In More Money Than God, Sebastian Mallaby describes how Alfred Winslow Jones, the world’s first hedge fund manager, made his money. Between 1949 and 1968, his fund posted a cumulative return of just under 5,000 percent, rivaling Warren Buffett’s performance.

I think Jones’ case study is interesting today as a datapoint around whether markets or efficient or individuals can beat them. His example indicates it’s possible to be genuinely more conceptually sophisticated than the competition.

Primarily, his “edge” lay in pioneering two strategies that are widely deployed by modern hedge funds: leveraged hedging, and the competitive multimanager structure.

Leveraged hedging

AW Jones used leverage (borrowing money) and short selling (betting against a company by borrowing its stocks from other investors at interest and selling them in the expectation that their price would fall, at which point they could be repurchased at a profit), both of which had earned a reputation for being too risky for professional investing following the 1929 Wall Street crash. But he used these techniques to “hedge” his stock picks against broad market movements (thus making his fund a “hedged” fund). He was using “speculative means for conservative ends”, as he put it.

To explain how his method de-risked his investing, it’s best to repeat Mallaby’s illustrative example from the book.

Suppose there are two investors, each endowed with \$100,000. Suppose that each is equally skilled in stock selection and is optimistic about the market. The first investor, operating on conventional fund-management principles, puts \$80,000 into the best stocks he can find while keeping the balance of \$20,000 in safe bonds. The second investor, operating on Jones’s principles, borrows \$100,000 to give himself a war chest totaling \$200,000, then buys \$130,000 worth of good stocks and shorts \$70,000 worth of bad ones. This gives the second investor superior diversification in his long positions: Having \$130,000 to play with, he can buy a broader range of stocks. It also gives him less exposure to the market: His \$70,000 worth of shorts offsets \$70,000 worth of longs, so his “net exposure” to the market is \$60,000, whereas the first investor has a net exposure of \$80,000. In this way, the hedge-fund investor incurs less stock-selection risk (because of diversification) and less market risk (because of hedging).

It gets better. Consider the effect on Jones’s profits. Suppose the stock market index rises by 20 percent, and, because they are good at stock selection, the investors in Jones’s example see their longs beat the market by ten points, yielding a rise of 30 percent. The short bets of the hedged investor also turn out well: If the index rises by 20 percent, his shorts rise by just 10 percent because he has successfully chosen companies that perform less well than the average. The two investors’ performance will look like this:

Traditional investor Hedged investor
30% gain on \$80,000 worth of stocks 30% gain on \$130,000 worth of stocks
  10% loss on \$70,000 worth of shorts
Gain: \$24,000 Net gain: \$39,000 - \$7,000 = \$32,000

The result appears to defy a basic rule of investing, which is that you can only earn higher returns by assuming higher risk. The hedged investor earns a third more, even though he has assumed less market risk and less stock-selection risk.

Now consider a down market: The magic works even better. If the market falls by 20 percent, and if the stocks selected by the two investors beat the market average by the same ten-point margin, the returns come out like this:

Traditional investor Hedged investor
10% loss on \$80,000 worth of stocks 10% loss on \$130,000 worth of stocks
  30% gain on \$70,000 worth of stocks
Loss: \$8,000 Net gain: \$21,000 - \$13,000 = \$8,000

In sum, the hedged fund does better in a bull market despite the lesser risk it has assumed; and the hedged fund does better in a bear market because of the lesser risk it has assumed. Of course, the calculations work only if the investors pick good stocks; a poor stock picker could have his incompetence magnified under Jones’s arrangement.

Jones was a decade ahead of academia in inventing the capital asset pricing model (CAPM):

[S]horting only works as part of a hedging strategy once a further refinement is brought in. It was here that Jones was way ahead of his contemporaries.

The refinement begins with the fact that some stocks bounce up and down more than others: They have different volatilities. Buying \$1,000 worth of an inert stock and shorting \$1,000 worth of a volatile one does not provide a real hedge: If the market average rises by 20 percent, the inert stock might rise by only ten points while the fast mover might shoot up by thirty. So Jones measured the volatility of all stocks—he called it the “velocity”—and compared it with the volatility of Standard & Poor’s 500 Index. For example, he examined the significant price swings in Sears Roebuck since 1948 and determined that these were 80 percent as big as the swings in the market average: He therefore assigned Sears a “relative velocity” of 80. On the other hand, some stocks were more volatile than the broad market: General Dynamics had a relative velocity of 196. Clearly, buying and selling the same number of Sears and General Dynamics stocks would not provide a hedge. If the Jones fund sold short 100 shares of volatile General Dynamics at \$50, for example, it would need to hold 245 shares in stodgy Sears Roebuck at \$50 to keep the fund’s market exposure neutral.

In his report to his investors, Jones explained the point this way:

We buy We sell short
245 shares in Sears Roebuck at \$50 = \$12,250 100 shares in GD at \$50 = \$5,000
\$12,250 x Sears’ velocity, 0.80 = \$9,800 \$5,000 x GD’s velocity, 1.96 = \$9,800

Jones pointed out that the velocity of a stock did not determine whether it was a good investment. A slow-moving stock might be expected to do well; a volatile one might be expected to do poorly. But to understand a stock’s effect on a portfolio, the size of a holding had to be adjusted for its volatility.

Jones’ approach was later vindicated by finance theory, especially James Tobin’s separation theorem, which held that an investor’s stock picks should be separate from the question of their risk appetite:

Most investment advisers in the 1950s assumed that certain types of stocks suited certain types of investor: A widow should not own a go-go stock such as Xerox, whereas a successful business executive should have no interest in a stodgy utility such as AT&T. Tobin’s insight was to see why this was wrong: An investor’s choice of stocks could be separated from the amount of risk he wanted. If an investor was risk averse, he should buy the best stocks available but commit only part of his savings. If an investor was risk hungry, he should buy exactly the same stocks but borrow money to buy more of them. Yet nine years before Tobin published his ground-breaking article, Jones was onto the same point. His fund made one judgment about which companies to own and a second about how much risk to take, adjusting the risk as it saw fit by using the device of leverage.

The competitive multimanager structure

Jones’ hedged strategy managed risk in a much more sophisticated way than his rivals did, but by itself it could not be the source of Jones’ marvelous profits. The Jones fund’s top-of-the-line stock choices came not from Jones himself (who actually lost the fund some money trying to time the market), but from an assembly of stock pickers. Jones’ innovation lay in coming up with a way to attribute fund performance to individual stock pickers, thus incentivizing individual competence.

He came up with a method to distinguish between the money made through skillful stock picking (what’s now called “alpha”) and that made through simple exposure to the market (“beta”). Each evening, he would look at the closing prices of the stocks in his portfolio and construct chains of reasoning like this one:

Our long stocks, worth \$130,000, should have gone up by \$1,300 to keep pace with the 1% rise in the market. But they actually went up by \$2,500, and the difference, attributable to good stock selection, is \$1,200 or 1.2% on our fund’s \$100,000 of equity.

Our short stocks, worth \$70,000, should have gone up also by 1%, which would have shown us a loss of \$700. But the actual loss was only \$400, and the difference, attributable to good short stock selection, is a gain of \$300 or 0.3%.

Being net long by the amount of \$60,000, the market rise of 1% helped us along by 1% of \$60,000, or \$600, or 0.6%.

Our total gain comes to \$2,100, or 2.1% of equity. 1.5 percentage points of the return were attributable to stock selection. The remaining 0.6 percentage points stemmed from exposure to the market.

Crucially, being able to measure alpha allowed Jones to proportionally compensate traders who generated it. As Mallaby describes it, in the 1950s, Wall Street was a “sleepy, unsophisticated place”. Mutual fund companies paid out thousands of dollars to salesmen who brought in investors’ capital, leaving little to compensate good research. Trustees at investment institutions were compensated by the volume of assets under management rather than a performance fee, and they reached decisions by committee. Even in the 1960s, Mallaby writes, “when Wall Street finally shrugged off its postcrash stupor, it was surprising how easily sheer diligence could set a man apart. Alan Dresher, one of the Jones stock pickers, had the idea of going over to the Securities and Exchange Commission offices to read company filings the moment they came out. The extraordinary thing was that he was all alone. The rest of the Street was waiting for the filings to arrive in a bundle from the post office.”

Jones convened remarkably few meetings; it was every manager for himself. Each in-house manager would be allotted a chunk of the partners’ capital and told the market exposure he could take on, and left to invest the money. At the end of the year, the managers who performed the best were the best compensated and given extra funds to manage; unsuccessful managers got less capital. Jones also required managers to have their own capital in the funds, sharpening incentives. He had essentially invented the modern pod shop.