to make good the drafts on the cash balances that the Government had with it. That, in brief, was the machinery employed in financing these short loans, which resulted in the growth of bank deposits, and bank loans and, in turn, in the growth of discounts at the Federal reserve banks. It was a policy that was in existence during the entire period, I should say, down to about March, April, or May, of 1920, at which time it was regarded as safe to permit trading in those certificates in the open market, so that they would find their natural interest level-investment level. From that time on the rates were more nicely adjusted, so that they would be distributed to the public. They then became a rather important element in contracting the loan account rather than in expanding it. I suppose to-day there can not be more than 10 per cent, or thereabouts, of all the outstanding issues held by the banking institutions of the countrypossibly a little more than that-whereas at the time of the heaviest borrowing, when the issues ran for a short period at the rate of $750,000,000 a fortnight, something like 80 per cent, or over 80 per cent, or thereabouts, were owned by the commercial banks and trust companies.

The CHAIRMAN. Referring again to this question of mortgage loans by banks, I have some figures here which indicate that the total loans to the farmers loaned on farm mortgages, in 1918, amounts to 4 per cent; in 1920, 5.52 per cent, the country over, but in the State of Minnesota the proportion was 19 per cent; in Iowa, 23 per cent; in Missouri, 23 per cent; in North Dakota, it was 14 per cent. These figures are fairly representative in the agricultural States. Gov. STRONG. Yes.

The CHAIRMAN. So that in the agricultural States it is quite clear, I think, that a very considerable proportion of the funds of the banks are employed in mortgage loans.

GOV. STRONG. How do those figures compare with similar figures in those States for 1913?

The CHAIRMAN. Unfortunately, I can not give you the figures for

1913.

GOV. STRONG. Is not that the answer, however, as to whether they did increase, in response to the amendment of the national bank act?

The CHAIRMAN. The proportion of loans made by national banks, as compared with State banks, I suppose, is very small. Gov. STRONG. Shall I proceed, Mr. Chairman?

The CHAIRMAN. If you will, please.

GOV. STRONG. One of the most important causes to which is assigned the responsibility for the decline in prices, and the difficulties that the farmers have encountered in marketing their produce, has been the closing of the foreign markets. There is difficulty in moving our surplus products of all kinds, including agricultural produce, to foreign markets, and I think it is generally attributed, in the popular mind at least, to the fact that adequate credit facilities do not exist for financing that movement, and especially for financing the purchasers abroad-the people abroad who would like to buy on credit, but can not get the credit.

In fact, I am inclined to other causes are more

Now, I am not sure that that is the reason. think that it is the least reason, and that important than that. The most important of these is the extent

to which the value of foreign currencies, measured in dollars, fluctuates up and down. There are, of course, a good many causes, and credit is one of them. At one time transportation was one cause, and when the war broke out the difficulty of getting insurance was a cause. But at the present time the condition of all the foreign exchanges is such that I am rather disposed to think that violent fluctuations in exchange rates are as great a deterrent against buying

MOVEMENT OF REPRESENTATIVE FOREIGN EXCHANGES.

Percentage above or below par of foreign exchanges at New York.

Source of information: European exchanges, Times Annalist, and Commercial and Financial Chronicle; Argentina, Davis & Co., New York City.

farm and other goods as anything; that is, as related to foreign markets.

The CHAIRMAN. Let me see if I understand you. The possibility of exchange rates changing adversely, increases an element of uncertainty in a hazard which otherwise would not appear?

Gov. STRONG. Yes, sir.

The CHAIRMAN. And I suppose it is true generally that every increased hazard, or every uncertainty that is injected into every

sort of business transaction not only increases the difficulty of the transaction, but usually widens the margin or price.

Gov. STRONG. It does; and that difficulty requires a little analysis to see just how it operates. Fluctuations in foreign exchange-and I will ask you to distinguish between the fluctuation in the exchange and the normal operation of an exchange market within a narrow range. Fluctuations in foreign exchange are generally illustrated as though they resembled a rise and fall in a barometer, say, or in a thermometer.

Now, that is not a good illustration, because a fluctuating foreign exchange is one of the effects of a condition which reacts upon the cause that is, fluctuations in foreign exchange and their effect should be likened, I think, to the operation of a thermostat. When exchange is rising-that is to say, when dollars are increasing in value all over the world as they have recently-almost since the World War started-the effect of that is gradually to close our export door and open our import door. The reason for this is obvious, because as a foreign purchaser buys our goods he has to pay for them in dollars, and the dollars cost him an increasingly higher amount, expressed in his own currency, and consequently the goods that he has purchased have got to be sold in his own country at an advancing price in his own currency. The effect of it is that while that advance is taking place, and before the readjustment in world prices is made to correspond to the new level of exchange, he is facing a new cost of those goods so that he must advance the price at home at which he sells them. If those goods are raw materials which must go through the process of manufacture, and if it takes time to transport them and he has not got dollars to lay down when he makes his original contract for purchase, then he faces the possibility of a very serious loss when he does buy his dollars in order to pay for the goods. But it works both ways. Suppose he has been shrewd enough to buy only those things for which he has ample funds in dollars to make cash payment and insures himself in that way against an advance in dollars, he is not. however, insured against a decline in dollars which may occur at any time, whereupon some competitor of his, having purchased in a declining market for dollars, is able to undersell him at home. I bring that out to illustrate the point that the difficulty is the fluctuation in exchange.

Now, as you have stated, the great elements of risk in every commercial transaction are the risk in the fluctuation in the value of the goods purchased between the time of purchase and the time of resale: and the second great element of risk in a commercial transaction is the goodness of the credit of the man to whom the goods are sold by the merchant in case he sells on credit.

Far beyond anything possible at the time when the world was settling balances by gold shipments, exchange fluctuation introduces an element that has such an effect upon the ability of the operators to operate with certainty that it has a very serious effect upon the amount they are willing to buy, when payment must be made in foreign currencies. In fact, those who buy goods in the United States to-day and subject their transactions to these risks of foreign exchange fluctuation are, in a sense, dealing in futures-that is, they are dealing in exchange futures, and are liable to be beaten in market prices by some one

who has the credit and employs it when exchange is more favorable. So that what we should aim at, if possible, in eliminating that difficulty is to eliminate the fluctuation in exchange. It is perfectly obvious that the American exporter or would-be exporter of goods is not going to take that risk. If the American producer wants to sell cotton or wheat or any other commodity abroad, he wants to get a contract in dollars, and then he knows that he is going to get his cost and a profit back, assuming that the credit is good. Now, the foreign buyer is in no different position whatever. He does not want to buy those goods unless he knows he has the means of payment without loss. It operates, if you please, at both ends of the line.

The American producer insists on the buyer taking the risk of the exchange, and the buyer would naturally seek to put that risk and responsibility on the seller. And, if you please, that operates more especially in the case of that kind of trade which is usually conducted upon credits of three months or six months or any length of time. The buyer of goods is subjected to all the hazards of exchange fluctuation over quite a long period. That is one of the answers to the question that has been so frequently raised, Why don't we get up big organizations to sell goods abroad on long credit? If we do get them up, and if we do have the goods to sell and to export, how are we going to extend long credits to buyers who do not know, within a range of 5, 10, 15, or 25 per cent how much they are going to be called upon to pay when that credit matures? And that is exactly the situation in an exchange market which fluctuates over as wide a range as all the exchanges have covered within recent months or years. It was not very long ago that we saw sterling exchange selling for $4 or better. A few weeks ago we saw it selling for $3.53, and all those who had entered into contracts for goods when sterling was at the rate of $4 and who had expected to get dollars at that rate four or five or six months thereafter have all lost, of course, in getting dollars costing the rate of $3.53, unless they have been fortunate enough to readjust prices in their own country to a point where they can recover that loss.

The readjustment of the international price level does not take place as rapidly as exchange fluctuates at the present time. It is a rather slow and ponderous movement as compared with exchange, which may vary 10 or 15 cents on the pound in a few days, as we have seen recently.

I have referred to what happened in India, due to various causes, when I was there. The rate expressed in sterling at which a merchant could convert rupees into sterling at one time, shortly before I reached India, was such that for one rupee he could get 2 shillings and 10 pence in English currency, and within a period of a few months or a few weeks, almost, rupees declined steadily to a point where he could buy only 1 shilling and 6 pence. And I believe it got below that; I believe at one time it was 1 shilling and 3 pence. Consider the plight of the importer who had purchased goods in England, say cotton cloth in Lancashire, and at the time of his purchase sterling was selling at the rate of 2 shillings 10 pence per rupee; he might have had six months' credit, and turned around and sold all those goods and made a handsome profit, and then when he came to convert the proceeds into sterling he would find that his rupees were only worth 1 shilling and 3 pence. It would cost him twice as much

as at the time he had entered into the contract. That is one of the extreme cases of fluctuating exchange, and of losses that may result, even with credits assured, and with credits perfectly good.

The point, summed up in one word, is that merchants will not take that third and unusual risk.

Now, there are a variety of reasons for these fluctuations. Of course, the principal reason is that in ordinary times, before the war, exchange in its fluctuations rarely got beyond the point where it was profitable for gold to move; that is, beyond the "gold point," as commonly expressed.

The CHAIRMAN. According to my understanding, that represents the cost of the shipment of gold from time to time?

Gov. STRONG. Yes; to illustrate it simply for the purpose of the record, if it is of any value, Mr. Chairman, the exact parity, in round figures, of the gold in a sovereign, expressed in dollars, is $4.863. Now, if an American importer before the war incurred a liability for a round sum of money for goods imported for which he was later to make payment, say, three months following the importation of the goods, and then found that sterling had advanced to $4.90 or $4.91, in response to various conditions entering into the market, it might be easier and cheaper for him to pay for those goods by going to his bank and getting the gold and shipping it over, rather than buying the exchange. And he can do that when exchange passes the "gold point", even considering the cost of shipping the gold, that is, the packing and shipping charges, insurance, loss of interest, abrasion on shipment, and so on. Therefore, the rates of exchange before the war were more or less automatically limited by the fact that the nations were settling these balances due to and from each other, when they got out of balance, by the actual shipment of gold from one country to the other. Therefore the exporters and importers were insured against heavy loss, because a limit was fixed upon the fluctuation of exchange by the actual shipment of gold itself.

Now, I was referring to the various causes of this fluctuation. The underlying cause, of course, lies in the fact that the world's trade is out of balance, and has been ever since the war started. People have imported more goods in some parts of the world from this country than they are able to pay for easily. Other causes have also added to the difficulty. If it had not been so greatly out of balance if the amount of international debt incurred in connection with the war had been less, there would have been no occasion for the suspension or control of gold shipment, and exchange adjustments or settlements would have been effected by gold shipments when exchange passed the "gold point," so to speak. I am purposely omitting references to depreciated currencies.

And one of the evidences of the fact that this indebtedness is considerable and is still not settled fully appears from the fact that we are now receiving very large amounts of gold in payment of bills or debts due in this country, and still sterling and other exchanges do not come back to anything like parity; that is, most of them have not. So that, in general, one of the underlying causes lies in the fact that heavy debts must be paid, and the reason that the exchange can not be controlled is that gold or other goods can not be shipped or credits arranged in sufficient quantity to pay those debts.