The Suntory and Toyota International Centres for Economics and Related Disciplines

Short and Long Rates of Interest Author(s): F. Lavington Reviewed work(s): Source: Economica, No. 12 (Nov., 1924), pp. 291-303 Published by: Blackwell Publishing on behalf of The London School of Economics and Political Science and The Suntory and Toyota International Centres for Economics and Related Disciplines Stable URL: http://www.jstor.org/stable/2548108 . Accessed: 17/04/2012 02:53
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Blackwell Publishing, The London School of Economics and Political Science, The Suntory and Toyota International Centres for Economics and Related Disciplines are collaborating with JSTOR to digitize, preserve and extend access to Economica.

http://www.jstor.org

Short and Long Rates of Interest
By F. LAVINGTON
IT is a little surprising that the abundant literature of the London Money Market should show so little sign of curiosity on what might well be expected to be one of its main interests: the relations between the different prices in the market; in other words, the relations between the different rates of interest on capital supplied for short and long periods. Why do the average levels of short and long rates differ ? How far is it true that these two sets of rates rise and fall together ? And, finally, are changes in long rates appropriate in magnitude to corresponding changes in rates for short loan money? To some of these questions, no doubt, market experience can readily give a general answer. This paper is concernedwith the answers given by such figures as are available for the purpose. The development of the London Short Loan Market has been greatly favoured by fortune. Its supply of short loan money could hardly have been so abundant had historical events given our banking system any other than its present form. For the practice of our English banks of offering either cheque-making facilities or a rate of interest in exchange for deposits is peculiarly efficient in attracting the spare funds of the public. At the same time the highly centralized character of the system makes it convenient for bankers to lend a substantial part of these deposits for short periods and by so doing to provide, perhaps, two-thirds of the total supplies available in the London Discount market. On the other hand, innumerable uses for short period supplies of capital arise from our vast merchanting operations. And these uses, expressed in the form of bills, and supplemented by the needs of the Stock Exchange, provide a wide and ample demand for short period funds. This great and even development of the two sides of the London market, in sharp contrast to that of New York, has naturally had its balance disturbed by the finance of war. A great increase in its supplies, together with a decline in the volume of bills to about

one-half of a pre-war amount estimated

at ?350,000,000,

left a

deficit in demand which has been filled, and much more than filled, by Treasury bills. It would be interesting to speculate on the level which short loan rates may reach when recovery in our foreign trade expands the volume of commercial bills. But we are concerned here with rates as they have been and as they are. First:
291

292

ECONOMICA THE RELATION OF THE AVERAGE LEVELS OF SHORT AND LONG RATES.

Is there any reason to expect any difference between the average annual rates for short and long loans ? The circumstancethat short loans, because they are more troublesome to the lender than long loans, should command a higher rate may be set aside, for they are also more troublesome to the borrower. The further circumstance that the rate on any loan is appreciably affected by the market conditions anticipated at the date of repayment is more important. But that influence works sometimes in one direction, sometimes in the other. It does not explain permanent differences in the average level of short and long rates. For that explanation we must turn to other considerations. Every lender, however strong his financial situation, is exposed to the incidence of i-mperfectly foreseen emergencies and opportunities which he can meet effectively only if his capital is readily recoverable from the uses in which he employs it. As a lender, therefore, his ideal situation is one in which his capital is invested in uses from which he can withdraw it at any time convenient to himself. Hence, in order to secure themselves against imperfectly foreseen contingencies, bankers and private persons alike will be willing to supply at least a part of their capital for such uses at rates of interest exceptionally favourable to the borrower. And if it were desired to extend the argument to trade and industry generally, one might fairly argue that the merchant who buys pig iron, which is readily saleable, will be content with a smaller net return on his capital than the iron-master who commits his resources almost irrevocably in the construction of blast furnaces. On the other hand, any limitation of the period for which capital is supplied limits the profitableness of its use to the borrower. The shorter the period of supply, the narrower is the range of uses in which he can safely employ it, and the greater are the risks to which he is exposed from his liability to repay at short notice. From his point of view, therefore, the ideal situation is one in which the borrowed capital is repayable at his option and is therefore available for employment in the widest possible range of uses. Accordingly he will make his strongest bid for capital which is available on these terms and will offer lower and lower rates as the conditions of repayment are less and less favourable to him. The greater willingness of lenders to lend and the less willingness of borrowersto borrow for short periods rather than long, work on rates in the same direction and lead to the expectation that rates of interest will be at their minimum for loans repayable on demand and will rise as the period of supply lengthens. A comparisonof actual rates bears out this expectation. Ordinarily the lowest rate is paid for day-to-day money; higher rates are paid for short loans; still higher rates for loans to stockbrokers; and

SHORT AND LONG RATES

OF INTEREST

293

these differences hold good for post-war as well as pre-war conditions. The following figures are sufficient for present purposes. A complete considerationof the relations between rates and period of supply would make it necessary to examine the yield on a continuous series of investments ranging from day-to-day money, through bills, through short-dated securities maturing at more and more distant dates, up to irredeemable gilt-edged stocks. And an explanation of their relative yields would need to take account of the numerotus post-war anomalies in the prices of stocks and the influence of the conditions anticipated at the various dates of redemption. For the sake of simplicity, therefore, such a comparison is omitted.
AVERAGE London.
Day-to-daymoney,4yrs.11919-22

RATES. New York. Call Loans, 22 yrs. I890-I9I1 60-go Day Bills prime comm. .. ... i2 Railway Bonds 3.67 4-7
4-23

Short Loans, Short Loans, 5yrs. Loans to Stockbrokers, 3 mos. Bills, ShortLoans,4yrs. 3 mos. Bills
,,

,, I9II-I5 ,,

3-44 3 go
2-85

,,
...

4 27 350
3'90 ...

1919-22

4 54

The failure of the discount rate always to conform to the expectations of the argument shows that the argument is not yet complete. Its final stage is more conveniently dealt with under the heading of the second part of the discussion, namely:
THE RELATION BETWEEN CHANGES IN SHORT AND LONG RATES.

It might very reasonably be argued that there are no good grounds for expecting any close connection between short and long rates. The supply of short loan money, it might be urged, is determined by the amount of capital which bankers and others find it convenient to lend on these particular terms; that is, it is largely determined in an arbitrary way by the historical development of our banking system; while the demand is governed mainly (apart from Treasury bills) by the volume of a particular set of merchanting and Stock Exchange operations in which short period money can suitably be employed; and accordingly that there is no obvious reason why the relative pressures of demand and supply for capital available for these particular uses should correspond with their relative pressures in that wider market for long period capital where demand and supply are governed by more general influences. It is not reasonable, for example, to suppose that heavy speculation on the Stock Exchange or an abnormally high price for raw cotton, either of which would increase the demand for short money, would always be accompanied by a corresponding increase in the demand for capital for the expansion of industrial equipment. It would be more' reasonable to expect that the

294

ECONOMICA

movement of the two sets of rates in the two markets would be largely independent of one another. This argument would be a sound one if it were possible to grant its assumption that the supplies of short and long capital were each tied to its partictilar market and were not free to move from one to the other. And this assumption would itself be a sound one, were it not for the development of a device, commonplace in form but cunning in function, which enables any lender to supply capital for long period employments without sacrificing the special advantage of being able to recover his capital at will. Owing to the development of an efficient market in securities, any party who buys, say, railway debentures and so supplies capital for long period uses, can recover the market value of his investment, without disturbing the borrower in his use of the capital, whenever his particular financial situation makes it desirable to do so. So, too, in the bill market. It is no doubt the fact that the Clearingbanks never re-sell their bills; yet the ability of other owners of bills to sell their holdings in the market in order to meet any unforeseen contingency, puts them almost in the ideal situation of lenders-that of having their capital recoverable at will. And this seems quite sufficient to explain the apparent anomaly that the average rate of discount for three-months bills may be below the rate paid on loans to the Stock Exchange, which are made for shorter periods but are yet less readily recoverable. The effect of the invention of the Stock Exchange in facilitating the movement into long period uses of capital which lenders are prepared to supply only for short periods of time, can hardly be less important than the effect of the more deliberate invention of the steam locomotive in facilitating the movement of goods. The influence of highly-organized markets in securities and bills must be very important, too, on the relative levels of rates in the short and long loan markets. In their absence the great bulk of the supplies of loanable capital would be available only for short period employments; the rates paid for capital to be employed in such uses as railways and industrial plant might well be far above their present level; while rates paid for capital for short-period merchanting operations might sink to zero. Such speculations are, however, not very relevant to the immediate practical question: the connection between movements in the average levels of short and long rates. The comparison most conveniently made is that between the average rate of discount for three-months fine bills and the average yield on Consols. In the former case, the rate of return is variable and the capital value fixed; in the latter, the rate of return is fixed and the capital value variable. In both cases the security is about equally good. It would be expected, therefore, that as capital is free to move between the two markets, the average level of the two rates would be closely similar, and that they would

SHORT AND LONG RATES OF INTEREST

295

change in close agreement with one another. The actual figures show that the Rate and Yield, when averaged over periods of years, have correspondedclosely; though in the last two years, for reasons which will be considered later, the two have shown a marked divergence.
I892-1903
I904-I9I3 1914-1923 1920-I923

...
... ...

Average of Daily Ratc on 3 mos. Bills. ... 2 88 ...
... * 3'24 4 IO 4'23

Yield on Consols.

3'23
302 4 43 4'8i

(Oct.), 3 yrs. IO mos.

Since the yield on Consols and the rate on three-months fine bills, when averaged over periods of years, are so closely similar, it is natural to expect a close connection between their annual changes. It is a little surprising, therefore, to find that when the annual average yield and the annual average rate are plotted on a chart, this connection is not exposed: to find that during the twenty-two years ending with i9ii the Yield and the Rate moved together only fourteen times and moved eight times in opposite directions. The close connection which one would expect to find between the two movements is, however, brought out by an alternative method of presentation adopted by Mr. Williams in a paper published in the Statistical Journal of March, i9i2. In that paper Mr. Williams compared the changes in the average annual yield and the average annual rate for the sixty-one years ending with and showed that, with few exceptions, every change in the i9io; average rate was accompanied by a change in the average yield in the direction of the new rate. He showed, that is to say, that the annual changes in the Yield and the Rate were not necessarily in the same direction, but that each change in Rate was accompanied by a change in the Yield which brought the two closer together than they would otherwise have been. That this connection still holds good may be seen by comparing the movements during the

subsequent periodI9II-23:

I9II
I9I2

... ...

... ...

Annual Average Rate. Yield. 2,92 3 15 363 328

Agreesmentor Difference. A

I913 I914 I915 I9I6 I9I7 I9I8 I9I9
I920

...
.

... ...

4-38
285 5 20 4.8I 3'59 3 95 6'39
5'2I

...
.'. ... ...
...

... . ...
'.. ...

370

3-4? 3y34 (7 mos.) 3.82
4'31 4 58 4'40 4,63 5 32 5 21 443 4 29 (IO mos.)

A A A
A A A D A A A A

...

...... ...
...

192I
I922 I923

2 63

...

...

2 68 (IO mos.)

296

ECONOMICA

In support of the evidence given by these figures it may be noted that during the period of twenty years ending with I909, the annual average yields of French Rentes and German Threes moved in close relation to their corresponding market rates. It happens, however, that in both these cases and in the latter period of comparison between Consols and bills the connection between Rate and Yield is brought out almost equally well by the ordinary method of presentation on a chart. These results may do little more than give precision and certainty to what is already known to common sense and experience. It is, however, something to be able to say with confidence that there is an intimate connection between changes in the annual average of the discount rate and changes in the annual average of the yield on Consols, a yield whose movements must be broadly representative of movements in gilt-edged stocks generally. In proceeding to test whether the annual movements in the yields of other types of securities also conform to movements in discount rates, one meets the difficulty that many of these yields, for reasonswhich do not need discussion, do not tend to approximate to the rate on fine bills, but fluctuate about a permanently higher or lower level. It would seem, therefore, that Mr. Williams' method is inapplicable. Nevertheless, if the method is slightly elaborated, it may be applied to show that the movements of some at least of these yields are intimately connected with movements in discount rates. Consider the figures for the average annual yield of twelve Americanrailway bonds and the average annual rate on four to sixmonths commercialbills in New York. For the twenty years ending with I909 the average rate was 5 80 per cent. and the average yield was 4-23 per cent., showing an average difference throughout the period of I-57 per cent. If, therefore, there is a connection between the annual movements of these bond yields and the annual movements of these bill rates similar to the connection between the yield on Consols and the London rate of discount on three-months bills, it is to be expected that the bond yield will constantly tend to approximate to a figure I-57 per cent. below the New York bill rate. In order to show whether or no this connection exists, the following table is given. In the first column are set down the annual average rates; in the second column these Rates minus 1-57 per cent.; in the third column the average annual yields. The fourth column gives the result of a comparison between the movements shown in columns two and three. A indicates that a change in Rate is accompanied by a change in Yield towards the new Rate. D indicates divergence.

SHORT AND LONG RATES OF INTEREST RATE. Rate 157 % Rate.
I 890 I89I I892 I893 I894 I895 ... ...
'..
,

297

Yield.
4 72

... ... ...
..

6&89 6'50 5-38 7,62
5-22

5-32

4-93 3.8I 6-o5
3-65

485
4.64

A
A A A A A A A A D A D A A A A A A A

4-75
4-59

.
. ..
,

i896 i897 I898

573 7-02 4'72
53I

I899 I900
I9O1
1902

...
. ...
,

5-48 5.7I
5.4I 5-75 6z2I

4. I6 5-45 3.I5 3 74 3.9t 41I4
3y84 41i8

4.48

4*54
4'38
4.21

3.96

395
379 377 3.96
3-92

...

I903 I904
I905

...
...
*

5.I3
517

4.64 3>56
3.60

3-82

I906 I907 i9o8
I909

...
.

...
. ...

6-24 6-55 495 4,67 5- 8o

4-67 4.98 3.38
310 4- 23

394
4-22

4 I6 4o00
4.23

Average

...

The table shows nineteen changes in the corrected rate. Seventeen of these changes were accompanied by changes in the yield, which brought the new yield closer to the new rate. Two of the changes were accompanied by very slight divergences (oi and o2,) in the yield. Plotted on a graph, these movements show no appreciable correlation. The use of Mr. Williams' method, however, makes it clear that there is here the same intimate connection between the long period yield and the short period rate as was seen to exist between the return on Consols and London bills. The yield on these railway bonds tends constantly to conform to a level I '57 per cent. below the discount rate for four to six-months bills. It is a fair presumption that there is a tendency for all investment yields to conform in this way to discount rates. But it must not be expected that the tendency can always be isolated and exposed; it will, of course, frequently be concealed by changes in yield due to changes in the public estimate of the worth of the particular security under examination. The selection for comparison of a yield which was an average of the yields of twelve sound bonds had the advantage of eliminating the effects of disturbances due to changes in the particular financial circumstances of individual securities. When the comparison is made with the yield of a single security, the conformity of its movements with movements of the discount rate is naturally less close by reason of these arbitrary changes in the yield of the security arising from changes in its credit. But the connection may still be clear. A particularly safe security is the West Shore Railway bond.

298

ECONOMICA

A comparison of its yield with the rate on 60-go day bills in New York for a period of thirteen years ending with i9ii gives the result: A==9, D=3. A similar comparison of the yield on L. and N.W. Deb. Stock with the rate on three-months bills in London for the period of seventeen years ending with I9IO, gives the result: D 4, No change=i. But, again, these two results are A-ii, brought out almost equally clearly by the ordinary graphical method. As the comparison passes from the average yield of a group of bonds, through the yield of individual sound securities, to the yield of stocks which are more and more subject to the disturbing influences of changes in their particular financial circumstances, the conformity of the movements of Yield and Rate is likely to become less and less apparent. But it seems to be a fair conclusion from these figures that the annual average yield of every sound fixed-interest stock is likely to show a tendency, more or less marked as its credit is more or less stable, to conform to a level x per cent. above (or below) the annual average rate in the bill market.
THE RELATIVE MAGNITUDE OF THE CHANGES IN SHORT AND LONG

RATES. There remains the more interesting, more important and much more difficult question: Are the changes in long rates appropriate in amountto correspondingchanges in short rates: more specifically, are their changes in periods of rising and declining business activity such as to make it always about equally profitable to invest in bills and in securities ? In order to investigate this, let us neglect any difference between the cost of investment in bills and in securities; let us assume that a considerable volume of capital is free to move between the two markets; and let us compare the rates for three-months fine bills with the yield on Consols for two periods of five years, having as their centre points the " peak " years I907 and ig20 respectively:
Rate.
1905 I906 I907 I908 1909
... ... ... 2'62

Yield. 2 78
2.83

...
....

...
... ... ...

...
... ... ...

3.97 4 49
2-29 2-28

297
2'91

... ...

2-98
2-89

Aver. of 5 years

3I3

The I907 Boom.-If the Rate and the Yield had clearly defined normal levels to which they returned when the effects of boom (or depression) had exhausted themselves, the comparison would be a simple and conclusive one. The two normal levels could be used as bases of measurement. The extent to which the Rate rose above its base in boom years would serve to measure the additional

SHORT AND LONG RATES

OF INTEREST

299

profitablenessof investment in bills; the extent to which the Yield rose above its base would similarly serve to measure the increased profitableness of investment in Consols. And a comparison of the results would show whether or no the rise in the one was appropriate in amount to the rise in the other. In fact, however, as there are no such clearly defined normal levels of Rate and Yield, it is necessary to substitute for them the average levels during the five years over which the comparison extends. These two averages were, it is seen, 3I3 per cent. and X289 per cent. In the year 19O7 the average yield on Consols rose to 2-97 per cent., or o8 per cent. above its five year average. Accordingly, anyone who invested &r,ooo in Consols at their average price in that year would have secured an additional yield of ?8 per annum. But this additional yield is, of course, a perpetual one; its value to the investor is the present value (at 2 89 per cent.) of that perpetual additional yield: i.e., ?28. Hence, in order that investment in discounts should have been equally profitable, ?i,ooo employed in the bill market should have given an additional return of about ?28 before the rate again fell to its average level of 3-I3 per cent. But in fact, as the table shows, ?i,ooo so employed would have earned an additional return of "84 per cent. (3.97-3.I3) in 1906, andof I-36 percent. (4 49-3 I3) in I907, givinga total of about?22. During this period of exceptional business activity, therefore, the rise in the yield on Consols was not appropriate in amount to the rise in the return on bills. It would appear that, for reasons which are discussed later, the market pressed the capital value of Consols somewhat too low and so raised the yield somewhat too high. The result is strengthened if account is taken of the fact that in the month of November, I907, the yield on Consolsrose to 3 04 per cent. The I920 Boom.-As might be expected, post-war statistics show the same effect in a more marked degree. The corresponding figures are:
I918 9I9i
1920

... ...
...

... *
.. '

,^ *e
.. ...

Rate. 3 59
3 95 6-39
5.2I

Yield.
4'40

4 63
5 32
5.21 4 43

I92I1 1I922

..

..

2-63

Aver. of 5 years

...

4 35

4 80

These figures show that a person who invested ?&,ooo in Consols at the average price prevailing in 1920 would have secured a perpetual yield of ?5 2s. per annum above the average yield for the five years. The present value of this additional yield at 4 8 per cent. is ?ioo. A similar sum employed in the discount market would have earned an additional return of not more than ?29 before the rate again fell to its five year average level.

300

ECONOMICA

It may be added that the tendency in years of extreme business activity for the yield on Consols to rise disproportionately to the rate on bills is found also in the five year periods whose centres are

the boomyears of I900 and I889-90.
The interpretation of these results needs caution. It is not claimed that they show the precise amount by which the profitableness of investment in Consols in boom years exceeds that of investment in, bills. For the purpose of that calculation it would be necessary to take into account the fact that the boom years of I889-90 were situated in a period in which the capital value of Consols was per-

sistently rising, while the boom years

I900

and

I907

were

situated in periods in which that capital value was persistently falling; and allowance for this fact would modify the arithmetical estimate of the superior profitableness of investment in Consols during periods of boom. It is claimed for the results that they isolate the effect of increased business activity on the relative levels of the Yield and Rate and show that in each of the four years of boom, investment in Consols was in some degree more profitable than investment in bills. What, then, is the explanation of the failure of the market to maintain the Yield and the Rato at levels of similar profitableness ? It is a matter of common sense and common knowledge that in times of high activity, when there are exceptional opportunities for the profitable employment of capital in industry, business men, in order to take advantage of these opportunities, take their money " out of store "; in other words, they sell their investments and by so doing depress the capital value of Consols and other giltedged securities. This action in itself need cause no discrepancy between long and short rates, for the increased profitableness of business which indirectly raises the yield on Consols is certain to increase the volume of commercial bills and so to raise also the discount rate. But it seems likely that the high expectation of profit in business, which indirectly depresses the capital value of gilt-edged securities, will make holders of those securities apprehensive of a further fall and so cause yet further selling. And that this further selling will continue until the prospective profitableness of investment in Consols exceeds that of investment in bills by an amount which compensates the investor for the possibility of a further fall in their capital value. If this is the right explanation of the discrepancy in boom years between the relative profitableness of investment in Consols and bills, one would expect a symmetrical opposite effect in times of depression. One would expect that in times of low business activity, a gloomy view of the prospective profitableness of capital employed in trade and industry would cause business funds to be pressed unduly far in the purchase of gilt-edged securities; that the yield on such securities would be forced, not necessarily below the level of bill rates, but below the level corresponding to the

SHORT AND LONG RATES OF INTEREST

30I

reduced profitableness of investment in bills, and hence that investment in gilt-edged securities would become less profitable than bills, the difference being a measure of the expectation of a further increase in their capital value. It may be that this is true. But if figures are taken which isolate the influence on the relative levels of Yield and Rate arising from conditions of trade depression, the answer which they give is uncertain. There seems to be no such firm basis of facts for concluding that the yield on Consols is forced unduly low in times of depression as there is for concluding that the yield is forced unduly high by periods of boom. Indeed, if the figures for the present depression are examined, they appear at first sight to show a precisely opposite condition of affairs.
THE PRESENT YIELD ON CONSOLS AND BILLS.

Considerthe following table showing the relative Yield and Rate since I9I7:
Yield.
I9I7
...

ANNUAL Rate.
48I

AVERAGES. Yield.
192I
..

Rate.
5 2I

4 58

21

i9i8
I9I9
1920

... ...
...

4'40 4 63

3-59

5 32

3 95 6.39

1922 *'' I923 (IQ mos.)

4 43
4-29

2

63

2-68

The figures are remarkable-the more so when it is remembered that the yield on Consols is very low in comparison with the yield on other long-dated gilt-edged stocks. They suggest that the long persisting tendency for the Yield and Rate to conform to the same level has suddenly weakened in a notable degree. Possibly since I9I8, certainly since I922, the yield on Consols has been maintained at a level far above the corresponding rate on bills. During I922 and the first ten months of I923 investment in Consols has yielded a return of more than ij per cent. per annum above that obtainable in the discount market. And this marked and persistent change in the relative profitableness of the two forms of investment has occurred in spite of the circumstance that demand in the short loan market has been greatly strengthened by immense quantities of Treasury bills. What explanation can be given of this change ? It is known that after the boom of I920 a good deal of capital usually employed in commerce, failing to find profitable employment in its ordinary uses, was diverted into the market for loans. If, as seems also to be the fact, the owners of this capital believed in a rapid trade recovery, they would have anticipated not a further rise, but rather an early fall in the capital value of Consols and would consequently have preferred to employ their funds in uses such as bills where the capital would be secure, although the rate of return might be unduly low. This distaste for long-dated stocks would naturally be reflected in the relative levels of rates.

302

ECONOMICA

The effect of such a temporary diversion of funds may be a partial explanation of the recent change in the relative profitableness of investment in Consols and bills, but it seems probable that the full explanation depends on a permanent change in the relative eligibility to bankers and others of the two forms of investment. Possible causes of such a change are not difficult to find. The continuous decline in the capital value of Consols and other giltedged securities from I896 to I9I4 and the violent fluctuations in their capital values since that date can hardly have failed to shake the confidence of financiers in their eligibility as a secure shortperiod investment. The effect of this increased uncertainty in their capital values has been reinforced by the recent creation of a large supply of short-dated securities, equally marketable, relatively more stable in capital value and hence more suitable than Consols for the investment of funds temporarily unemployed. This twofold change in market conditions, the more unstable capital value of Consols and the new competition of securities whose capital value is steadied by the promise of early redemption, seems sufficient to explain the relatively very high current yield on Consols as compared with bills. It does not suggest that the intimate in connection between movements the Rate and Yield which existed for at least sixty years up to the end of the war is now broken. But it does imply that the tendency to equality in Rate and Yield is now markedly and perhaps permanently impaired.
CONCLUSION.

In order to bring the relations between short and long rates into a rough unity it may be convenient to summarize the various facts, inferences and suggestions which have been detailed. The argument suggests that the relative levels and the relative movements of short and long rates would show no clear connection, but would be governed mainly by arbitrary and incalculable influences, were it not for the development of such organizations as the Stock Exchange and the Bill Market, whose effect is to impart to investments in the " long " market most of the peculiar advantages attaching to the loan of capital in employments from which it is readily recoverable. That it is only in consequence of the presence of those organizations that there is sufficient mobility of funds between the long and short markets to bring the levels and the movements of their rates into calculable conformity. It shows that for many decades up to the end of the war the yield on Consols and the three-months bill rate, when averaged over periods of a few years, were substantially identical. This strong tendency towards.conformity was impaired in years of great business activity by influences which forced the yield above its appropriate level. In recent years it has given way as a result of special circumstances due to war. It shows, further, that changes in the Yield and the Rate were

SHORT AND LONG RATES

OF INTEREST

303

intimately connected; that up to I923 it almost invariably occurred that each change in the annual average rate was accompanied by a change in the yield, whose effect was to bring the two into closer agreement with each other. But this connection between movements in bill rates and movements in the yield of securities in general was often not susceptible of exposure by ordinary graphical methods. An adaptation of a method employed by Mr. Williams was necessary for the purpose. It shows, again, that a similar close connection exists between movements in the average annual level of, e.g., the yield on a group of American railway bonds and of a New York discount rate; and accordingly bears out the natural expectation that the connection holds good for all sound securities, though it may be partially concealed in the case of individual stocks by disturbances of yield arising from changes in their particular circumstances. The argument has dealt only with the relative levels and relative movements over periods of a few years of two sets of rates roughly distinguished as short and long. It has said nothing of the causes promoting these movements. But little doubt remains of the nature of the principal cause when one further connection is made: when movements in prices are attached to movements in discount rates and through them to movements, each in its own degree, of the whole set of associated rates and yields. It is well known and easily proved that during the past sixty years upward and downward movements in the annual average level of general prices have been intimately associated with corresponding upward and downward changes in the discount rate. It is equally well known that rising prices increase the profits of business, expand the demand for capital and so raise short and long rates alike; similarly, that falling prices, by reducing the profits of business, set in motion an opposite train of effects. So far, therefore, as we are concerned with those fluctuations in rates and yields which are completed within periods of a few years, the moving cause is evidently to be found, not in changes in the supply of capital, but principally in those changes in the profitableness of business which govern the volume of its demand. The figures used have been obtained from the following sources: EconomicService, The StatisticalAbstract,The Londonand Cambridge The Economist,documents of the U.S.A. Nat. Monetary Commission, the Bank of England, Mitchell's Business Cycles, articles by Mr. Williams and Mr. Macdonald in The Statistical Jou:rnalfor March, 1912, and by Mr. Spring Rice in The Bankers' Magazine for March,
1923.