The Stock Market, Credit and Capital Formation

CHAPTER VII

THE DEMAND FOR LOANS BY THE STOCK MARKET

49. There are a considerable number of authors who regard the volume of loans to stockbrokers as a measure of the funds which have flowed onto the stock exchange or even as a measure of the funds that have been absorbed by the stock exchange. In actual fact it is neither of these things.

The total volume of lending to brokers within any period may be substantially smaller than the amount of money capital which has flowed onto the stock exchange, or it may be greater. There is no definite relation between these items, nor even any necessity for them to move in the same direction. Why this is so will be explained in the course of this chapter.

The volume of brokers’ loans tells us nothing about the amount of funds that has flowed onto the stock exchange,—

The analysis given in the previous chapters should have made it sufficiently clear that the volume of loans to brokers has nothing to do with any tie-up of purchasing power in stock exchange transactions. The subsequent sections will complete that exposition.

—and still less about a tie-up of funds.

First of all it is necessary to explain the essential facts relating to the mechanism of brokers’ loans.1 When does a broker borrow? He borrows, like any other person, when he expects his receipts from ordinary sources to be less than his outgoings. His receipts, apart from new borrowing, consist mainly in the proceeds of sales to other brokers and in receipts from his customers (or for their account). His outgoings, apart from loan repayments, consist mainly in payments to other brokers for purchases from them and payments to customers.

Brokers borrow when they have to pay out more than they receive.

The cash ledger of any individual broker will thus show on the receipts side, receipts from brokers and receipts from (or on account of) customers, and on the expenditure side, payments to brokers and payments to customers. If, however, we were to take all the brokers as a group, then the payments between brokers would of course cancel out, and only the payments from and to customers would remain. The reason why we select for inspection all brokers together instead of a single broker is that our object is to explain the total of brokers’ loans and not loans to individual brokers. The individual broker will of course need to borrow when he has to pay a clearing difference to another broker, but the latter will repay a loan or grant a loan at the same time and no later.2 Consequently, the payments of clearing balances between brokers do, it is true, lead to shifts in the person of the borrower, but they do not give rise to changes in the total volume of borrowing. Changes in the total volume of borrowing are caused exclusively by differences between payments by customers and payments to customers. If the payments to customers (mostly in respect of the proceeds of sales) are the larger, then the brokers need to take up new loans; if the payments by customers (mostly in respect of purchases and also dividends received on their behalf) are the larger, then the brokers are able to pay back old loans.

Payments between brokers cancel out for all brokers taken together.

Payments from and to customers determine the brokers’ borrowings.

The process may be made clearer by the aid of examples depicting schematically the course of events.3 (Ledger balances are shown in Appendix A.)

Monday: Mr. A pays his broker the sum of $20,000 and informs him that he will give him an order to buy in due course. The broker credits A with the amount and in the meantime applies it to the purpose of reducing his bank debts.4

First illustration,—

The total of brokers’ loans declines on this day by $20,000.

Tuesday: Mr. A gives his broker an order to buy. The broker buys the shares ordered by his customer for $19,500 from another broker who is selling the shares on behalf of his customer Mr. B.

Settlement does not take place until the next day.5 To-day there is no change in the positions.

Wednesday: A’s broker borrows $19,500 from his bank6 in order to pay to Mr. B’s broker. The latter credits his customer with $19,500, but as B has not demanded payment he (B’s broker) uses the $19,500 to pay off part of his own debt to the bank.

—which shows that stock purchases and sales do not affect brokers’ loans—

On this day the total of brokers’ loans has not changed.

Thursday: Mr. B gives his broker an order to buy $12,000 worth of stocks and asks for payment of the rest of what is due to him. The broker borrows $7500 in order to pay this amount to Mr. B. He buys the required shares from another broker who is selling for the account of a Mr. C.

On this day brokers’ loans have risen by $7500.

Friday: B’s broker borrows $12,000 in order to pay C’s broker.7 The latter credits his customer with the $12,000 and reduces his own debts.

On this day the aggregate of brokers’ loans has not changed.

Review of the week: The total amount of brokers’ loans outstanding has fallen by $12,500. This is explained by A’s paying in $20,000 and B’s withdrawing $7500. Mr. A has acquired a brokerage deposit of $500 and Mr. C a brokerage deposit of $12,000: these new brokerage deposits of together $12,500 correspond to the decline in brokers’ loans. The decline in brokers’ loans corresponds in turn, if they are loans from the banks, to a decline in bank deposits which involves an increase in the “excess reserves” of the banks. In so far as the flow on to the stock exchange has not flowed off the stock exchange, it has caused a paying back of bank credits and thus made it possible for new bank credits to be granted to the same amount. The new flow of money capital to the stock exchange is counterbalanced in the case described by a reduction in bank lending to the stock exchange.

—that an excess of customers’ deposits over customers’ withdrawals causes a decline in brokers’ loans;—

—and that a new flow of saved funds to the stock exchange causes a fall in bank loans and deposits.

50. The fact that a large number of speculators buy more stocks than they can pay for out of their own resources, i.e., that they borrow “margin loans,” would not in itself necessitate any growth in brokers’ loans. For if the people whose stocks are sold to the “margin speculators” do not withdraw the sales proceeds from the stock exchange, that is to say, if they do not take them away from their brokers, the brokers have nothing to pay out and do not need to borrow anything. The buyer of the stocks will have run up debts with his broker, but the brokers do not need to borrow new money from anybody so long as the seller does not demand payment of the money due to him. The buyer will have bought without paying in the amount due, and the seller will have sold without being paid the amount due. If the buyer and the seller both keep their accounts with the same broker, then there will not even be any alteration in the borrowing positions of the individual brokers. If the buyer and the seller keep their accounts with different brokers, then the broker of the buyer will have to take a loan and the broker of the seller will be able to pay back a loan: or in the case that this latter broker has no debts he will himself lend to the broker of the buyer.8

When customers buy “on margin” and thus borrow from the brokers, the brokers need not borrow unless the sellers demand payment.

Thus, the total amount of brokers’ loans is directly dependent neither on the stock exchange turnover, nor on the level of security prices, nor on the new margin debts incurred by speculative buyers. It is dependent only on the difference between payments in by customers who have bought shares (plus dividends received for customers) and withdrawals by customers who have sold shares (or are collecting dividends). What these sellers, whether they be owners of old shares or issuers of new shares, do with their money is of course not apparent from the statistics of brokers’ loans. All that may probably be concluded from the statistics9 when they show a rise in brokers’ loans is that larger sums have been withdrawn from the stock exchange, that is from the brokers, than have been paid in to the brokers.

Neither security turnover, nor security prices, nor speculators’ borrowings directly determine brokers’ loans.

Let us again illustrate the relationships by taking another week’s transactions. (Ledger balances are shown in Appendix A.)

Second illustration,—

Monday: Mr. A, who has deposited securities to the value of $20,000 with his broker, is optimistic and desires to buy more securities to the value of $10,000, i.e., he takes up a “margin loan.” The broker buys the shares from another broker who is selling them for the account of Mr. B.

Settlement does not take place until the following day.

Tuesday: A’s broker borrows $10,000 and pays this sum to B’s broker. The latter credits B with the amount and reduces his own debts.

Thus brokers’ loans in the aggregate have not risen. The margin debts of customers to their brokers have risen (since Mr. A now has a debit of $10,000 against his account) and the brokerage deposits of customers have risen (since Mr. B now has a credit of $10,000 to his account). The margin debts of the brokers, i.e., the sum of brokers’ loans, have however remained unchanged.

Wednesday: Mr. B orders his broker to pay him $5000 (of the $10,000 due to him) and to buy certain shares for $8000. He thus incurs a margin debt to the extent of $3000. The broker borrows $5000 to pay out to B. He buys the $8000 worth of shares from a broker who is selling them on behalf of C.

On this day brokers’ loans have increased by $5000. Settlement of the stock purchase takes place to-morrow.

Thursday: B’s broker borrows $8000 and pays this amount to C’s broker. The latter credits C with the $8000 and uses them to reduce his own debts.

On this day the total of brokers’ loans has not changed. It will not change until C, or somebody from whom he buys other shares, withdraws money from the broker.

Review of the week: The sum total of brokers’ loans has risen by $5000. This is the amount withdrawn by B. The margin debts of customers have risen by $13,000 (A borrowed $10,000 and B $3000) and the brokerage deposits of customers have risen by $8000 (which were credited to the account of C). The difference between the growth in customers’ debts to brokers and the growth in customers’ deposits with brokers (which are identical with brokers’ debts to their customers) is balanced by the growth in brokers’ debts to the banks. The increase in lending by the banks amounts to $5000; the new brokers’ loan led to the creation of a bank deposit which was placed to the account of Mr. B on Wednesday. What he does with it we do not know. He may use it to increase the stocks of materials or the equipment of his firm, he may buy his wife a fur coat with it, he may lead the money and earn interest on it (see § 52), or he may, in certain circumstances, leave it idle (see Chapter VIII).

—which shows that the excess of new customers’ borrowings over new brokerage deposits held by customers equals the rise in brokers’ loans;—

—and that the now brokers’ loans create bank deposits to account of those who withdraw funds from the stock market.

The expansion of bank lending was here the source of the flow of money capital to the stock exchange, but it had no sooner flowed onto the stock exchange than it flowed out again, since the brokers’ loan was only borrowed for the specific purpose of making payments to customers.

These interconnexions have been described with remarkable clarity by Eiteman. He comments on the enormous figures of brokers’ loans in New York in 1929 in the following terms: “Since increases in the total of brokers’ loans represent an excess of customers’ withdrawals over deposits, it follows that the huge brokers’ loan total of 1929 indicated the amount of funds withdrawn from speculation rather than the amount diverted into speculative channels for purposes of aiding stock gamblers to trade on margin. Whether these loans also deprived legitimate business of needed funds depends upon the uses to which the funds were put by those who made the withdrawals.”10 If the sellers who withdrew these funds had themselves used them to purchase other securities a couple of days later (as many other authors thought was likely), brokers’ loans would have declined again, or in the case that the next sellers had immediately withdrawn the proceeds, would at least not have increased any further. “But some group must have sold stocks without repurchasing, for the total of brokers’ loans did increase. During 1928 and the first nine months of 1929, corporations whose stocks were listed on the New York Stock Exchange are known to have printed and sold shares of new issues for which they received $3,042,120,000 in cash.”1 In short, a substantial part of brokers’ loans were taken up in order to pay out new capital to corporations.

The vast sum of New York brokers’ loans in 1929 could be interpreted as indicating hug withdrawals of funds from the stock market,—

—especially by corporations which sold new issues of shares.

51. Many authors were not prepared to accept this interpretation of the heavy increase in brokers’ loans.2 The statistical correlation between brokers’ loans and stock prices was too striking. “The great increase in brokers’ loans was a function of stock price increases.”3 And in saying this Professor Beckhart was undoubtedly expressing the opinion of many of his colleagues. It is interesting to note that in this statement (true to the tradition of the Banking School) it is not the stock prices which are treated as a function of the volume of credit, but the volume of credit which is treated as a function of stock prices. If what is meant by this is that in consequence of the higher prices the value of the turnover rises and the brokers require larger cash holdings to deal with this turnover, it may be said at once that it is simply not true (see § 40 above). Yet Professor Ellis also believes that “Local brokerage houses can no more expect to carry through a larger volume of business with the same credit balances than can a local grocer.”4 It seems to me that this misses the essential distinction. The local grocer cannot help having his till fuller at the end of a busy day than on a day when business has been slack. But the broker who has heavier receipts from customers, and in addition expects an active balance in the stock exchange clearing, will use his receipts even before the end of the day’s business (he may use part of them even before the stock exchange clearing) either to repay his debts or to lend out at call. There is no reason why he should keep larger bank deposits in consequence of the higher turnover or merely as a result of higher stock prices. But even where rules or conventions or convenience induce the brokers to keep larger bank balances when their turnover or their debts increase, the effect is of a ridiculously small order of magnitude. In proportion to the turnover and to the volume of brokers’ loans in times of boom, or indeed in proportion to the total increase in circulation, the bank deposits of brokers are almost microscopically small.

Some writers believed the increase in brokers’ loans was caused by stock price increases,—

—especially because brokers would need larger balances for handling a larger turnover

This view is faulty:—

—there is no inherent necessity for larger balances,—

—and their relative size, even if increased, is trivial.

Thus, in order to give the argument a generous interpretation and to make sense of the statement that brokers’ loans are a function of stock prices, the level of stock prices must somehow be linked up with withdrawals of customers’ funds. The link is not difficult to discover. High stock prices lead (1) to withdrawals of their gains by those who want to consume their additional “income,” (2) to withdrawals of the whole of the sales proceeds by those who want to “get out” of the stock market, and (3) to the flotation of new shares and the withdrawal of the sales proceeds by the issuing corporations.5 All these withdrawals are, so far as is necessary, financed by new brokers’ loans. The rise in stock prices may thus be said to explain the volume of brokers’ loans just in so far as it explains withdrawals of sales proceeds; to treat it as being in some way antithetical to these withdrawals is a grave misunderstanding.

High stock prices, however, encourage withdrawals by profit-takers, and by corporations issuing new stocks.

Brokers’ loans finance these withdrawals.

Earlier we drew a contrast between the thesis that the volume of credit is a function of stock prices, and the thesis that stock prices are a function of the volume of credit. This antithesis is found very frequently, but unfortunately no care is taken to make clear the not unimportant fact that “the volume of credit” means in one case the “demand for credit” and in the other case the “supply of credit.” An ample supply of loanable funds will lead to a lower interest rate, and, under certain circumstances, to increased business activity and higher stock prices. The higher stock prices lead to withdrawals of funds from the stock market by the sellers of stocks, and thence to a demand for credit by the broker. In what follows we shall try to show that the money which flows out of the stock exchange may sometimes reappear as part of the credit supply and make further rises in stock prices possible.

Supply of easy credit may lead to demand for brokers’ loans.

52. The high stock prices offer corporations a rare opportunity to cover their past, current, and future capital needs on the most favourable terms. Capital may have been raised in the past through unfunded debts or through the issue of fixed interest-bearing bonds. The high stock prices provide the corporation with the incentive to alter its financial capital structure by paying back the debts or the bonds, and so reducing the interest charge and raising its profits. The high stock prices also encourage corporations to raise capital for all kinds of new investment, including investment which is undertaken only because the conditions for obtaining capital are so favourable. Finally, they encourage the raising of capital for which there are as yet no specific investment plans: corporations do not want to let so favourable an opportunity for obtaining capital pass even if they have not drawn up their investment plans. (In the United States, the regulations of the Securities Exchange Commission have made it impossible to raise capital for as yet undetermined purposes. Stock issues of this kind were not infrequent during the boom of 1928–29.)

Soaring stock markets induce stock issues—

—for refunding,—

—for new investment,—

—and sometimes for undefined purposes.

In all three cases (refunding, new investment, and indefinite plans) the proceeds of the stock issues are in the first instance withdrawn from the stock exchange, so that if the purchasers have procured funds by taking up margin loans there will be an increase in brokers’ loans. In the first and third cases, i.e., where the newly raised money capital is not immediately used for the purposes of real investment, the money withdrawn from the stock exchange may return there in one of two ways. The corporations (or the creditors who are repaid or the previous owners of bonds that are now redeemed) may themselves use their increased cash balances to purchase stocks; this would result in payments to brokers and a consequent decline in the volume of brokers’ loans outstanding. Or the corporations (and other recipients of the funds) may use their increased cash balances to grant loans to brokers. This is very attractive if the rate of interest on call money is high. The result is that the same funds (as originated in an initial broker’s loan) may serve to finance withdrawals by other people and capital issues by other corporations—and the statistics will register a further rise in brokers’ loans.

The funds withdrawn from the stock market may return again—

—for other stock purchases,—

—or for other brokers’ loans.

Some part of the funds withdrawn by a corporation thus returns to the stock exchange. Strictly speaking, of course, they do not go back “to the stock exchange” for they are only used there in order to be paid out to other persons and corporations. Thus, the cash balances of one corporation may be transformed into cash balances of another corporation, but not without causing the statistical returns to show a rise in the aggregate of brokers’ loans.

The withdrawals by one seller may be used, through new loans to brokers, to finance the withdrawals by another seller.

The analysis becomes increasingly complicated as we proceed, and it may be helpful to give a new illustration of the transactions of another “week.” (Ledger balances are shown in Appendix A.)

Third illustration,—

Monday: Mr. A, who has a large deposit of fully paid-up stocks, orders his broker to buy $35,000 worth of stocks. The broker obtains them from another broker who carries out the sale on behalf of Mr. B. At the same time as he ordered his broker to sell these stocks Mr. B also directed him to purchase other stocks to the value of $53,000 of which $32,000 worth are a new issue of corporation M. The broker buys the remaining $21,000 worth from another broker who is selling for the account of Mr. C.

Settlement takes place on the following day.

Tuesday: Mr. A receives certain funds that he had been expecting and pays in $10,000 to his broker. He thus remains in debt to the extent of $25,000. A’s broker borrows $25,000 from his bank and pays $35,000 to B’s broker. B’s broker pays $32,000 to corporation M for yesterday’s sale6 and $21,000 to C’s broker. B’s broker must therefore borrow $18,000 from his bank for his customer’s purchase. C’s broker, who, so far, has received no further orders from his customer, credits him with the $21,000 sales proceeds and reduces his own bank debt by the same amount.

On this day the margin debts of customers have risen by $43,000 and brokers’ loans by $22,000 ($25,000 plus $18,000 minus $21,000).

—which shows that buying, on margin, of new issues increases brokers’ loans because the corporations withdraw the funds;—

Wednesday: Mr. C orders his broker to buy him $20,000 worth of a new issue of corporation N. The broker makes the purchase. Settlement takes place to-morrow.

Thursday: Corporation N withdraws the $20,000 deriving from yesterday’s sale of stock. The broker C takes up a loan offered by corporation M7 to the amount of $32,000 and (after paying $20,000 of it to corporation N) pays $12,000 back to his bank. Mr. B asks his broker to pay him out the sum of $6000: his broker finds Mr. B has sufficient margin so he borrows the $6000 from his bank and remits them to B.

—that these funds may be used for other brokers’ loans, which may serve to liquidate bank loans to brokers;—

On this day the margin debts of customers have risen by $6000. Loans granted by the banks for their own account have diminished by $6000 ($12,000 minus $6000): brokers’ loans on account of others have risen by $32,000. The aggregate of brokers’ loans has thus increased by $26,000.

—that they may also finance new margin buying of other new issues and thus withdrawals by other corporations.

Review of the week: The aggregate of brokers’ loans has risen by $48,000 of which $16,000 are on account of the banks and $32,000 on account of others. The margin debts of customers to brokers have risen by $49,000 (Mr. A $25,000; Mr. B first $18,000 and then another $6000). Brokerage deposits of customers have risen by $1000 (Mr. C sold $21,000 worth of stocks and bought $20,000 worth).

The rise of $48,000 in the total of brokers’ loans is explained by the fact that $58,000 have been withdrawn from brokers and only $10,000 have been paid in to brokers. This payment came from Mr. A, and the withdrawals were those of the two corporations totalling $52,000 (M $32,000 and N $20,000) and the $6000 taken away by Mr. B.8

What does this figure of $48,000 signify? Was the amount of funds which flowed onto the stock exchange $48,000? Or was it $58,000 so as to count the remittances by customers as well as the loans? Neither the one nor the other can be seriously argued. It would be quite unreasonable to calculate that on the Tuesday in our example $32,000 ($10,000 paid in by customers and $22,000 derived from bank loans) flowed onto the stock exchange and that on the Thursday a further $26,000 (in the form of loans) followed. These “additional” funds were in fact still the same funds, parts of which were transferred from the account of corporation M to the account of corporation N and to the account of Mr. B. If corporation N also offered credits which were used to finance company O, and this process continued, one and the same dollar would wander on and on from one account to another and cause the total of brokers’ loans to rise with each successive transfer.

How much money has gone to the stock exchange?—

—by no means the total increase in brokers’ loans, since it may count the same amount several times.

53. What is the amount which can really be considered to have “flowed on” to the stock exchange in our example, and what has become of it? Let us assume that the $10,000 paid in by Mr. A had been saved by him out of his current income. These $10,000 must undoubtedly be treated as having flowed on to the stock exchange. In addition the banks have granted credits to the extent of $16,000 net. This sum which may be assumed to be the result of credit expansion by the banks has also flowed on to the stock exchange, making in all $26,000. And where were these $26,000 at the end of our week? $20,000 has passed into the account of corporation N to await further allocation by the treasurer of this corporation, and $6000 were in Mr. B’s account on Thursday, but by Friday they had probably already been transferred to the account of somebody else, since B will not have withdrawn his money just for fun.9

The deposit by a saver—

—and loans extended by banks for own account represent the real afflux,—

—which went to those who withdrew funds.

The $26,000 thus registered stock exchange credits of no less than $48,000. If, on the following day, corporation N lends its $20,000 to brokers, who use it to finance the purchase for Mr. D of $20,000 worth of the stock newly issued by corporation O, then the figure for stock exchange credits will already have risen to $68,000, with every prospect of gaily rising further. The volume of funds which “have flowed onto the stock exchange” will still, however, be no more than the $26,000 subscribed out of the savings of Mr. A and the credit expansion of the banks.

But does not this clearly prove, it will be asked, that funds were absorbed in stock exchange transactions? Have not $20,000 been shifted from one account to another in a series of unproductive transfers? If it is granted that funds, which incidentally owe their existence chiefly to new credit expansion by the banks1, are taken up for a certain period of time2 by this “merry-go-round” on the stock exchange, can we say that the level of brokers’ loans is any sort of indication of the amount of funds that is thus tied up? If one dollar passes from brokers’ loans to new share capital, back to brokers’ loans and again to new share capital, and the same process repeats itself many times, the aggregate of brokers’ loans may rise very high, but the one dollar remains one dollar.

Part of these funds may be tied up in a chain of financial transactions,—

—which may raise the statistics of brokers’ loans to a multiple of these funds.

54. In our example the $10,000 of Mr. A and the $16,000 from the banks ($26,000 in all) performed the incredible task of financing a withdrawal of $6000 by Mr. B and new issues of $52,000 by corporations. But the task does not appear to be so incredible once it is realized that part of the financing of corporations is of a rather peculiar character, viz., that the share capital subscribed is lent to the share purchasers. It would of course be a mere coincidence if the new capital of any particular corporation were lent, through the agency of the banks and the brokers, to the corporation’s own shareholders; but there is nothing extraordinary about corporations in general lending part of their new share capital, via brokers’ loans, to the buyers of shares in general. What this amounts to is that the buyers of shares remain in debt to the corporations for a part of the price of the shares.

Corporations, by making loans to brokers, finance in part the purchase of their own shares by the new stockholders.

This fact is of no small significance. The circumstance that the buyers of the shares, even if only indirectly, have become debtors of the corporations, means in the first place that the money capital actually received by the corporations was less than the full amount of shares sold. These purchases of shares required neither the money of the shareholders nor bank credit.3 Since those who bought the shares and borrowed the sales proceeds (indirectly) from the issuing corporations did not need either the savings of the public or bank credit for the purpose, it is obvious that in this case neither the savings of the public nor bank credit can have flowed onto the stock exchange or have been absorbed by it.

For this borrowed part of new stock purchases neither savings nor bank credits are required.

The fact that the corporations, through their loans to brokers, have become creditors of the purchasers of shares is significant for another reason also. The credits are repayable at short notice. It would be rather surprising then if these credits were to be called in4 so gradually that the owners of the shares could pay them off out of current savings. In the absence of this possibility the repayment of the funds to the corporations can come from the following sources: (1) the gap may be filled by an increase in bank credit; (2) owners of liquid funds may buy the shares at low prices; (3) corporations may buy the shares, or other securities,5 at low prices.

Eventual payment to corporations may take place—

—out of current savings,—

In case (1) loans to the brokers by the banks on their own account are substituted for loans on the account of others. The new bank loans create bank deposits held by the corporations. In case (2) the aggregate of brokers’ loans undergoes a sharp decline. Bank deposits previously held by somebody as liquid reserves now become deposits of the corporations. If the corporations have a use for these cash balances, an act of dishoarding can be said to take place. In case (3) the total of brokers’ loans falls sharply just as in case (2). The corporations obtain shares especially cheaply from those who owed “them” (only indirectly “them”) the purchase price. Here bank deposits are neither transferred nor created nor destroyed. The corporations simply take securities in payment of loans outstanding.6

—through new bank credit,—

—with idle funds,—

—or by selling securities to the corporations;—

Those who believed rather naively that an enormous amount of stock exchange credit was absorbed by a rising stock market also thought, when they were consistent in their reasoning, that the credits absorbed were set free, either wholly or in part, when stock prices fell.7 A glance at the list given above of the possible ways in which brokers’ loans may be liquidated shows that the story of the release of the credits is no truer than the story of their absorption.

—and, of course, no funds are “released” by the decline in stock prices.

55. As has already been shown one dollar is capable of creating many dollars’ worth of brokers’ loans. One dollar may become sales proceeds and brokers’ loans and again sales proceeds and so on, and in this way produce a continual increase in the figure for brokers’ loans.

Large sums of brokers’ loans are made out of one dollar,—

But many dollars’ worth of brokers’ loans may also be created by no dollar at all. This may best be explained by going straight to an example. (Ledger balances are shown in Appendix A.)

—but also without any dollar.

Monday: Mr. A who possesses a large deposit of securities and therefore has sufficient margin wishes to buy $30,000 worth of shares on credit. His broker obtains them from another broker who is selling for the account of Mr. B.

Fourth illustration,—

Settlement takes place on the following day.

Tuesday: A’s broker borrows $30,000 from his bank and pays this sum to B’s broker. B’s broker has received no further orders from his customer and so credits him with the $30,000 and reduces his own bank debts by the same amount.

Mr. C asks his broker to buy him $20,000 worth of shares on credit. His broker obtains them from the broker of a Mr. D. This transaction will not be settled until to-morrow.

Up to now the total amount of brokers’ loans has not changed although the margin debts of customers to brokers have risen.

Wednesday: Mr. B withdraws his $30,000 from his broker. The broker borrows the money from his bank. Mr. B offers to lend the $30,000 at call. The amount is borrowed by Mr. C’s broker. Mr. C’s broker uses $10,000 to reduce his bank debt and $20,000 to pay Mr. D’s broker. Mr. D’s broker credits his customer with the $20,000 and uses the funds to reduce his own bank debt.

—which shows that brokers’ loans from banks rise when sellers withdraw their funds, but fall again immediately when sellers use the funds for making loans to brokers;—

On this day the total of loans granted to brokers by the banks on their own account has not risen, but brokers’ loans on account of others have risen by $30,000.

—that total brokers’ loans rise when sellers withdraw and lend 50 brokers;—

Thursday: Mr. D buys $5000 worth of shares. His broker obtains them from Mr. A’s broker who is selling for the account of Mr. A.

Settlement takes place to-morrow.

Friday: Mr. D withdraws $10,000 from his broker. The broker borrows these $10,000 and the $5000 which he owes to A’s broker from his bank. Mr. D offers to lend the $10,000 at call. They are borrowed by Mr. A’s broker, who adds them to the $5000 which he received for the shares sold on behalf of Mr. A, and pays back $15,000 to his bank.

On this day loans granted to brokers by the banks again remain unchanged. Brokers’ loans on account of others have increased by $10,000. The margin debts of customers to their brokers have been reduced by $5000 (through Mr. A’s sale).

—and that brokers’ loans may increase while customers’ margin debts decrease.

Review of the week: The total of brokers’ loans has risen by $40,000. The whole of the increase was on account of “others” and the loans granted by the banks remained unchanged. The margin debts of customers to their brokers have risen by $45,000 (Mr. A $30,000 minus $5000, Mr. C $20,000) and the brokerage deposits of customers have risen by $5000 (of Mr. D).8

In our example the rise in brokers’ loans on account of “others” did not lead to any fall in loans to brokers from the banks, because those who lent the call money used the proceeds of their sales and not funds which they had held previously. If the call loans had been financed, say, by Messrs. X and Y out of their bank balances instead of by Messrs. B and D out of their sales proceeds, these funds would have led to a net repayment of brokers’ loans of the banks.9 Here, however, it was the sales proceeds which were employed to make the loans; the brokers took the loans in order to pay out the sales proceeds; the customers asked for payment of their sales proceeds in order to make loans.

Brokers loans by “others” reduce bank loans, if the funds come from existing bank deposits but not if they come from sales proceeds and old brokerage deposites.

56. If the sellers leave their sales proceeds with their brokers, the volume of brokers’ loans does not rise despite the rise in margin debts of the buyers. The brokers can lend to those who want to buy on margin without themselves borrowing for the purpose, provided the sellers leave their sales proceeds on deposit with the brokers. In this case the sellers wait for payment by holding brokerage deposits.

The sellers may actually wait for payment, either by holding brokerage deposits,—

But now the owners of the brokerage deposits may decide to withdraw their funds and transfer them back to the brokers in the express form of loans. The only difference between this and the previous situation is that the brokers now have to pay interest and that the statistics of brokers’ loans show an increase. This does not, however, alter the fact that the buyers of securities still owe the price to the sellers.

—or by making, brokers’ loans;—

—only the second method affects statistics of brokers’ loans.

It is evident that this process does not involve either any inflow of funds or any tie-up of funds, but nevertheless the volume of brokers’ loans rises. Mr. M buys shares from Mr. N but does not pay for them. Mr. N lends the sales proceeds due to him to the broker and through the latter to the purchaser M. This simple fact that M buys from N because he expects share prices to rise and N lends the sales proceeds because he gets interest on them, is registered in the statistics as a rise in “brokers’ loans on account of others” or “from others than banks.’ And this statistical phenomenon has misled a large number of authors into concluding that the stock exchange absorbed an alarming proportion of the country’s credit supply.1

There is not any afflux, and still less, any tie-up, of funds; and yet figure of brokers’ loans rise when the seller indirectly lends to the buyers.

This lending by the seller of the shares to the buyer—however dangerous it may be from the point of view of market stability—has deprived nobody of either money or credit. The purchaser did not take money away from anybody else by making the purchase, because he did not have or use any money. The seller did not take credit away from anybody by lending it to the stock exchange, because he could not have lent to anybody other than the buyer of his shares since he was only able to sell the shares at a favourable price by disposing of them to the buyer who had no funds.2 M had no money with which to finance productive investment and neither had N. N had shares which he sold to M on credit. This transaction appeared in the statistics as an increase in stock exchange loans.

The credit given by the seller to the buyer could not be given to anybody else.—

—because neither of the two has liquid funds.

57. When these loans are called in, it usually happens that the owner of shares who is in debt is at best able to scrape together a small part of what he owes by compelling himself to save out of his current income from other sources (salaries, business profits). But this is, of course, far too little, and he is forced to liquidate his holdings of shares. And the person who is most able and willing to buy is the person who lent the call money.3 He is in a liquid position, not in the sense that he has bank notes or bank deposits, but because indirectly he holds the claims against the owners of the shares, claims which have to be paid by the sale of those shares.

When the lender of call money buys, after the crash, the stocks from the margin debtor,—

The shares which our Mr. M now has to sell at a low price are bought by Mr. N. The seller does not obtain any money for them because he was in debt for them: and just as the seller has no money to receive, the buyer has no money to pay. The volume of loans outstanding may fall, just as they rose previously, without there being any “inflow,” “outflow,” “creation,” or “destruction” of bank credit. The claims against the unlucky speculators disappear when their creditors buy up their shares from them.

—the buyer has nothing to pay and the seller nothing to receive;—

—ye figures of brokers’ loans fall.

To complete the exposition, the chain of operations may again be illustrated by an example. (Ledger balances are shown in Appendix A.)

Fifth illustration,—

Monday: Mr. A receives a demand from his broker to put up more margin because the securities held for him have depreciated in value. Mr. A decides to sell securities which realize $15,000. The shares are bought by Mr. B’s broker on Mr. B’s behalf.

Settlement takes place to-morrow.

—which shows that customers’ margin debts are settled mainly through forced sales:—

Tuesday: Mr. B calls in $15,000 of the call loans he has outstanding. The broker who had borrowed these funds now borrows a bank loan in order to repay B. B pays the $15,000 to his broker who in turn pays it to A’s broker as the price of the shares he bought from him. Mr. A has paid in another $1000 in cash to his broker. A’s broker uses the $16,000 to pay back to his bank.

—that brokers’ loans “by others” fall through stock purchases by call-money lenders;—

On this day the volume of loans to brokers granted by the banks on their own account has diminished by $1000 only, whereas the volume of brokers’ loans on account of others has diminished by $15,000. The margin debts of customers have diminished by $16,000.

—and that brokers’ loans by banks fall, when outside funds are paid in.

It is unnecessary to give further examples of the transactions leading to the liquidation of brokers’ loans. The process is not really a very complicated one. It will suffice to add that it is also possible that B may ask for his call loan to be repaid before he decides to purchase shares. In this case loans to brokers from the banks for their own account will rise for the time being and those on account of others will fall. The volume of brokers’ loans from the banks will fall again later when B buys the low-priced shares.

The fall in the total volume of brokers’ loans following on a break in stock prices comes about essentially in three ways:

“Brokers’ loans fall—

(1) speculators whose accounts are undermargined pay in what they can afford in order to maintain their positions;

(2) owners of bank deposits buy the shares sold at low prices by speculators who are forced to reduce their debts;

(3) owners of funds previously lent out as call money buy the shares sold at low prices by speculators who are forced to reduce their debts.

What is there to be said about these possibilities as regards their effects on the amount of purchasing power going to other markets?4 In case (1) sums which would otherwise have become effective demand on the markets for goods are used to repay loans which brokers had previously borrowed from the banks: this will result in the disappearance of a certain quantity of bank assets and of circulating bank deposits. There will, of course, be a rise in excess reserves and therefore in the capacity of the banks to grant credit to other borrowers, but for the moment there will undeniably be a deflationary effect. This case is the real counterpart of the rise in loans from the banks to the brokers which had an inflationary effect on the markets for goods.

—when margin debtors put up more money through new saving, which is deflationary in its effect on other markets;—

Case (2) can only be judged if we know how the bank deposits of the buyers of the shares would have been used if the share purchase had not taken place. It is possible, though not very probable, that funds will be withdrawn from the markets for goods in this case also.5 It is more probable that the funds will come out of idle balances. If these now lead, through the share purchase, to the eventual wiping out of a certain amount of bank credit, there is no net deflationary effect. The raising of the excess reserves of the banks through the cancellation of these inactive deposits actually increases the potential supply of new credit in the future.6 In case (3) where call loans are withdrawn in order to purchase shares, and shares are sold in order to pay back call loans, there is nothing which would have any effect either actual or potential on the effective demand in other markets.

—or when holders of bank deposits buy shares from the margin debtors, which mostly means the disappearance of idle balances;—

—or when call-money lenders buy shares from the margin debtors, which leaves other markets untouched.

58. The conclusions of the last nine sections are sufficient to shake all confidence in the significance of the statistics of brokers’ loans. As brokers’ loans can rise for so many different reasons, it is quite impossible to diagnose the situation merely on the basis of the aggregate figures for these loans. It remains impossible, no matter how perfect a correlation can be shown to exist between the volume of brokers’ loans on the one side and the turnover of stocks, the level of stock prices or the velocity of circulation of bank deposits on the other. The most naϊve interpretation of all was that which said that brokers’ loans represented credit tied up in stock exchange transactions. But even rather more “enlightened” interpretations prove to be untenable when regard is had to the analysis of this chapter. Take, for instance, the contention that the total of brokers’ loans represents the amount of funds that have flowed through the stock exchange into industry; or the idea that the truth lies somewhere in the middle, i.e., that brokers’ loans represent funds which have flowed onto the stock exchange, and part of these funds flows out into “productive” markets and part is tied up. None of these arguments is tenable since a rise in brokers’ loans does not necessarily warrant the conclusion that there has been a flow of funds onto the stock exchange.

The analysis shows no valid conclusion can be drawn from changes in brokers’ loans

Below is an attempt to draw up a list of the various kinds of operations which may lie at the back of an increase in brokers’ loans. The list is undoubtedly incomplete but will nevertheless be sufficient for our purposes. In all cases it is assumed that somebody has bought securities on borrowed funds.

Ten types of operations leading to a rise in brokers’ loans are summarized.

This transaction may be connected with any of the following operations:

(1) industrial corporations have issued new shares, received money (bank deposits) for them and spent it on real investment;

(2) individual business men or firms, who had previously invested part of their funds in shares, have sold shares, received money (bank deposits) for them, and spent it on real investment in their own businesses;

(3) individuals, who had previously invested part of their funds in shares, have sold shares, received money (bank deposits) for them, and spent it on consumption;

(4) individuals, who have made capital gains as a result of the rise in share prices, have realized their gains, withdrawn them in the form of money (bank deposits), and spent them on consumption;

(5) corporations have issued new shares but have used the proceeds immediately to grant loans to brokers;

(6) individuals or firms have sold shares from their holdings but have used the proceeds immediately for granting loans to brokers;

(7) corporations have issued new shares and have the money proceeds (bank deposits) in their accounts, for a few days, until their use in further financial transactions;

(8) individuals and firms have sold old shares from their holdings and have the money proceeds (bank deposits) in their accounts, for a few days, until their use in further financial transactions;

(9) corporations have issued new shares and leave the money proceeds (bank deposits) lying idle as liquid cash reserves;

(10) individuals or firms have sold shares from their holdings and leave the money proceeds (bank deposits) lying idle in their liquid cash reserve.

A real inflow of funds (money capital that has either been newly saved, or newly dishoarded, or newly created out of bank credit) has taken place in cases (1–4) and (7–10). In cases (5) and (6) there has been no inflow at all, or at least not as far as the end effect is concerned. (See §§ 54, 55, 56. Bank credit, for instance, which was created for the purpose of paying out funds to the sellers was, if they used it for granting loans to brokers, repaid and so disappeared again.) In cases (1–4) the funds flowing onto the stock market were spent on the markets for commodities; in cases (1) and (2) they were used for production, and in cases (3) and (4) for consumption. In cases (7–10) the funds flowing onto the stock market were not spent on the markets for goods; in cases (7) and (8) the money (bank deposits) continued to circulate in the financial sphere, and in cases (9) and (10) it went into the idle cash reserves of pessimistic hoarders.

No inflow of funds in two cases;—

—funds used for buying producers’ goods in two cases; for buying consumers’ goods in two cases tied up in financial transactions in two cases; and locked up in idle balances in two cases.

Nowhere are there any statistical data to show how the total volume of brokers’ loans at any time is distributed over these ten items. To anybody with a sense of proportion, however, it would appear that cases (7) and (8), the tie-up of funds (bank deposits) in stock exchange transactions, cannot be responsible for more than an extremely small fraction of brokers’ loans. Neither can it reasonably be held that cases (9) and (10), hoarding by pessimistic holders of money, are responsible for the whole or the greater part of brokers’ loans as some writers seem to think. (This topic will be taken up in the next chapter.)

The smallest part is played by the absorption cases.

A high official of the Federal Reserve System expressed only recently the following opinion7: “We are inclined to conclude that the best evidence on whether expansion of credit through an increase in security loans has a stimulating effect on business or is ‘absorbed’ by the stock market, is to be found in data on changes in business volume and in prices . . .” He proposes to investigate whether “the expansion in business [was or] was not in proportion to the expansion in credit if all brokers’ loans are included in the credit figures.” Here, then, brokers’ loans on account of others are expressly included in the credit expansion, and if they are found not to have resulted in any business expansion they are to be regarded as having been absorbed by the stock exchange. The preceding sections have demonstrated that this point of view is untenable because an increase in brokers’ loans on account of others than banks seldom means a further expansion of credit. Lending by the seller of the shares to the buyer, which finds expression in an increase in brokers’ loans, can certainly not lead to a business expansion, but neither can it be regarded as being “absorbed by the stock exchange.”

To repeat once more our main conclusion: figures giving the sum total of brokers’ loans tell us absolutely nothing about the absorption of credit by the stock market.

____________________

J. M. Keynes has hinted several times at the case of the lending of sales proceeds (e.g., Treatise on Money, Vol. I, p. 267, and Vol II, p. 196). When he came to the interpretation of the level of brokers’ loans, however, he did not think of the possibility that loaned sales proceeds might be included. We shall comment on Mr. Keynes’ interpretation in the next chapter.