no effective control whatever over the aggregate of their deposits — ^unless they are in a position to control the aggre- gate of their Reserves, to which possibility I wiU re- turn in Book VII.
The conclusion that banks lend up to the hilt, in the sense that they never maintain idle reserves in excess of their normal ratio, depends, nevertheless, upon two conditions being satisfied :
(i.) It must be the case that interest-yielding assets are available, the liquidity of which is beyond ques- tion. Otherwise, a bank may sometimes have to keep excess reserves as the only means of avoiding a deficiency at some subsequent date. If a situation was to arise, such as has happened before now in Italy, where the Treasury could not be relied on to pay its Treasury Bills in cash at their maturity and where the Central Bank was not making new advances to the market against them, or if there was a fear that such a situation might arise, then this condition would not be satisfied. But with Central Banking on modem lines such a situation, or even the fear of it, is abnormal. In the United States the Federal Reserve Banks virtually undertake to dis- count bills of certain specified types ; and in Great Britain the Bank of England does so by long un- broken custom. The obligation of modem Central Banks to buy for cash certain specified types of interest-bearing documents is almost as absolute as their obligation to give legal-tender for their notes (if their notes are not themselves legal-tender). The Bank Bill, therefor^, is as good as gold — and better, because it earns intent. There is no reason, there-
OH. 25 RATIO OF BANK MONEY TO RESERVES 56
fore, why a bank should sacrifice the interest thus obtainable (directly, or indirectly by call loans to the market) by holding cash in excess of what is reqimed by the written, or the unwritten, law.
(ii.) It must also be the case that the cash reserves which a bank habitually keeps, in virtue of law or of binding custom, are in excess of the maYimum which it requires for the convenient transaction of business. In former times this was not always the case. It is not the case everywhere to-day. But the diminished use of cash, as compared with cheques, the greater rapidity of the means of transport from head- quarters, and the improbability of a serious “ run ” % the depositors of the banks (at any rate in England in the case of the “ Big Five ”) to turn their deposits into cash, have combined to reduce the amount of cash in proportion to their deposits which the banks strictly need as till-money. In addition to their till- money the Banks require a balance at the Central Bank to meet the contingency of an adverse balance at the “ clearing But beyond these requirements the Banks, provided they have a sufficient amount of abso- lutely liquid interest-bearing assets to meet all reason- able contingencies, have no need to hold any cash re- serves at aU — ^if it were not for law or binding custom.
Let us now test these generalisations by reference to the actual facts in England and in the United States.
1. England. — ^In England there is no law governing the proportion of cash which must be held by the member banks against their deposits. The figure is determined by custom and convention ; though, once the figure is determined, it would be bad for the prestige of a bank to lower its own ratio of reserves below the prevailing level. But there are two pecu- liarities about the English custom in this matter.
In the first place, since there is no law governing the question, it is more important to satjsfy the custom on the dates for which the figures are pub-
66
A TREATISE ON- MONEY
BK. ▼
lished thjin on those for which they are not published. Formerly they were only published half-yearly ; now they are published monthly. Partly, perhaps, as a survival from the days when these were tne only published figures, the banks are still accustomed to woilk to a much higher customary figure in their annual accounts at the end of the calendar year than in their monthly accounts, publishing indeed in their annual accounts a total of cash in hand and at the Bank of England as much as 50 per cent higher than they normally carry except on the annual balance- sheet day — which seems a stupid practice, whether or not it is intended to deceive. It is also the case — ^but to what extent no details are available — that the figures published in their monthly accounts, being averages of four days in the month, namely, of the day in each week on which their weekly balance-sheets are made up, are higher than their true daily average.^ Nor is this all : the “ Big Five ” Banks which follow this practice, being four in number (excluding the Midland Bank), can and do choose difierent days of the week for their little manoeuvre. That is to say, each takes it in turn to call from the money market a certain quantity of resources which will swell their balance at the Bank of England on the day of the week sacred to the particular Bank which is calling. In this way a certain part of the published reserves of the “ Big Five ” is a stage army which appears four times over. When Bank A’s sacred day has passed, it lends to the Money Market that part of its Bank of England balance wMch is no longer required for publication purposes, for the Money Market to pass on as promptly as possible to Bank B whose sacred day has arrived, so that a Bank of England balance which
^ Except, I think, in the case of the Midland Bank. It is of course essential to the smooth working of the system that a stable average should be compatible with sharp daily fluctuations for individual banks above or below this flgure.
OH. 25 BATIO OF BANK MONEY TO EESERVES 67
belonged to Bank A’s reserve at dawn has put in. a public appearance before sunset as part of Bank B’s ; and so on day by day. In short, as Dr. Leaf, when chairman of the Westminster Bank, frankly caressed it, the published reserves are “to a certain extent fictitious In this way the traditional “ strength ” of the British Joint-Stock Banks is safely preserved and handed on for the admiration of posterity.
In the second place, the banks do not all work to an identical figure for their proportion of reserves. These differences of practice may be due to differences in the class of business transacted or to differences in the amoimts held of the next most liquid asset. But they may also be mere survivals from past con- ditions, which no longer have any distinct meaning, or they may represent different estimates of the advertisement or prestige value of working to a proportion higher than the average.® In spite of these differences, each bank — the figures show — is, generally speaking, steadfast to its own figure, with the result that the proportion for the average of the banks as a whole is — allowing for the half-yearly movements upwards — ^remarkably stable also.
Apart from the higher proportions for the half- yearly balance-sheets, particularly in December, these figures represent for the period 1921 to 1926 something as near absolute stability as one can expect to find. Over these troubled and changing years the figures exhibit no correlation whatever with the state of trade or with the bank-rate or with the fluctuating propor- tions of the advances (given below, p. 68), or with any such influences ; and for the average of each year from 1921 to 1926 the reserves bore a ratio to the deposits stable to within one thousandth part of the latter. The drop in the average since 1927 was mainly due to a
^ BanJdngf p, 133. . t j -d i
• This probably explains the high figure to which the l^dland^ Bank worked up to 1927 — a figure which dates back to Sir Edward Holden s day, who introduced it and made much of it.
58
A TREATISE ON MONEY
BE. T
change in the practice of the Midland Bank, which will be referred to later, and to a decrease in “ window- dressing ” at the end of each half-year.
Percentage of the Amount of Cash in Hand and at the Bank OF England to the Amount of Deposits held by the Nine ^ Clearing Banks
|
1921.* |
1922. |
1923. |
1924. |
1926. |
1926. |
1927. |
1928. |
1929. |
|
|
Jan. |
11 3 |
114 |
120 |
11*7 |
11*9 |
11*7 |
116 |
11-3 |
10-9 |
|
Feb. |
110 |
11-6 |
116 |
11-6 |
11*8 |
11*7 |
11*6 |
11-0 |
10*6 |
|
Mar. |
IM |
11-6 |
11-7 |
11-6 |
11*6 |
11*7 |
11-6 |
IM |
10*6 |
|
April |
11-7 |
11-7 |
120 |
11 6 |
118 |
11-7 |
11*7 |
IM |
10*8 |
|
May |
11 7 |
11-9 |
11-8 |
11-6 |
11*6 |
11-8 |
11-6 |
IM |
10-9 |
|
Junef |
12-3 |
11*8 |
119 |
11*9 |
122 |
12*1 |
11*8 |
11*2 |
10-9 |
|
July |
11*7 |
11-6 |
118 |
11*6 |
11-9 |
11-8 |
11*6 |
110 |
10*7 |
|
Aug. |
11-7 |
11-7 |
11*8 |
11*6 |
11*9 |
11*8 |
11-6 |
IM |
10-7 |
|
Sept. |
11-8 |
11-7 |
11-9 |
11-7 |
11*8 |
11*7 |
11*6 |
11-2 |
10-9 |
|
Oct. |
120 |
11-6 |
11-7 |
11*6 |
11 6 |
11-7 |
11-4 |
11-0 |
10-7 |
|
Nov. |
11-3 |
11*6 |
11-6 |
116 |
11-6 |
11*6 |
11-3 |
11-0 |
10-6 |
|
Deo.f |
124 |
121 |
12-2 |
124 |
121 |
120 |
11*7 |
11-3 |
11*3 |
|
Aver. |
1 11-7 1 |
11-7 |
11-8 |
11-7 |
11-8 |
11*8 |
11-6 |
IM |
10*8 ; |
♦ The relatively low figures for the first quarter of 1921 are probably to be explained by the fact that this was the beginning of the post-war pub- lication of monthly figures, so that the conventional ratios of the different banks had not yet had time to settle down.
t It will be noticed that the outburst of display at the end of the year has the effect of raising somewhat the weekly average for December, and that, until recently, the same thing has also occurred on a smaller scale in June. The figures since 1927, however, suggest that this practice is disappearing.
As between the different banks there are, however, appreciable differences of practice. Of the ten Clear- ing Banks which publish monthly figures, the Mid- land Bank used to maintain a reserve ratio much higher than the others, namely from 14-5 to 15 per cent, but now in the neighbourhood of 11 per cent — ^which may be relatively higher than appears, since it is believed that the Midland Bank’s monthly figures not “ window dressed ” ; • Lloyds Bank and
^ The Natidhal Bank, which is now included in the Clearing House figures, is excluded from this table so as to secure continuity of record.
Xl * It is possible that the true duly averages of the other banks, apart from window-dressing might not work out at much above 9 per cent.
OH. 25 EATIO OF BANK MON^Y TO RESEKVES 69
WiUiams Deacon’s Bank, publish reserves of 11 to 11-5 per cent ; then five others from 10 to 11 per cent ; and finally Coutts’ Bank, the business of which is rather different from that of the others, with 8 to 9 per cent. On the other hand the Midland Bank partly balances its high cash reserves by holding less “cash at call and short notice” than the average, and Coutts’ Bank by holding more.
If we turn to pre-war statistics, we find a tendency for the reserve-ratio to change slowly ; but changes were very gradual and in no way correlated with bank-rate or the state of trade. The publication of monthly figures was first made by thirteen banks in 1891 at Goschen’s suggestion.^ The proportion of the reserves then amounted to 13 per cent. By 1898 the proportion had risen to 14 per cent, and by 1908 to 15 or 16 per cent, which was still the figure in 1914 before the outbreak of war. The pre-war fibres, however, included in the reserves “ Balances with and Cheques in course of Collection on other Banks in the United Kingdom ”, as well as cash in hand and Deposits at the Bank of England. If these were to be included now, they would raise the proportion by something between 3 and 3-3 per cent, thus bringing the post-war average, prior to the Midland Bank’s change of ratio, up to about 16 per cent, or nearly the same as the pre-war figure. Such excess as remains in favour of the pre-war figure may be due to the fact that the figures now represent averages for a single day in each week (different for different banks), whilst the pre-war figures, being only for a single day in each month (different for different banks), may have been raised by “ window-dressing ” even higher above the true average figure than they a?^ now.
As to the division of the banks’ reserves between
^ The deposits of these banks were then £122,000,000, which was abont^^- 30 per cent of the aggregate deposits of the country.
60
A TREATISE ON MO^Y bk. t
cash and Bank of England deposits, it is possible to make an estimate. One bank (The Union of London and Smith’s) was accustomed in pre-war days i» state separately its cash in hand and its cash at the Bank of England, showing about half its reserves in each form. At the present time (1928) it would seem that fully two-thirds of the aggregate reserves of the Banks must be held in Notes and not more than a third as cash at the Bank of England. For the total deposits at the Bank of England, first pub- lished in November 1928,^ on the occasion of the amalgamation of the Bank and Currency Note Issues, of all British Banks doing mainly domestic business (which include several institutions besides the ten Clearing Banks), stood (Nov. 1928) at no more than £62,500,000 ; whereas the total pubhshed reserves of the ten Banks range from£l90,000,000 to £200,000,000. If we assume that the published figures are increased by (say) £22,500,000 as a result of window-dressing, we have left a figure of (say) £172,500,000, of which perhaps as much as £115,000,000 is held in the form of Notes and £57,500,000 at the Bank of England. There is no reason why we should not know precisely, but — ^like so much else in banking statistics — the figures are kept a secret.
The proportion of the published reserves of the nine Clearing Banks to the total Note Issue (Treasury Notes and Bank Notes combined) plus the total private deposits at the Bank of England is shown below (p. 61).
The figures in the last column, giving the proportion of State Money and Central Baifir Money held in re- serve by the nine Clearing Banks to the total of such money, show, recently, a steady and fairly substantial iiicrease. Since “ window-dressing ” has probably been on the decline these figures may understate the increase
^ * That is to say, since 1877. IVom the passing of the Bank Act in 1844 to 1877 the London bankers* balances were ^own as a separate item.
OH. 25 RATIO OF RANK MONEY TO RESERVES 61
♦i
Monthly Avbbages
|
T — C — at B. of £. (1) |
Bank and Currency Notes. (2) |
Total of columns 1 and 2. (3) |
Index of this. (1923 » 100.) (4) |
Deposits of 9 Clearing Banks. (1923*. 100.) (6) |
Reserves of 9 Clearing Banks. (1923*. 100.) (6) |
• Proportion of col. 6 to col. 4. (1923-100.) (7) |
|
|
1921 |
£Md. 124 |
£Mn. 435 |
£Hn. 659 |
113 |
108 |
107 |
95 |
|
1922 |
118 |
399 |
517 |
104 |
106 |
105 |
101 |
|
1923 |
110 |
386 |
496 |
100 |
100 |
100 |
100 |
|
1924 |
110 |
389 |
499 |
101 |
100 |
99 |
98 |
|
1925 |
111 |
383 |
494 |
100 |
99 |
99 |
99 |
|
1926 |
105 |
374 |
479 |
97 |
100 |
100 |
103 |
|
1927 |
101* |
373 |
474 |
96 |
103 |
101 1 |
105 |
|
1928 |
102 |
372 |
474 |
96 |
106 |
99 |
103 |
|
1929 |
99 |
361 |
460 |
93 |
108 |
99 |
106 i |
♦ This figure is believed to have been vitiated in the spring of 1927 for purposes of comparison, to the extent of at least £5,000,000, by the disappearance of some item arising out of the loan from the Bank of England to the Bank of France which was then paid ofi.
rather than otherwise. The absolute amount of the published reserves of the nine Clearing Banks was the same in 1929 as in 1924, whereas the total of column (3) had fallen by 8 per cent. This suggests that the notes in the hands of the public must have fallen appreciably, whilst cash deposits — as we have seen (p. 40) — ^were practically unchanged. Thus the indi- cations are that the use of Notes is still sensibly de- clining relatively to the use of cheques. The table also brings out in a striking manner the effect of the Midland Bank’s change in its reserve ratio ; for, although the Bank of England kept the aggregate of State and Central Bank Money in 1929 4 per cent below that of 1926, of which 1 per cent was re- flected in a reduction of the reserves of the Clearing Banks, the Member Bank Money increased by 8 per cent.
2. The United Stofes.— The United States mainly differs from Great Britain in that the reserve proper-
62*
A TEEATISE ON MONEY
BK. T
tions are a matter of law and not merely of custom. Hie actual reserves maintained are, as we shall see, milch the same in practice.
The Member Banks ^ of the Federal Reserve System are required to hold reserves in the form of deposit accounts with their Reserve Bank in the following percentages : ®
|
Demand Deposits (payable on 29 days’ notice or less). |
Time Deposits* (payable on 30 days’ notice or more). |
|
|
Central Reserve Cities |
||
|
(New York and Chicago) |
13 per cent |
3 per cent |
|
63 Reserve Cities |
10 „ |
3 „ |
|
Country Towns |
7 „ |
3 |
* Before 1914 there was no distinction in the reserve requirements against Demand and Time Deposits respectively. At the inception of the Federal Reserve System the reserve against the latter was fixed at 6 per cent, and was reduced to 3 per cent by an amendment to the Act in 1917.
In practice (for reasons into which I need not enter) it is even more complicated than the above table would suggest to compute the amount of the legal reserve from the published figures of the Demand and Time Deposits. The amounts of the legal reserves, as compared with the actual reserves, are, however, compiled at irregular intervals by the Comptroller of the Currency, in order that he may satisfy himgelf that the baj^ are duly complying with the require- ments of the law. The legal proportion averaged over
^ At the end of 1926 the Member Banks numbered 9260. There were also no less than 17,824 non-member banks. But the assets of the Member Banks represented about 60 per cent of the total.
* Cash in hand does not reckon towards the required proportion. Con- sequently its amount has been worked down to the minimum needed as till-money. The actual percentage of demand deposits so held on Dec. 31, 1928, was less than 2 per cent for Member Banks in reserve and central reserve cities and 5 per cent for country banks, so that the difference in the legal proportions between reserve cities and country towns was offset by the larger amount of cash held in the latter.
OH. 25 RA.TIO OF BANK MONEY TO EESERVES 63
the net demand deposits generally works out at between 10 and 11 per cent. Since the proportion against Time Deposits is only 3 per cent, the legal proportion against the total deposits falls when there is a relative increase of the Time Deposits, which cause has reduced it from about 9 per cent in 1918 to 8 per cent in 1922, and 7-2 per cent in 1927. As, however, till-money does not reckon in the United States to- wards the legal reserve, we have to make an addition to the above of about 2 per cent, to render the figures comparable with those given for England. If an allow- ance be made for the English practice of window- dressing, it would seem, therefore, that the cash and Central Bank deposits, which law or custom requires the Member Banks to hold in proportion to their own total deposits, are now very nearly the same in the two countries, at (say) 9 to 9-6 per cent.
Furthermore, the percentage of reserves actually held in the United States does not normally differ to any material extent from the legal miniiiim. In the unsettled conditions of the war period the actual reserves were occasionally in excess of the minima by as much as 1 per cent of the deposits ; but since the war they have seldom exceeded the minima (except on Dec. 31) by more than -5 per cent of the deposits and are often only -1 or -2 per cent above the minima. For example, in the middle of 1926 the reserves actually held by the National Banks (which represent more than 80 per cent of the member banks) were 7-6 per cent, whilst the prescribed l^al minirmim was 7*4 per cent. This means that the member banks are using their reserve-bank deposits practically up to the hilt.
These results are confirmed by the express state- ment of the Federal Reserve Board in their Annual Report for 1924 as follows (p. 9) :
** While the proportion of the amount of reserve bank credit outstanding to loans and investments of
64
A TREATISE', ON MONEY
BK. T
member banks is only a fraction of wbat it was in 1920, the ratio between reserve balances maintained by member banks at the reserve banks to member bank deposit liabilities has remained practically con- stant at about 10 per cent,^ which represents on the average the minimum required by law.”
That the member banks should wish to work as close as possible to their legal minima is explained by the natural incentive to keep down the assets not earning interest. That they should be able to do so is explained by the existence of certain special facilities. In the first place the law allows the banks’ reserves to fall below the prescribed figure on a parti- cular day provided the weekly average is sufficient. Any deficiency in reserve can be quickly made good by discounting at the Federal Reserve Bank. Any igexcess can be immediately loaned in the Call Market, and used in due course to let bills run off. Thus it is not surprising that it should be the practice of nearly all banks (especially in New York, where the Call Market is near at hand) to shave their reserve close to the legal minimum.^
' Thus the facts show that, in the banking conditions which now exist in England and the United States, the aggregate of bank-deposits (represented by M in our monetary equation) is a multiple, as nearly as possible constant, of the “ reserves ” of the member banks.®
3. Elsewhere. — Dr. Burgess (op. dt. p. 36) has quoted figures which indicate that the percentage of reserves (including cash) to deposits in several other countries
^ This must mean, I think, reserres held against net demand deposits.
* As regards current practice in the United States in the above respects I have to thank Prof. B. H. Beokhart for useful information. Some interesting figures to the same effect as the above are also given by Dr. Burgess 0^he Reserve Banks and the Money Market, pp. 152-155).
* Subject, of course, to the appropriate arithmetical modification when different reserve-ratios are maintained against time-deposits and demand- deposits respectively and the proportion in which the total deposits are divided between the two is changing.
OH. 25 EATIO OF BANK MONEY TO RESERVES 65
is not very different from what it is in and
the United States. His estunates, based on the years 1925 and 1926, are as follows :
Typical Reserve
United States Member Banks . . .9-5
Ten London Clearing Banks . . .11-5
Four French Credit Companies . .11-5
Swiss private banks «... 8*0
Chartered banks of Canada . .11-0
In Germany, however, the position is protected neither by law nor by convention. In the case of the principal Berlin banks the percentage of their deposits covered by cash and balances with banks of issue has been as follows :
End 1900 . . . 12*5 per cent
„ 1913 • . • 7*4
„ 1924 ... 6*1
„ 1925 ... 5*0
.. 1926 ... 4*4
At other periods of the year the position is still worse. For example, on March 31, 1928, the six prin- cipal Berlin banks held in cash, foreign currency, coupons and balances with banks of issue and clearing banks no more than 2^ per cent of their total deposits, which was also the percentage held by the 83 leading credit banks of the country (including the six Berlin banks) against deposits aggregating £528,000,000. It is true that a smaller percentage of the deposits is repayable at call than in Great Britain (see p. 19 above) ; but the main e:^lanation of this low per- centage seems to be the facility with which cash can be obtained, when it is required, by re-discounting bills at the nearest branch of the Reichsbank, so that a German bank’s portfolio of bills eligible for re-discount consti- tutes its real reserve and the proportion of this port- folio to its liabilities the effective check on expansion. This means, however, that when trade is reviving and the supply of eligible bills consequently expanding.
66 A TREATISE ON MONEY bk. v
it may be unduly easy under the German system to force the Reichsbank into a corresponding creation of additional credit. In other words, the German banking system is somewhat unprotected, compared with other countries, against the development of conditions of profit inflation contrary to the wishes of the Central Bank. The authorities of the Reichs- bank are aware of the danger and have occasionally put some pressure in recent times on the Berlin banks to reform their ways.^ As a last resort in times of emergency, the Reichsbank resorts to the rationing of credit, i.e. it sets an arbitrary limit to the amount of eligible bills which it will discount for a given institu- tion. German banking law appears to an outsider to be in urgent need of amendment in the direction of requiring minimum reserves (preferably in the shape of deposits with the Reichsbank) from the member banks.
(ii.) The Interchangeability of Non-Reserve Bank Assets
People often speak as if the aggregate of invest- ments, discounts and advances, etc., held by the Member Banks was, within fairly wide limits, decided by them by an act of intelligent judgment. In arriv- ing at this decision, the banks are supposed to pay attention to the demands of trade, to the soundness of underlying business conditions, to the general condition of their customers’ accounts, to the pre- valence of speculation, and so forth.
The statistics given above demonstrate, however, that this is a “ vulgar error Apart from the rare occasions of a deliberate change in the conventional ratio, such as that made by the Midland Bank fit
^ Cf. Report of the Commiasioner of the Beichebankt Deo. 1927» p. 37 : '' The Reichsbank objects to this development (».e. the dwindling percentage of reserves) and has thought it advisable to urge the banks to maintain a sufficient reserve
OH. 25 RATIO OF BA2^ MONEY TO RESERVES 67
1927, and from the possibility of the Member Banks being in a . position to influence the amount of their own reserves (a discussion which we are postponing to Book VII.), what bankers are ordinarily deciding is, not how much they will lend in the aggregate — ^this is mainly settled for them by the state of their reserves — ^but in what forms they will lend — in what propor- tions they will divide their resources between the different Hnds of investment which are open to them. Broadly there are three categories to choose from — (i.) Bills of Exchange and Call Loans to the Money Market, (ii.) Investments, (iii.) Advances to Cus- tomers. As a rule, advances to customers are more profitable than investments, and investments are more profitable than bills and call loans ; but this order is not invariable. On the other hand, bills and call loans are more “ liquid ” than investments, i.e. more certainly realisable at short notice without loss, and investments are more “ liquid ” than ad- vances. Accordingly bankers are faced with a never- ceasing problem of weighing one thing against another ; the proportions in which their resources are divided between these three categories suffer wide fluctuations ; and in deciding upon their course they are influenced by the various considerations men- tioned above. When, for example, they feel that a speculative movement or a trade boom may be reach- ing a dangerous phase, they scrutinise more critically the security behind their less liquid assets and try to move, so far as they can, into a more liquid position. When, on the other hand, demands increase for ad- vances from their trade customers of a kind which the banks deem to be legitimate and desirable, they do their best to meet these demands by reducing their ii#estments and, perhaps, their bills ; whilst, if the demand for advances is falling off, they employ the resources l^us released by again increasing their investments.
68 A TREATISE ON MONEY bk. v
It will be sufficient, in illustration of this, to give the hgmres for the proportion of the “ advances ” of the leadii^ English banks, i.e. of their loans to their customers, to their deposits.
It is evident that the proportion of the banks’ deposits employed in advances to their customers is capable of wide fluctuations in conformity with the demands of trade ; but it does not foUow that the banks lend more or lend less in the aggregate.
Percentage of the Amount of their Advances to the Amount of THE Deposits held by the Nine English Clearing Banks ^
|
1921. |
1922. |
1924. |
1925. |
1927. |
1928. |
||||
|
Jan. |
46-7 |
41-2 |
455 |
49-5 |
524 |
52*8 |
62-9 |
||
|
Feb. |
481 |
41-5 |
44-8 |
47-2 |
500 |
63-8 |
54-6 |
54-1 |
54*5 |
|
Mar. |
60-3 |
42-7 |
46-6 |
524 |
552 |
55-8 |
66-6 |
564 |
|
|
April |
49-8 |
429 |
463 |
550 |
66-6 |
66-3 |
66-6 |
||
|
May |
486 |
422 |
462 |
48-7 |
531 |
552 |
663 |
555 |
66-4 |
|
June |
460 |
41-2 |
460 |
47-9 |
62-4 |
63-6 |
64-3 |
53-7 |
56-3 |
|
July |
45-5 |
41-7 |
45-6 |
48-2 |
52 1 |
63 1 |
64-6 |
53-3 |
.55 4 |
|
Aug. |
45-3 |
423 |
461 |
521 |
534 |
53-8 |
55-7 |
||
|
Sept. |
44-4 |
42-8 |
46-2 |
493 |
617 |
541 |
55-2 |
52*7 |
554 |
|
Oct. |
43-7 |
42-9 |
45*8 |
494 |
515 |
53-7 |
63-7 |
53-6 |
55-0 |
|
Nov. |
43-2 |
498 |
51-7 |
538 |
53-8 |
554 |
|||
|
Dec. |
41-9 |
43-6 |
462 |
51-4 |
52-7 |
52-9 |
524 |
54-9 |
|
|
Aver. |
461 |
424 |
45-6 |
48-5 |
61-7 |
538 |
54-5 |
53*9 |
55*3 |
(iii.) How OUGHT THE Rbsebve-Ratio to be fixed ?
In Great Britain the reserve-ratio to be main- tained by the banks has never been fixed by law. During the earlier period of English Joint-Stock Banking it was left to the banks themselves to keep
^ Namely, the Bank of Liverpool, Barclay’s, Coutts’, Glyn Mills, Lloyds, Midland, National Provincial, Westminster and Williams*. The National Bank, which now publishes monthly figures, is excluded from all the tables in this chapter, since figures for it are not available throughout the period.
OH. 25 RATIO OF BANK MONEY TO RESERVES 69
such amount as was dictated to them by considera- tions of prudence and of convenience ; and since the amoimt so kept was never published except in half- yearly balance-sheets which bore little relation to the normal position, there was no motive for them to keep more. In February 1891, however, Mr. Goschen, the Chancellor of the Exchequer (as he then was), made his celebrated speech at Leeds, in which he took advantage of the uneasiness ensuing on the Baring crisis of the previous year and its attendant circum- stances, to argue that the amount of reserves which the banks were in the habit of keeping was insufficient for the safety of the system. The impression pro- duced on public opinion was so great that the banks felt that they must do something. Moreover, Mr. Goschen intimated that he had a scheme of currency reform ready which he proposed to introduce into Parliament. ** Owing ”, as the Economist put it at the time, “ to the antagonism of the bankers and the obstructive attitude of the Gladstonian party ”, nothing came of the scheme, but the banks somewhat strengthened their reserves and began the publication of monthly statements. Since these monthly state- ments were primarily for public consumption, and since there were no legal requirements to satisfy, the practice of “ window-dressing ” them, which has con- tinued to thifl day, was unfortunately adopted. As time went on, the diminished use of cash for daily business, and the progress of amalgamation of the banks amongst themselves^ (which may be said to have begun on an intensive scale with the formation of Barclay’s Bank in 1896), served to reduce the amount of reserves which tlxe banks really required to meet contingencies, whilst the conventional fig^e
» In 1891 there were only four Joint-Stock Banks with £10,000,000— namely, the London and County with London and Western with £27,000,000, the Union with £16,000,000, ana the London Joint-Stock with £13,000,000.
VOL. n ^
70
A XREATISE ON MONEY
BK. y
increased sKghtly on account of the banks, for reasons ,of advertisement and prestige, tending to aim at the ratio which was maintained at first only by the stronger amongst them. By 1926 the position was •such that Dr. Leaf, writing as Chairman of the Westminster Bank, described the ratio as “purely arbitrary
In the United States the position is, as we have seen, rather different. The ratio, var3dng somewhat for differing tjrpes of banks, is laid down by law. Moreover ih.e ratio is much lower for time-deposits than for demand-deposits. Further, till-money does not reckon at all as part of the lawful reserves, so that in the United States any pretension of the amoimt of the reserves being governed by the actual requirements of the banks to meet sudden contin- gencies has been given up.
Why, then, we must ask, should law or custom impose on the banks the maintenance of greater re- serves than they really require ? In part, no doubt — ^like so much else in our currency and credit arrange- ments— this practice is a mere survival of a past state of affairs, and a result of our purporting to conduct a representative money system as though it were a kind of commodity money system. But there is also a sound practical reason Ijdng behind it. The custom of requiring banks to hold larger reserves than they strictly require for till-money and for clearing purposes is a means of making them contribute to the expenses which the Central Bank incurs for the maintenance of the currency.
Under conceivable monetary systems this might not be necessary. But where there is an intemationai standard to maintain, the Central Bank must keep a certain amount of idle resources, not earning inte^st dr not eammg adequate interest, in the forik of •metallic or owei reserves. If, by virtue bfe ti|ie,
* Banking, p. 190.
OH. 25 RA.TIO OF BANE MONEY TO RESERVES 71
pr&ctice adopted, by the Central Bank of recognising certain bills or certain types of security as “ eligible ”, there exist perfectly liquid interest-bearing invest- ments, the member banks would be able, in theory, to cut their reserves very fine-limiting them, indeed, to their requirements for till-money and the day’s clearing. It is reasonable, therefore, in such cir- cumstances that the member banks should be com- pelled to share part of the expense of maintaining the ultimate reserves of the system as a whole — without the existence of which the convenient prac- tice of “ eligible ” bills and “ eligible ” collateral might sometimes break down ; and this is effected by their being compelled to maintain a certain pro- portion of non-interest-bearing deposits at the Central Bank or of non-interest-bearing notes.
The greater part of the central metallic reserves of the modem banking system is still generally provided, it is true, by the active note-issue in the hands of the public ; but the silent trend of evolution is, I think, in the direction of the Government regarding the note-issue as a legitimate source of revenue * and of the member banks being required to provide their quotas towards the maintenance of the central reserves.
So long as the proportion of cash reserves which the member banks are expected to maintain is determined solely by what they need for their own safety and con- venience, there is much to be said for the traditional British practice of leaving it to the banks to de- cide for themselves what the appropriate figure is — especially when a stage has been reached, as is now
iSince the above was written, the new Bank Note Act (1928) in Great Britain has assigned the whole of the profits of the fiduciary issue to the ’Treasury and the whole of the profits of the non-interest-bearing deposite of the member banks to the Bank of Bngland. Since the fiduciary issue is not 6(]^ual to the active circulation in the hands of the public, whilst the . memlw banks’ deposits at the Bank of England are not the whole of their conveaj||ozial reserves, this arrangement is not as logical as it looks. Never- fiie^i|iB the figures work out in practice not very far from the arrangement B^iggested above.
72 A TREATISE ON MONEY sk. v
the oaae in Great Britain, when mushroom bank- ing is extinct. But as soon as further considerations are considered relevant — considerations, such as the above, afEecting the safety and efficiency of the bank- ing system as a whole rather than the interests of individual banks — ^it is by no means so clear that this is the best plan. Post-war developments in the technique of Central Bank management — develop- ments both inevitable and desirable which we shall explain more fully in Book VII. below — ^have brought us to a transitional stage between the two criteria of reserve proportions. Modem methods of Central Bank control, particularly the use of “ open-market operations ”, necessitate that the Central Bank should be ia a position to dispose of resources of a certain size in relation to the scale of the banking system as a whole ; and it is right and reasonable that the member banks should contribute, in the shape of non-interest-bearing deposits held by them with the Central Bank, the bulk of the resources which are required for the safety of the highly economical and efficient system the existence of which enables the member banks to be so comfortable and profitable. Now it is probable that the normal level of bankers’ deposits, required to make the Bank of England strong beyond a doubt and able in all circum- stances to impose its will on the market without straining its own earning capacity unduly, is greater than the member banks strictly require for the safe and convenient transaction of their own business. If this is true, then both what one observes and what one would expect concur in raising a doubt, whether it is desirable to go on leaving the matter entirely to the discretion of the member banks.
If we were to go back fifty years, we should probably find that the reserves of the member banks were at that time no Idgheir than were required by their own safety and iwffivemenqe— except, perhaps.
OH. 25 RATIO OF BANK MONEY TO RESERVES 73
on balance-sheet dates. Between balance-sheet dates there was indeed no sufficient motive for locking up resources unnecessarily. The growth of amalgama- tions, the diminished use of cash by the public, and the general progress of the speed and efficiency of banking machinery generally during the intervening period, have tended to dimmish the proportion of reserves which is strictly necessary. Our banks, how- ever, owing to their conservatism, and owing to the unwillingness of any one of them to appear to be the first to lower its reserve ratio below those of its neighbours, have not taken full advantage of this, with the result that we have arrived at the p.viating rather artificial position in which the banks voluntarily maintain, for reasons which are probably not very clear to their own minds, somewhat larger reserves than they really require. Nevertheless the natural desire to make as much profit as is compatible with safety and with appearances has caused them to employ such a very undesirable ruse as the “ window- dressing ” operations described above, whereby their real proportions are appreciably lower than their apparent proportions. Nor is there, in fact, any substantial obstacle to their nibbling at their con- ventional ratios as the result of concerted or s)mi- pathetic action continued by slow degrees over a period.
The possibilities of the situation may be illustrated by an interesting episode which occurred in 1927-29. The Midland Bank had, as pointed out above, main- tained for some years past a reserve proportion a good deal higher thfl.Ti those of its competitors. It is not obvious that this had really been worth while^ from its own point of view. Accordingly, beginning in the latter part of 1926, a gradual downward movement became apparent in the Midland Bank’s proportion from about 14"6 per cent in 1926 to about 11 ‘5 per cent in 1929. Since the Midland Bank’s deposits
74
A TBEATISB ON MONEY
BK. ▼
amounted during this period to at least £370,000,000, this whs e<^uivalent to a release of more than £11,000,000 m cash, and in fact enabled the banks as a whole to increase their deposits (and their advances) by about £100,000,000 without any increase in their aggregate reserves. Thus the result of the total operation is either an acquiescence on the part of the Bank of England in a substantial increase in the volume of bank-money without any iucrease in the volume of reserves, or a depletion of the Bank of England’s resources by £11,000,000 — ^the Midland Bank gaining and the Bank of England losing the interest on this sum.
Now, as it happened, this relaxation of credit was in the particular circumstances greatly in the public interest. For an increase in the total of member bank deposits was needed to balance an increase in their fixed deposits — an increase which the Bank of England, if left to itself, might not have allowed in a sufficient degree. Thus the Midland Bank was, in fact, not only acting in its own interest along lines which no one could criticise (for its ratio of reserve was still higher than the average of the other banks), but was also materially aiding the general situation. Neverthe- less such an expansion of the resources of the member banks should not, in any sound modem system, depend on the action of an individual member bank, even though it remains within the power of the Central Bank to counteract its effect by appropriate measures. For we ought to be able to assume that the Central Bank will be at least as intelligent as a Member Bank and more to be relied on to act in the general interest.
I conclude, therefore, that the American system of regulating by law the amoimt of the Member Bank Eeserves is preferable to the English system of depend- ing on an ill-defined and somewhat precarious con- vention.
Moreover, the possibility of all or some of the “ Big
OH. 25 RATIO OF BANK MONEY TO RESERVES 76
Five ” nibbling at their proportions is not the only reason for regularising the present system so as to render the control of the Bank of England unchallenge- able. At present the reserves of the member banks consist, as pointed out above, not only of their deposits at the Bank of England but also of what they call their “ cash ”, and it is open to them to vary freely the proportions in which their total reserves are divided between the two categories. Normally the “ cash ” consists of Bank (or, formerly. Treasury) Notes, but it might also consist of gold or, conceivably though never yet in practice so far as I am aware in the case of English banks, in deposits in a Central Bank abroad. Now each of these options contains a possibility of surprise and inconvenience for the Bank of England, as follows :
(1) So long as the Bank of England calculates its reserve “ proportion ” by its present formula, the position can be artificially influenced within wide limits if the member banks turn their “ cash ” into deposits at the Bank of England or vice versa. Indeed the adequacy or inadequacy of the existing fiduciary issue in its effect on the Bank’s free reserves largely depends on the policy of the member banks in the above respect.
(2) If the member banks were to hold a part of their reserves in actual gold, the effect might be to diminish either the profits or the free gold of the Bank of England below the point of safety. A clause was, indeed, wisely inserted in the Bank Note Act of 1928 to provide against this contingency, by which the Bank of England is given the power of compulsory purchase of any gold so held within this country. This clause does not, however, prevent a member bank from reckoning as part of its reserves gold held abroad or in transit ; and cases are beheved to have occurred recently in which a bank has taken advantage of this liberty.
76
A TEEATISE ON MONEY
BK. T
(3) There is nothing to prevent an English member bank from holding a part of its reserves “ earmarked ” with the Federal Eeserve Bank of New York. Froth the point of view of safety, and even of convenience, such deposits would be practically “ as good as gold But it is obvious that this practice would greatly diminish the Bank of England’s power of controlling the volume of Bank-money in England by depriving it of the necessary volume of resources and profits. I am not aware that this option has ever yet been exercised. But the risk is quite a real one. For in the case of Continental member banks the practice of holding some part of their reserves at foreign centres is very common. And sometimes — ^for example, in Switzer- land, as I am informed — ^the practice is carried so far as to impair very greatly the authority of the Central Bank. In Germany at the present time the large proportion of their liquid resources which the leading member banks hold in foreign centres is cap- able of interfering seriously with the hegemony of the Beichsbank.
Furthermore, there are a great many banks, in- cluding London branches of overseas banks, other than the ten Clearing Banks, publishing no monthly or average figures, which take advantage of the facilities of the London Money Market without necessarily con- tributing proportionately to the resources of the Bank of England.
Finally, there are certain important technical advantages which are impracticable without a legal reserve proportion, the full significance of which will appear in subsequent pages, namely a differentiation between the reserves required against Demand and Time Deposits respectively, and the power to the Bank of England to vartf the legal proportions from time to time within prescribed limits.
The question of the principle of establishing legal reserve proportions for all banks is distinct from the
OH. *5 RATIO OF BANK MONEY TO RESERVES 77
question as to what these regulations should be. I suggest, however, something on the lines following ;
(i.) The prescribed proportions to cover all bank » deposits payable in sterling in Great Britain.
(ii.) Deposite payable on thirty (or, perhaps, four- teen) days’ notice or more to reckon as Time Deposits.
(iii.) The prescribed proportions to relate to daily averages over monthly periods.
(iv.) The cash reserves to consist either of Bank of England Notes or of deposit balances at the Bank of England, but not less than 40 per cent of the total to be held in the latter form.
(v.) The normal prescribed proportions of cash reserves to deposits to be as follows :
Demand Deposits • . 15 per cent
Time Deposits . • • 3 ,,
(vi.) The Bank of England to have the power on thirty days’ notice to vary the prescribed proportions to a hgure between 10 and 20 per cent in the case of Demand Deposits and to a figure between 0 and 6 per cent in the case of Time Deposits.
These regulations would greatly strengthen the power of control in the hands of the Bank of England — ^placing, indeed, in its hands an almost complete control over the total volume of Bank-money — ^without in any way hampering the legitimate operations of the Joint-Stock Banks. The importance for this purpose of (vi.) will transpire more fuUy in Chapter 32 (vi.).
Thus in countries where the percentage of reserves to deposits is by law or custom somewhat rigid, we are thrown back for the final determination of M, the Volume of Bank-money, on the factors which determine the amount of these reserves. How far
^ The legal deSnition of a “Bank” presents some— not iMuperable- difficulties which are legalp rather than economic, in character.
78
A TREATISE ON MONEY
BK. T
are they determined by the Central Bank, with the behaviour of the Member Banks necessarily following in a more or less passive way ? And how far, if at all, is the amount of the reserves mfluenced by the Member Banks themselves acting on their own initiative ? Only in so far as the latter is true need we modify the statement that the Member Banks do no more than determine in what form they will lend, and not how much they will lend altogether.
The answers to these questions we must postpone to Book VII.
CHAPTER 26
THE ACTIVITY OP BUSINESS
(i.) The Influence op Business Activity on the Velocity of the Business-deposits
It has long been held that, when business is brisk, cash is turned over quicker. Indeed the velocity may increase in such circumstances more than in proportion to the transactions ; so that an increase of transactions, instead of being associated with a falling price-level as it would be if “ other things remained equal ”, is in fact regularly associated in the minds of busi- ness men with a rising price-level. We have seen in Chapter 24 that there are obvious reasons why this conclusion should be plausible. When business is brisk the whole process of exchange is accelerated, and this acceleration diminishes the average amount of time for which it is necessary or convenient to retain cash between transactions : receipts and disburse- ments follow hotter on one another’s heels. Moreover, when markets are good, satisfactory sales can be fore- seen with greater confidence, so that, at least in the belief of the business world, there is less need for pro- vision against the contingencies of frozen stocks or debts which cannot be collected. There is also the further reason that when business is brisk there may be more strain on the resources of traders for the pro- vision of working capital, so that they economise their holdings of cash to the utmost possible extent. The
79
80
A TREATISE ON MONEY
BK. T
argument may nev^heless rely too much on the mere fact that in actual esmerience brisker business is generally accompanied by a rising, not by a falling, price-level. For this fact may be better explained by over-investment causing the second term of the Fundamental Equation to increase, than by influences which directly affect true velocities.
The conclusion, that an expansion in the real volume of transactions tends to be accompanied by an increase in velocity more than in proportion, is a view often held. But there is a class of writers who maintain that the variations in the two factors tend to be equal — i.e. that there is an almost strict proportionality in their movements. Professor Angell has quoted Wieser as an early adherent of this doctrine, and Baron Charles Moume as a recent advocate of it in France.^ The exposition on these lines most familiar to contemporary students® is, however, that of Mr. Carl Snyder of the New York Federal Reserve Bank.®
Mr. Snyder’s conclusions are empirical and are based on a computation of the variability of the ratio of American Bank Debits {i.e. Bank Clearings) to Demand Deposits in past years, or, in my terminology, of the velocity of the Cash-deposits. He compares this not with the volume of “ transactions ” but with that of “ trade ”. Since Mr. Snyder’s tables represent the fullest investigation yet made of the
^ The Theory of Internolional Prices, p. 327 and p. 279. This volume i» a most valuable source for the history of monetary doctrines.
* Professor Angell (op, ciU p. 180) quotes Fk’ofessor Worthing (“ Prices and the Quantity of Circulating Medium/’ Quarterly Journal of Economics, 1923) as the first recent American writer on these lines.
* See his articles ** New Measures in the Equation of Exchange ” (American Economic Review, 1924), ** A New Index of Business Activity ” (Journal of the American Statistical Association, 1924), and ‘‘ A New Index of the Ceneral Price Level from 1875 ” (Journal of tlU American Statistical Association, 1924) ; and his book Business Cycles and Business Measure- ments, of wUch chapter vii. summarises these articles. Mr. Snyder seems to claim somewhat less validity for his generalisations in his book than in the articles which preceded it.
OH. 26
THE ACTIVITY OF BUSINESS
81
variability of tbe velocity of the Casb-deposits, it will be useful to reproduce them, whether or not we accept the generalisation which he bases on them.
The following table shows the average annual per- centage variability from the normal velocity (defined as the ratio of Clearings (or Debits) to Demand De- posits) for 141 cities, including New Jork City, as compared with the volume of “ trade ” :
|
Annual Average. |
Velocity of Circulation. (100 = 1919-1925 average.) |
Volume of U.S. Trade. |
|
1919 |
102 |
104 |
|
1920 |
102 |
101 |
|
1921 |
94 |
92 |
|
1922 |
98 |
102 |
|
1923 |
99 |
108 |
|
1924 |
99 |
105 |
|
1925 |
105 |
111 |
|
-- ■ ■■ ■■■■ |
Even if we take this table at its face value, there does not seem to be very much in the alleged correlation, except as regards the drop m both of them in the slump of 1921. If, however, we take the velocity of the Cash-deposits for the whole country , iMtead of 141 cities, the alleged conelation less impressive, as is shovm m the correlation whatever l^avmg P
1922. If, on the other hand, we take the fig^®® for New York City, also shown fair degree of correspondence as regards the dtrectum
of the movements :
[Tablk
82 A TREATISE ON MONEY bk.t
(100=1919-1926 average)
|
Velocity of Cash-deposits. |
Volume of Trade, U.S. |
||
|
For Whole Country. |
For New York City. |
||
|
1919 |
no |
98 |
104 |
|
1920 |
106 |
96 |
101 |
|
1921 |
94 |
89 |
92 |
|
1922 |
99 |
98 |
102 |
|
1923 |
98 |
103 |
108 |
|
1924 |
97 |
103 |
106 |
|
1925 |
96 |
114 i i |
111 |
Thus it might be that if Mr. Snyder had alleged a correspondence between the velocity of the Business- deposits, instead of the velocity of the Cash-deposits, and the volume of Trade, he would have been nearer the truth. On the other hand, this statement also might be too loose ; for, as we have seen, the Business- deposits are employed not only for the handling of “ trade ” but also for hnancial and Stock Exchange transactions, the real volume of which may sometimes move with that of trade, but may often not.^ Thus the statistical foundation of the theory is scarcely sufficient to support it.
(ii.) The Relation between the Banc Clearings AND THE Volume op Trade
Can we obtain out of the Bank Clearings a satis- factory index to the volume of trade or output ? It
^ Mr. Snyder points out that the influence of Stock Exchange transactions on the Clearings is notg even in New York, so great as might have been supposed, since 80 per cent or more of the transactions on the New York Stock Exchange is cleared by the Stock Exchange’s own clearing corporation, (Review of Economic StaUeiics^ Feb. 1928, p. 41). Nevertheless, allowing for cheques between brokers and their clients, the influence must be very considerable.
OH. *6
THE ACTIVITY OF BUSINESS
depends on whether we can isolate transactions relat- ing to the Industrial Circulation from those relating to the Financial Circulation, or at least obtain an approximate index of changes in the former.
Let us write
B (the volume of Bank Clearings) = QiRi + QjRs
where Ri = volume of Wages and current production of goods (finished and imfinished) traded,
Rj = volume of Bonds, Shares, Real Estate, and other financial obligations changing hands,
each weighted in proportion to its cash-using im- portance ; and Qi and Q*, the price-levels of each of these, weighted on the same system.
It is then seen that B is only a reliable index to QiRi in so far as Q2R2 is small compared mth QiRi, or in so far as it varies in the same way. For the bank clearings of the country as a whole tee is no justification for regarding Q2R2 as small. But it is sometimes possible to isolate a particular sec- tion of the bank clearings for which this assumption is more nearly justified. For example, it used to be a common practice to exclude the London btock Exchange clearing days from the tel clearmgs m order to obtain a figure which might be a useful indication of the state of British trade ; but smce there are now no special settling days m the London gilt-edged market, this correction is no longer s^ient The most serviceable means of partially 8
Q,Ri in England at the present time is to take the Country^Clearings and the Provmml (namely, the inter-Bank Clearmgs in t e ^
vincial towns), since by this means preluded
purely financial transactions are probab y »
inasmuch as they pass through London.
Nor does vary in the same i^y
The price-index of shares, may, it 1 j
84
A TREATISE ON MONEY
BK. V
move in the long ran with Qi, the price-index of wages and commodities, and an exceptional activity of the share market will often accompany an exceptional activity of trade. But at times of a rapidly changing price -level the index of gilt-edged securities will tend to move, not in the same, but in the opposite direction from Qi. This can be illustrated from the post-war price and banking statistics of England. In the following table A is the price of fixed-interest securities and B is the price of industrial Ordinary Shares, as calculated by the London and Cambridge Economic Service ; C is the Consumption Index Number for Great Britain (see Vol. i. p. 62) ; and D is the Board of Trade Index Number of whole- sale prices :
(1913= 100)
Since, moreover, the activity of financial trans- actions is far more variable than that of industrial transactions, it seems clear that the total volume of bankers’ clearings will be almost worthless as an index to the volume of trade transactions arising out of current production and consumption, unless we ex- clude St^k Exchange transactions. This conclusion is confirmed by the figures of the total clearings. The
OB. 26
THE ACTIVITY OP BUSINESS
85
total bank clearings (B), corrected for price by the Board of Trade* Index-Number of wholesale prices (D), give the following result :
B/D.
71
67
77
101
100
104
110
117
127
136
Now if we were to argue from the Fisher equation MV = PT that, since B MV, B/D is an index to the volume of trade, we should arrive at absurd results. For we should find that the so-called “ volume of trade ” was 50 to 100 per cent greater during the slump of 1921-22 than it had been during the boom of 1919-20 ; that it was nearly 50 per cent greater in 1925-26 than before the war ; that it was 7 per cent greater in the great strike year, 1926, than in 1925 ; and that it is nearly double to-day what it was before the war. This is a rednctio ad absurdum of calculations on these lines. Clearly the inclusion of all kinds of financial transac- tions and the correction of the price-level by reference to the Wholesale Standard have made enough differ- ence to vitiate the results completely.
Let us, however, consider whether, if we confine our calculations to the Coimtry and Provincial Clear- ings in which Stock Exchange transactions do not play an important part, the result of correcting their money volume by an appropriate index-number of prices will conform with what we know from other sources about the fluctuations in the volume of trade.
VOL. n o
86
A TREATISE ON MONEY
BK. T
Wliat is the most appropriate index-number for this purpose ? I believe that an* index made up mainly of wholesale prices for a period three months earlier and of wages for the current period would be serviceable, though something much better could be contrived by the Board of Trade with the whole apparatus of Government statistics behind it. For present purposes, however, we may take the Con- sumption Index, the method of compilation of which has been explained above (p. 61), and then obtain a “ Clearings Index ” of the Volume of Production by dividing the Country and Provincial Clearings (com- piled as on p. 40 above) by this Consumption Index.
If now we compare the annual average of the Clear- ings Index, thus arrived at, with Mr. Rowe’s Index of Production, compiled for the London and Cambridge Economic Service, which is mainly based on the con- sumption of raw materials, and with Mr. Rokeling’s ^ Index of the Volume of Employment, the result is as shown in the table opposite.
Generally speaking, the agreement between the fluctuations of these entirely independent calculations — ^they have not a single direct constituent the same — ^is remarkable. I believe that the most comprehen- sive Index of Production for Great Britain might be obtained by taking an average of these three indexes, asshownin the last column opposite. For the somewhat slower rate of recent growth in the Clearings Index compared with the other two, more purely industrisd, indexes may be a necessary corrective where we are dealing with the economic activities of the country as a whole ; whilst the recent tendency of the Raw Material Index to outstrip the Employment Index may reflect an increase of efficiency.
I This index is part of an Index of Frodnction for Great Britain com- piled by lir. Bokel^ {Economitt, Oot. 6, 1928) as follows :
1920 . . . 101 1928 . . .93 1926 . . .90
1921 ... 75 1924 . . . 100 1927 . . .106
1922 . . 88 ] 1926 ... 99
OE. 26
THE ACTIVITY OF BUSINESS
87
(1924»100)
|
Clearings Index (Keynes). |
Employment Index ♦ (Bokeling). |
Raw Material Index (Rowe). |
Composite Production Index (average of the three). |
|
|
1920 |
101s |
103 |
104*5 |
103 |
|
1921 |
78 |
89*5 |
75*5 |
81 |
|
1922 |
91-6 |
93*5 |
89*5 |
91*5 |
|
1923 |
97 |
97 |
91 |
95 |
|
1924 |
100 |
100 |
100 |
100 |
|
1925 |
101 |
101 |
101 |
101 |
|
1926 |
95 |
95*5 |
90 |
93*5 |
|
1927 |
101*5 |
104*5 |
110 |
105 |
|
1928 |
102*5 |
104*5 |
108*5 |
105 |
|
1929 |
104 |
106 |
116*2 |
108*7 |
* These figures have been kindly supplied to me by Mr. Rokeling, and are slightly revised from those published by him in the Economist, They represent A + B-C-D-E, where A = number of agricultural labourers, number of insured workers aged 16>64, number of insured workers unemployed, D = number of insured workers absent from sickness, accident, etc., E = number of persons directly involved in trade disputes. They make no allowance for increased efficiency on the one hand, or for short time on the other.
An Index of the Volume of Trade based on the Country and Provincial Clearings corrected for price may indeed be superior in some respects to indexes based on statistics of production and output ; since, as Mr. Snyder has pointed out in dealing with similar Statistics for the United States,^ most indexes of pro- duction are overweighted with the basic commodities, whereas the volume of cheque transactions offers a more comprehensive picture embracing all those numerous miscellaneous activities which, whilst they are too small individually to be caught in the mesh of the statistician’s net, are nevertheless very im- portant in the aggregate. It is a further advantge of such an index that the figures are available
^ Business CjfcUs and Measurements^ p. 79.
88 A TREATISE ON MONEY bk. t
promptly without the necessity for any special cal- culation.
Mr. Snyder’s calculations for the United States confirm the value of the Bank Clearings, outside the chief financial centre, corrected for price (mean- ing, however, by “ price ” — as in my own calcula- tions above — ^not the Wholesale Index but the Con- sumption Index) as an index of the Volume of Trade. The following table is based on a three months’ moving average of the Bank Debits for 140 cities out- side New York City ^ corrected for price by Mr. Snyder’s “General Price - Level ”, and on his independently compiled index of the Total Volume of Trade.
|
Bank Debits outside N.Y.O. corrected for price. |
Volume of Trade. |
|
|
1919 . |
105 |
104 |
|
1920 . |
102 |
101 |
|
1921 . |
92 |
92 |
|
1922 . |
101 |
102 |
|
1923 . |
106 |
108 |
|
1924 . |
102 |
105 |
|
1925 . |
110 1 |
111 |
(iii.) Statistical Summary
The estimates of the preceding chapters are summed up in the following table. Although the reader must not overlook the large element of guess- work, approximation and probable error which this table involves in the present deplorable state of our banking and other statistics, it has, I think, some value
^ The Harvard Economic Service now exclude, in addition to New York City, the Debits of seven other important cities which are more or less financial, as well as business, centres, for the purposes of their B curve, which is intended to refiect business conditions.
OH, 26
THE ACTIVITY OF BUSINESS
89
as a general indicator of the variability of the different factors involved.
I have no doubt that in a few years’ time such guesses will appear hopelessly amateurish and in- accurate, and will be replaced by scientific estimates. But by indicating what figures it would be in- teresting to know and by making doubtful attempts at their value, I may stimulate others, who are in a position to get better dxUa, to correct me. The compilation of official figures has generally been pre- ceded in England by the rash and imperfect efforts of individuals.
(1924=100)
|
Total Deposits,* M. |
Proportion of Cash - deposits to Total Deposits.t w. |
Cash- de]X)8lt8, |
Velocity, v.t |
Industrial Cash Turnover. MVti’.§ |
Volume of Production, 0. (Rowe and Kokeling.)* |
|
|
1920 |
100 |
Ill |
111 |
137 |
162 |
104 |
|
1921 |
108 |
100 |
108 |
93 |
101 |
82 |
|
1922 |
106 |
100 |
106 |
91 |
96 |
91-6 |
|
1923 |
100 |
102 |
102 |
94 |
96 |
94 |
|
1924 |
100 |
100 |
100 |
100 |
100 |
100 |
|
1925 |
99 |
98 |
97 |
104 |
101 |
101 |
|
1926 |
100 |
96 |
96 |
97 |
93 |
93 |
|
1927 |
103 |
96 |
99 |
99 |
98 |
107 |
|
1928 |
106 |
94*5 |
100 |
98 |
98 |
106-5 |
|
1929 |
108 |
93 |
100 |
98 |
98 |
111 |
* As computed on p. 10. t ^ computed on p. 10.
t As computed on p. 40.
§ This has brought us back again to the index of the Country and Pro- vincial Bank Clearings as computed on p. 40, since we arrived at V by dividing this by Mw,
If we assume that the variations in the amount and in the velocity of the cash-deposits applicable to in-
^ 1 have taken the mean of the indexes of Rowe and Bokeling given on p. 87, since to introduce the Clearings Index in this context would be begging the question — though in fact it would not make much difference to the result.
90
A TREATISE ON MONEY
BK. T
come transactions and current business, apart from financial faransactions, are approximately measured by Mw and V respectively, i.e. that the turnover of the Industrial Circulation is measured by the Country and Provincial clearings, then, if there is no Profit Inflation or Deflation, MVw should furnish us with an index to the aggregate cost of current output, and MVw/0 should give us an index of the price- level of current output. If, therefore, our statistics could be relied on, the divergence between MVw/0 and the actual price-level of output would be a partial indication of the divergence between cost and value, i.e., of the degree of Profit Inflation or Deflation. For with Profit Inflation (or Deflation) present we should expect the actual price-level to be higher (or lower) than the price-level indicated by MVm>/0. As a first approximation, therefore, for deducing the degree of Profit Inflation (or Deflation) let us set side by side the theoretical price-level given by MVw/0, the Con- sumption Price-Index (Vol. i., p. 61) and the Whole- sale Price-Index (Board of Trade Index) :
|
MVw/0. |
Consumption Price Index. |
Wholesale Price Index. |
Profit Infiation or Defiation.* |
|
|
1920 |
146 |
150 |
186 |
127 |
|
1921 |
123 |
132 |
120 |
98 |
|
1922 |
106 |
106 |
96 |
91 |
|
1923 |
102 |
99 |
96 |
94 |
|
1924 |
100 |
100 |
100 |
100 |
|
1926 |
100 |
101 |
. 96 |
96 |
|
1926 |
100 |
98 |
89 |
89 |
|
1927 |
91 |
96 |
86 |
04-6 |
|
1928 |
92 |
96 |
86 |
94'6 |
|
1929 |
88 |
94 |
82 |
93 |
♦ Given by dividing the Wholesale Index by MVw/0.
Unfortunately the Consumption Index is not appropriate for the present purpose, because our Index
OH. s6
THE ACTIVITY OF BUSINESS
91
of Outj)ut excludes services,^ whereas we have ex- pressly weighted the Consumption Price Index so as to include them. On the other hand, the Wholesale Price Index is too much concerned with imported raw materials and too little with finished goods. If, how- ever, failing anything better, we take the latter as our index of the price-level of output exclusive of services, the degree of Profit Inflation (or Deflation) will be very roughly indicated by dividing this index by MVw/0, as in the table above. By taking 1924 as our base year, we have, it must be noticed, tacitly assumed that this was a year of equilibrium in which neither Profit Inflation nor Profit Deflation was present in any sensible degree.
The possible sources of error have piled up by now to such an extent that I am scarcely entitled to attach any serious value to the final column of the above table. But at least it does not contradict the con- clusions of the fundamental argument of Book III. For the degree of Profit Inflation or Deflation, as here shown, is quite in accordance with our reasonable ex- pectations. The boom of 1920, persisting into the first half of 1921 but balanced by the slump of the latter half of that year, the continuing slump of 1922-23, the acute distresses of 1926, and the long drawn out profit deflation of 1927-29 display them- selves appropriately.
^ This also vitiates MVwlO for the above purpose to some extent ; but not too much, since the amount of monetary turnover corresponding to the remuneration of services is relatively small.
BOOK VI
THE RATE OF INVESTMENT AND ITS FLUCTUATIONS
CHAPTEK 27
FLUCTUATIONS IN THE RATE OF INVESTMENT
I. FIXED CAPITAL
When there is a disequilibrium between savings and investment, this is much more often due to fluctua- tions in the rate of investment than to sudden changes in the rate of savings, which is, in normal circum- stances, of a fairly steady character. To understand, therefore, the genesis and the severity of the dis- equilibria which we have analysed in Volume i., it is chiefly necessary to consider what causes the rate of investment to fluctuate and to estimate the order of magnitude of such fluctuations. In this chapter and the two following, we shall treat in turn the causes and the degree of fluctuations of investment in Fixed Capital, Working Capital and Liquid Capital. These chapters are in the nature of a digression, which is doubtfully in place in a Treatise on Money- but has to be included because the fluctuations in the rate of investment have not been treated, sufficiently for my ptLTpose, elsewhere.
In the case of Fixed Capital it is easy to understand why fluctuations should occur in the rate of invest- ment. Entrepreneurs are induced to embark on the production of Fixed Capital or deterred from doing so by their expectations of the profit to be made. Apart from the many minor reasons why these should fluctu- ate in a changing world. Professor Schiimpeter’s ex- planation of the major movements may be unreservedly
96
96
A TREATISE ON MONEY
BK.TI
accepted. He points to the innovations made from time to time by the relatively small number of ex- ceptionally energetic business men — their practical applications of scientific discoveries and mechanical inventions, their development of new forms of in- dustrial and commercial organisation, their introduc- tion of unfamiliar products, their conquests of new markets, exploitation of new resources, shifting of trade routes, and the like. Changes of this sort, when made on a large scale, alter the data on which the mass of routine business men have based their plans. But when a few highly-endowed individuals have achieved success, their example makes the way easier for a crowd of imitators. So, once started, a wave of innovation gains momentum.” ^
It is only necessary to add to this that the pace, at which the innovating entrepreneurs will be able to carry their projects into execution at a cost in interest which is not deterrent to them, will depend on the degree of complaisance of those responsible for the banking system. Thus whilst the stimulus to a Credit Infiation comes from outside the Banking System, it remains a monetary phenomenon in the sense that it only occurs if the monetary machine is allowed to respond to the stimulus.
Fluctuations such as those just considered are due to a change in the readiness to invest at a given rate of interest. Besides these we also have fluctuations in the rate of investment due to a change on the side of the rate of interest. In a manner which we have already examined in Chapter 13, a change in the rate of interest will afiect the advantages of owning a particular piece of Fixed Capital so long as the income derived* from it remains unchanged. But there will be no reason for this income to be changed until the supply of Fixed Capital has been changed relatively
^ This conyenient summaiy of Professor Schnmpetor's views is taken from Wesley Mitchell, Biuintat Cytles, p, 21.
OH. 27
FIXED CAPITAL
97
to the demand for it. The process of changing the supply of Fixed Capital, until the income derived from it is again in equilibrium with the rate of interest, amounts, however, to the same thing as a change in the rate of investment.
Thus, whenever the rate of interest changes for reasons other than a change in the demand-schedule for the use or enjoyment of Fixed Capital, it is reason- able to expect a change in the rate of investment.
It is worth while to note in passing that the transi- tion, except when the change required is small, is likely to be easier in respect of an increase than of a decrease in the supply of Fixed Capital. For the rate of obsolescence of existing Fixed Capital sets a limit to the rate at which the total supply of it can be decreased ; and since different kinds of Fixed Capital will be affected unequally (for there is not the same elasticity of demand for all of them), the actual maxi- mum rate of decrease will be determined within still narrower limits.
(i.) The Statistical Indications
When we turn, however, to the relevant statistics to find some exact measure of the degree of these fluctuations, we find that they are few and unsatis- factory. There is no single set of figures which measures accurately what should be capable of quite precise measurement — ^namely, the rate at which the community is adding to its investment in Fixed Capital. The best we can do, therefore, is to take a number of partial indicators and to judge as well as we can from their combined results.
It might have been supposed that the volume of new issues on the investment market would provide a reasonably accurate index. But this total does not adequately represent the rate of investment in houses, which are largely financed in other ways than through
98
A TREATISE ON MONEY
BK. TI
the new issue market ; yet house-building is probably larger than any other one kind of investment. On the other hand, many so-called new issues merely represent the transfer of existing assets from one party to an- other ; whUst in the case of holding, finance and in- vestment companies there may be a large element of duplication. Moreover, even in the case of those classes of investment which are mainly financed by bond issues, there is a lack of synchronisation between the date at which the bonds come on to the market and the date at which the corresponding investment takes place. Thus bond issues are not a good index of the short-period fluctuations in the types of invest- ment which they finance. The fluctuations in the rate of investment may, therefore, be either greater or less than those in the rate of issue. Nevertheless, fluctuations in the volume of new issues is one of the partial indicators of which we must take account.
Much the greater part — ^probably not less than three-quarters — of the Fixed Capital of the modem world consists of Land, Buildings, Roads and Rail- ways. Thus — turning from the financial side to the physical side — any statistics directly bearing on the activity of productive effort in these directions will be of some assistance. For the United States there is a set of statistics of this character which is very significant for our purpose — namely, the monthly values of Building Permits. Since the term “ Build- ing” includes in this connection Constraction and Contracting Works generally (including, I think, roads, sewers and the like), these figures go a long way towards giving us what we want. In Great Britain we have no comparable figures ; but the volume of employment in the Building and Contracting in- dustries, and the incomplete quarterly returns of building published in the Labour Gazette, give some indication of the volume of investment in these directions.
OH. 27
FIXED CAPITAL
99
Since there are to-day comparatively few kinds of investment in Fixed Capital which do not employ a certain amount of iron and steel, some writers ^ have argued that the consumption of these materials — for wmch fairly accurate figures are available over a long term of years — ^affords a reliable barometer of the rate of fixed investment. Since, however, technical methods change and different t3^es of investment, even where they consume iron and steel, consume them in widely varying proportions (compare, for example, house- building with ship-bmlding), it is better not to exag- gerate the value of this indicator by itself, but to be content with including it as one amongst several.
The result of statistical inquiries along these various lines {vide Wesley Mitchell’s Business Cycles, passim, for a summary of the results) cannot, unfortunately, be reduced to tabular form, so as to allow us to make any satisfactory numerical estimate of the order of magnitude of the fluctuations in the rate of invest- ment in fixed capital between one year and another. They are sufficiently definite, however, to make it clear that the fiuctuations are substantial and that they are correlated with the phases of the Credit Cycle in quite as high a degree as our theory would lead us to expect.
(ii.) Theories op the Credit Cycle based on
Fluctuations in Fixed Capital Investment
Indeed the fact of fluctuations in the volume of fixed investment and their correlation with the Credit Cycle has long been familiar, and has been made by numerous writers the basis of a solution of the Credit Cycle problem. Whilst — ^if my theory is right — these solutions have been incomplete, particularly through their neglect of fluctuations in working capital, most of them, even when they have appeared to reach opposite results, seem to me to have had hold of some
^ Especially Hull and Spiethoff.
100
A TREATISE ON MONEY
BK. TI
part of the truth. Some of them have attributed the Cycle to under-saving and some have attributed it to over-investment. Take, for example, the following contrast made by Professor Wesley Mitchell (op. cU. p. 151) : “ Professor Tugan-Baranovski contends that crises come because people do not save enough money to meet the huge capital requirements of prosperity. Professor Spiethofi holds that crises come because people put their savings into too much industrial equipment and not enough consumption goods.” If we interpret the first of these statements to mean that Saving falls short of Investment and the second to mean that Investment runs ahead of Saving, we see that the two authorities mean essentially the same thing — ^and also the same thing that I mean.
Accordingly I find myself in strong sympathy with the school of writers ’ — Tugan-Baranovski, Hull, Spiethoff and Schumpeter — of which Tugan-Baranov- ski was the first * and the most original, and especially with the form which the theory takes in the works of Tugan-Baranovski himself, and of two American amateur-economists (cranks, some might say), Rorty ® and Johannsen.* The fault of Tugan-Baranovski lay
^ For a most excellent short summary of the views of this school see Wesley Mitchell, Business Cycles, pp. 20-31.
* His theory was originally published in Russian in 1894.
• Colonel Rorty 's theory of “ over-commitments ” is more directly applicable to cycles due to the growth of Working Capital at a rate in excess of Savings. But it has the merit of recognising that it is of the essence of the problem in these cases that the purchasing power is created as soon as expansion begins, whereas the goods come along at a later date governed by the duration of the productive process.
^ Mr. N. Johannsen originally published his theory in A Neglected Point in Connection with Crises, 1908, and followed it up with pamphlets in 1925, 1926 and 1928. His doctrine of Impair Savings ”, t.e. of savings withheld from consumption expenditure but not embodied in capital expenditure and so causing entrepreneurs who have produced goods for consumption to sell them at a loss, seems to me to come very near to the truth. But Mr. Johannsen regarded the failure of current savings to be embodied in capital expenditure as a more or less permanent conation in the modem world due to a saturation of the Capital Market, instead of as a result of a temporary but recurrent failure of the banking sjrstem to pass on the full amount of the savings to entrepreneurs, and overlooks the fact that a fall in the rate of interest would be the cure for the malady if it were what he diagnoses it to be.
OH. 27
FIXED CAPITAL
101
in his holding, or at any rate implying, that savings can in some way accumulate during depressions in an uninvested form and that this accumulated fund is then gradually used up during booms, and also in his suggesting that this failure of savings to become materialised in investments at a steady rate is due to the unequal distribution of wealth instead of to Schumpeter’s “ innovations ” in conjunction with a failure of the banking system to respond in such a way as to preserve the desirable degree of stability. But none of these writers clearly apprehend the direct effect on prices of disequilibria between savings and investment and the part played by the Banking System. The pioneer work at this point is due to Mr. D. H. Kobertson {Banking Policy and the Price Level). Moreover, laclang a version of the Quantity Theory of Money applicable to the problem of Credit Cycles, they have not got to the root of the matter or perceived that Cycles due to a growth of Working Capital are at least as “ characteristic ” as those primarily due to a growth of Fixed Capital.
VOL. n
H
CHAPTEE 28
FLUCTUATIONS IN THE BATE OF INVESTMENT — n. WOBEXNO CAPITAL
Subject to the necessary conditions (which will be elucidated in the later sections of this chapter) an increase in the volume of employment will usually require a more or less proportionate increase m the volume of working capital. Thus fluctuations of investment in working capital will be closely correlated with fluctuations in the volume of employment. An increased volume of employment may result either from abnormal activity due, for example, to an invest- ment boom, or to a recovery from a preceding slump. In any case a Credit Cycle will, as we have seen, tend to be associated with an increased investment in working capital — ^if not in its primary phase, then in its secondary phase. Furthermore, it is generally impossible to increase the volume of employment (even when it is at a level far below the optimum) unless it is practicable to increase pari passu the volume of investment in working capital.
Now the practical importance of the fluctuations in the amount of the revolving fund of working capital, which result from these conations, depends on their order of magnitude. If these fluctuations are sub- stantial relatively to the time-rate at which new in- vestment can be made available to replenish working capital either £rom new savings or by diminiahing the volume of investment in liquid capital, then the ques-
102
OH. 28
WORKING CAPITAL
103
tion before us is of great practical significance ; and our analysis may furnish us with an important clue to the explanation of the time-element in booms and depressions. The phenomena of the boom may repre- sent a struggle, concealed under the veil of the credit system, to replenish working capital faster than would be feasible under a regime of stable prices.
If, on the other hand, the possible variations in the demand for working capital are quite small compared with the other elements, so that any deficiency can be rapidly made good out of stocks of liquid capital ^ and out of current savings, then the practical bearing of the above analysis is not important.
Let us proceed, therefore, to an attempt to estimate the various factors quantitatively. How much work- ing capital, for example, is required to run the in- dustrial system of Great Britain at full steam ? How much does it fluctuate between good times and bad ? What relations does this fluctuation bear to stocks of liquid goods, to the flow of current savings and of total investment, and to the variability of investment in fixed capital ?
(i.) The Statistical Indications
In order to reach an approximation to the order of magnitude of the working capital of Great Britain (t.e. the value of goods in process) under normal conditions, we have to consider mainly the value of the output and the average length of time occupied by the productive process. For example, if the produc- tive process lasts six months on the average, and if the product grows in value at a steady rate so that its average value during the six months is half its final value ; then it follows that the working capital required is equal to three months’ output.
It is evident that the amount of working capital
^ Ab Mr. Hawtr^ maintains t vide Trade and CredU^ pp. 126 and 156.
104
A TEEATISE ON MONEY
BK. TI
required per unit value of output varies enormously between different products, corresponding to varia- tions in the length of process — from next-to-nothing in the case of personal services up to the equivalent of a year’s output or more in certain cases.
When we are calculating the working capital required by an individual country, we have also to consider at what stage of the trading and manufactur- ing process that country has to pay for its imports and at what stage it gets paid for its exports; for the former is the date at which, for that coimtry, the demand for working capital begins where imports are concerned, and the latter is the date at which the demand comes to an end where exports are concerned. For Great Britain the date of commencement is generally prior to the date of actual import, and the date of termination is generally after the date of export. The “ process ” of a Chinaman’s shirt begins with the preparation of the cotton fields for sowing and ends when the Chinaman goes shopping. The working capital for the first stage is furnished by America ; from the time when the raw cotton is bought, or soon after, it is furnished by Lancashire ; when the auction has been held at Shanghai, or at some near date to that, it is furnished by China. The com- plete process may average from a year and a half to two years, out of which Lancashire may be responsible for six to nine months, whilst the product will already have acquired a considerable value before Lancashire takes it over. But the distribution between the three parties of the burden of finding the working capital may vary greatly, according as the China merchant pays promptly or slowly, and according to the length of time, mainly governed by the relative rates of interest, during which the cotton-bill is carried in the money-market of New York or in that of London. This example illustrates the difficulty of an exact computation in the case of a trading country. The
OH. 28
WORKING CAPITAL
105
case of agricultural produce grown at home, and con- sumed at home at an even rate through the year, offers, on the other hand, a simpler example ; it has been calculated that the working capital required is equal to about a year’s output in the case of a mixed farm growing cereals, and half a year’s output in the case of a grass and dairy farm.^
Another example for which fairly definite figures can be given is that of copper, statistics being