In 2004 I enjoyed my 15 minutes of fame for an article I wrote called The Tao of Alpha, in which I explained the concept of alpha as a zero-sum game. Sources of alpha in 2004 were much different than those available in the mid-1990s when I started my career, and they are also different from today’s. Alpha is highly transient and has been coming and going for as long as capital markets have existed. 

The term “alpha” entered the vocabulary of Wall Street sometime in the 1980s, gaining widespread use in the 1990s with the proliferation of hedge funds. It reached cliché status around 2000 by which time everyone was pontificating about it, with many having no idea what they were talking about. 

An alpha source is a circumstance that affords an investor the opportunity to take non-market risks with the promise of positive returns. For example, if you can continually get some actionable information before the market does (legally or otherwise), you would have an alpha source. 

There are innumerable alpha sources in the world, but none ever persist. This follows from the inevitability of diffusion; as soon as enough people have access to a particular alpha source its value disintegrates. Think of some information advantage: its potency is inversely proportional to the number of investors who can access it. Once enough people have the information, it degrades into an advantage for no one. 

Over the past hundred and fifty years, many alpha sources have come and gone. This article reviews some of the most potent sources available to investors since the late 1800s. 

Early Alpha Sources 

For the decades and indeed centuries before the 1930s, the single most potent alpha source available to investors was information, whether real or manufactured. All the legends of the Gilded Age, from Jay Gould and Cornelius Vanderbilt to J.P. Morgan and Andrew Mellon, leveraged their ability to obtain what today would be labeled “material nonpublic information” from both industry and government sources. 

Gould (a.k.a. The Mephistopheles of Wall Street) decided in 1869 that he wanted to corner the gold market but to do so, he needed to know of the Grant administration’s plans to sell their own reserves. So he paid his bureaucrat brother-in-law, Abel Corbin, $1.5 million to place a mole inside the U.S. government. Corbin used his political influence to install General Daniel Butterfield as U.S. sub-treasurer and he, in turn, was paid to provide advance notice of any government gold sales. This information edge, obtained in what would today be an egregious violation of several federal laws, helped Gould profit liberally and even partially caused the Black Friday of September 24, 1869. 

Those less connected still achieved information advantages via the questionable — but at the time legal — practice of manufacturing disinformation. 

Trade Circulars from the 1860s record all kinds of episodes of deception. For example, in 1865 several members of the Chicago Board of Trade hired French actors to visit the exchange and inquire about a winter’s worth of wheat for the army of Napoleon III. Speculators rushed to buy wheat futures and the schemers short sold them for vast profits. Napoleon’s emissaries, of course, vanished within hours. 

What could easily be labeled the golden age of insider trading, which had lasted in America for a century and half, faded away in the first decades of the 20th century. Congressional acts in 1922 (Grain Futures Act), 1934 (Securities Exchange Act) and 1936 (Commodity Exchange Act) drastically curtailed the mechanisms by which an investor could gain an information advantage. 

By the end of the 1930s, the SEC and CFTC stood directly between investors and any hope of the information advantages they could once procure. Information as an alpha source was, for all intents, dead. (As it turned out, it was actually only dormant, albeit for 80 years or so. Keep reading.) 

A.W. Jones and the Birth of the Hedge Fund 

Other than over a long horizon, the return on particular stock is primarily dictated by the overall direction of the market. This is universally understood among professionals. 

By virtue of the variety of derivatives available to investors today, it is easy to take no market exposure and instead focus on precisely the risk one wants to take. But in the 1940s, this was not the case; investors had to accept that buying even one stock meant taking overall market risk. 

For this reason, conventional wisdom was that market timing was essential to successful investing; no matter what stock you wanted to buy, you best buy it when the market is trending up. The idea of being either long or short while remaining ever cognizant of your timing was the foundation of a good investment strategy. 

Alfred Winslow Jones would destroy this notion with a simple innovation: his portfolios contained both long positions and short positions at the same time. Furthermore, he would use leverage on both sides. Against the backdrop of conventional wisdom, this seemed capricious if not insane. 

But of course what Jones was doing was ingenious because it solved the market timing problem. He found a way to build a portfolio whose returns would be primarily driven not by the direction of the market, but rather by the performance of the stocks he was exposed to, relative to each other. 

By hedging out market risk and adding leverage, Jones could create material exposure to secondary risk factors. Thus he could be rewarded for being right about stock-specific factors regardless of the market’s direction. He had in fact, unlocked a new alpha source. 

The analysis of company specifics offered alpha for the simple reason that few people studied them as intensely as macro factors. Year after year, Jones could exploit relative value opportunities for perpetual profit. 

Jones offered colleagues access to this strategy via an LLC partnership he created in 1949. He took some fees for managing the fund and a percentage of its returns. The fund, being hedged against market risk, became known as a “hedge fund”. As it turned out, it would be the first of 10,000 and counting. 

Non-Market Risk Factors 

Today we know there are dozens of secondary factors moving stocks. But Jones and a handful of copycats would enjoy that edge for over 15 years before the rest of the world caught on. Carol Loomis’ 1966 article, “The Jones Nobody Keeps Up With”, is often cited as the major reveal of Jones’ strategy but it came out 17 years after he formed his hedge fund. 

In the meantime, the 1950s and 60s brought about significant advancements in financial theory. Starting with Harry Markovitz; continuing with Bill Sharpe, Barr Rosenberg, Michael Jensen; culminating in Eugene Fama and Ken French, academics gradually illuminated all the secondary factors that were truly moving markets. Much as the periodic table was filled in one element at a time, the set of factors that drove equity markets were “discovered” over the next 20 years. 

And thus Jones’ edge became diffused. Hedge funds today will not get credit for taking exposure to non-market risk factors because they are too well understood. Professional allocators will not pay hedge fund fees for the execution of strategies that are on the first year curriculum of any Masters of Finance program. 

Mathematics and Technology  

By 1985 there were still less than 100 hedge funds in existence, thanks to a bear market in the 70s that killed many funds who had yielded to the temptation of a long bias. But by 1990 that number had risen again to 500 and by 2000 there were about 3800 hedge funds globally. 

The alpha sources of this period could usually be connected to one of two themes: mathematics or technology. For example, the mere use of computers to price bonds in the 1980s afforded several prop desks a delightful advantage as their IBM PCs priced all bonds in the market every few seconds while their contemporaries computed one yield at a time on their HP financial calculators. 

Of course advantages like using computers could not and did not persist as an alpha source for long. But the clever use of processing power did. For 10 years or so, computer science powered the golden age of fixed income arbitrage. 

The rise of interest rate derivatives in the 1980s and 1990s, specifically bond futures, libor futures (aka Eurodollar futures), swaps, swaptions, caps and floors, created exploitable complexity. These derivatives were complicated but they all fundamentally depended on only two factors: libor rates and government bond yields. So in theory, they had to be related to each other in some way. Cue the mathematicians and computer scientists. 

Here’s one example: Bond futures stipulate that, on expiry, the seller must select one bond from a set of eligible issues for delivery. What this really means is that a put option is embedded in every bond future. That option has value. But the market had always completely discounted it. If you gave a smart quant a PC in the early 1990s, she could run some Monte Carlo simulations and not only value this embedded option, but craft a long position in the future hedged with a delicate balance of deliverable bonds that was guaranteed to be profitable. This was called “cheapest to deliver” or CTD arbitrage. 

Convertible bonds — where the holder could exchange the bond for shares in the company — offered a similar opportunity during the 1990s because again the market simply ignored what was an embedded call option. CBs traded too cheaply. An investor who could speak Black-Scholes and had a computer could simply buy these bonds, short the issuer’s stocks and gamma trade to a risk-free profit. 

The common theme across these and numerous other examples was that highly skilled mathematicians leveraged the power of stochastic calculus and microprocessors to trade against men who were, literally, ignorant. The analogy of a bunch of scientists showing up with explosives on a medieval battlefield would not be inappropriate here. Between 1990 and 2000, nerds usurped control of the bond market from the megalomaniacal Luddites who had reigned for thirty years. 

The Fall of the Quants 

Again showing the transience of alpha, by the start of the 21st century the low-hanging fruit for Wall Street mathematicians was gone. But the buy side was still armed with quants, and hedge funds now numbered over 4000. With hedge funds responsible for the majority of global trading in several asset classes, the sell side realized there was a market for complexity. 

While there were many factors that led to the rise of the CDO in the 2000s, one should not discount the importance of the appetite quantitative hedge funds had for these instruments. In fact, without quant funds eager to “arbitrage” the lower tranches, it is not clear if these derivatives would have thrived as they did. 

As it was, quants relished the challenge of CDOs, devising multivariate probability distributions to bet on the correlation between tranches. Like the generation before them, they were all set to profit from their deep understanding of these derivatives. 

An Icarus/Daedalus analogy is irresistible here. Where the first generation of quants came to Wall Street bringing sober and meticulous mathematics to their task, this second generation arrived with mostly hubris. 

Secure in their knowledge that inter-tranche correlation could never approach one, quantitative hedge funds took massive, leveraged positions in CDOs. It turns out math is infallible only when the underlying assumptions are right. There was no difference between the any of the mortgages that backed these CDOs. The default correlations were, in fact, one. 

The irony here is that this was visible from day one; not through the prism of Gaussian Copulas but rather through the long-forgotten craft of fundamental analysis and due diligence. Other than high frequency trading, the alpha extractable from technology and mathematics was never really there after the end of the 20th century. It took a decade and a market crash for investors to learn this alpha source was mostly tapped out. 

Alternative Data and the Return of Information Advantages 

Today there are over 10,000 hedge funds; still armed with quants and still seeking elusive alpha sources. There are two global phenomena that are powering a return of information as a potent alpha source. The first is the Internet of things. We are marching towards a regime where every device powered by electricity will log what it is doing. The implication is that we can take measures and aggregations of the state of just about everything humans do. 

Take modern farming as just one example: via sensors in tractors, the soil quality of every field in the Americas will soon be known in real time. Take this data, add weather information and you can infer crop yields to power your futures trading strategy. 

Even more potent is the ascendancy of the data-driven company. Tens of thousands of companies are now data-centric, measuring everything that is going on around them. Every one of them has their finger on the pulse of some part of the global economy. 

The data to measure the economy accurately, in real time, exists right now. As investors gain access to it, they will unlock all kinds of new alpha sources. Companies like Neudata and 7Park saw this opportunity early and generate powerful analysis and reports powered by this new “alternative” data. 

Quandl (now Nasdaq Data Link) too, is at the forefront of this trend. As a marketplace we naturally attract suppliers. But recently the most interesting supply side inquires we get are from companies that have never had anything to do with Wall Street. They bring us virgin data that, when mined properly, reveals nuggets of alpha buried within. 

Alpha sources are transient. The era of alpha from microprocessors and mathematicians is ending. But the era of alpha from pure information advantage is just beginning. Data Scientists are about to become Wall Street’s new best friend.