by investors, and are not carried in quantity by brokers, hence, could not be readily borrowed. But, in active stocks, there is no difficulty whatever in borrowing.

The reason is this: Every broker who carries many stocks employs a great deal more money than he possesses. In theory, a broker carrying for a customer 100 shares of Union Pacific at par would make up the money for the purchase by using $1,000 belonging to the customer, $1.000 of the money of the brokerage firm, and then borrow $8,000 from a bank on the security of the 100 shares of stock purchased.

An active broker, consequently, is always a large borrower of money, and when he borrows from a bank he is expected to put up 20 per cent. margin on his loan. But if he can lend stocks he gets the full value of the stock and does not have to put up any of his own money or of his customer's money. Hence, every broker is willing to lend stocks, particularly when the demand for stock is sufficient to make the rate of interest lower than the market rate, as the broker in this case makes a profit by charging his customer who is long 5 or 6 per cent. interest, while he perhaps secures his money without any cost through lending the stock flat.

This, from the standpoint of the short seller, is what makes his operation practically safe. Ordinarily, it is just as easy to borrow active stocks as it is to borrow money, and squeezes of shorts through inability to borrow are little if any more frequent than squeezes of "longs" through the difficulty of brokers in borrowing money.

Squeezes of shorts sometimes develop themselves and

are sometimes manipulated. When friends of a property see a large short interest they sometimes try to persuade holders of the stock to agree not to lend it for a day or two and thus scare shorts to cover by difficulty in borrowing. If this undertaking is successful brokers are notified to return borrowed stock, and when they try to borrow elsewhere they find little offering. The loaning rate possibly runs up to 14 per cent. a day, or perhaps higher.

Shorts are alarmed and cover, advancing the price of the stock and enabling holders to sell at a profit. Such a squeeze usually lasts only two or three days, as by that time the advanced price leads those who have the stock to either sell it or lend it, and the price then usually goes lower than before. Sometimes there is a short interest so large and so persistent as to keep a stock lending at a premium for some time. This is usually almost certain evidence of decline, but the expenses of premiums and the necessity of paying dividends sometimes eat up the profits so that but little remains even after considerable fall in price. Mr. Gould is said to have once remained short of New York Central over four years, and to have had a large profit as between his buying and his selling price, but to have had the greater part of it eaten up in dividends.

In picking out a stock to sell short, the first consideration ought to be that the price is above value, and that future value appears to be shrinking. It should be an active stock and, if possible, a stock of large capital. It should be an old stock by preference, which means having wide distribution instead of concentrated ownership. By

preference it should be a high priced stock with a reasonable probability that dividends will be reduced or passed.

Such a stock should be sold on advances and bought in on moderate declines, say 4 or 5 points, as long as the market seems to be reasonably steady. But, if the market becomes distinctly weak, only part of the short stock should be bought in with the hope that some short interest may be established at a price so high as to be out of reach of temporary swings. The best profits in the stock market are made by people who get long or short at extremes and stay for months or years before they take their profit.

CHAPTER XVI.

*SPECULATION FOR THE DECLINE.

The question is frequently asked whether in taking a bearish view of the general market it is expected that all stocks will go down together or that some will fall and others not.

The answer to this question takes two forms—the first 18 The speculative movement; the second the effect of values. When the market goes down, especially if the decline is violent or continued, all stocks fall; not perhaps equally, but enough to be regarded as participating fully in a general decline. Indeed, it often happens that a stock of admitted large value will fall more in a panic than a stock of little value.

The reason is that when people have been carrying various stocks, some good and some bad, and a time comes when they are obliged to suddenly furnish additional margin or reduce their commitments, they try to sell the stocks for which they think the market will be best, namely, their best stocks. But the very merit of such stocks prevents the existence of a short interest, hence when considerable amounts are offered in a panic there is no demand for covering purposes, and, in fact, no demand except from investors who may not know of the decline or who may not have money for investment at that particular moment. Consequently the good stock drops until it

*Dow's Theory.

meets an investment demand somewhere. This condition was illustrated by the action of Delaware & Hudson in the panic of May 9, 1901. It had nearly, if not quite, the largest decline of any stock on the list, falling in half an hour from 160 to 105, chiefly because people generally did not know the price at which stock was being offered.

It may be accepted, therefore, that in a general decline merit in a stock will not count for the time being. Good and bad will decline measurably alike. But here comes in a marked distinction. When the recovery comes, a day or a week later, the good stock will recover more and hold its recovery better than the poor stock. Delaware & Hudson is again a good illustration. After the quotation of 105 was printed on May 9 orders to buy the stock came from all sections, and in another hour the price was in the neighborhood of 150.

Value will always work out in the course of time. A stock intrinsically cheap and a stock intrinsically dear may be selling at the same price at a given time. As the result of six months' trading they may have presented the appearance of moving together in most of the fluctuations, but at the end of the period the good stock will be 10 points higher than the poor one, the difference representing a little smaller decline and a little better rally in each of five or six swings.

This exactly describes what will occur all through the market during the next bear period, whenever that period There will be a sifting of the better from the worse, visible enough at a distance, but not conspicuous at any particular stage in the process.

comes.