Facing Off with the Terminator of Wall Street
The Economist magazine has found the new Master of the Universe and it’s… Arnold. Not in his role as the RINO former governor of California… but in all his Terminator glory. We’re just talking about a different type of machine…
You see, The Economist writes that we are witnessing the “rise of the financial machines,” which now initiate a majority of all stock market trades, and are “taking over Wall Street.” Every day, apparently, the machines trade so many securities back and forth — with humans none the wiser — that recording them would require the entire population of Manhattan working flat out on triple shifts. (Or something like that.) It ultimately adds up to roughly a gazillion bazillion trades a day, give or take, and that’s just counting by securities. The dollar figure is a lot bigger.
My first reaction, I admit, was that if the machines want Wall Street, they can have it.
These Machines Can Never Create Real Wealth
As a wellspring of wealth creation, Wall Street has been over for a long time. The machines are more a sign of the change, not really the source.
With that said, we can still make money — a lot of money — investing in stocks trading on the NYSE and elsewhere (like Hong Kong). But most of the billions of daily Wall Street trades are an economically barren exercise irrelevant to creating real wealth.
Even for those who do it profitably, the sums earned by most quantitative investors are tiny compared to wealth produced by the next Qualcomm, Apple, Google, Amazon, Tencent, or Ping An.
Besides, while the super-quants have upped the pace of trading, most of the activity is just the machines playing with other machines. If I build two computers and tell them to trade IBM back and forth all day, and I rev them up to a million trades a second, our volume will vastly exceed the volume traded by anyone who actually takes time to learn anything about IBM. Some such diligent researcher may put his knowledge to good use, and make a profit by shifting capital in a useful direction. Our two computers will never do that — except entirely by accident. An accident that would be reversed in the next round. Yes, the computers in the real market are doing a bit more than that. But not much more. At that speed, the computers must be trading mostly against some other quant’s computers.
Yet quants can make money, mainly because the stock market is not entirely efficient, and some of its inefficiencies show up as patterns in stock prices.
Building Wealth from Inefficiency
In the latter half of the last century, a group of quite brilliant economists sought to settle whether large liquid stock markets were efficient, in the sense that prices responded only to external information — including opinion — about the market, the company, the stock, rather than prices responding to prior prices in a predictable way. This was the hunt for price independence, which in turn would prove that the market was efficient. The economists studied millions, then billions of actual stock market trades to see if they could derive any valid, strictly quantitative trading rules that could beat the market. If they could — if prices had patterns driven in part by prior prices — then the market was not efficient.
The greatest of these economists and the great original advocate of efficiency was Eugene Fama. Fama was an honest man. In the end he could say only that the market is very close to efficient — but not quite. Momentum, for instance, was a real thing. Rising prices have a slight statistical tendency to rise for a few more trades, falling prices to continue downward. In other words, prices had a very slight tendency to chase their own tail. And these patterns could not be attributed to any event in the real world.
Once that was proved, it was game-on. As Moore’s Law supplied the power, the algorithms got bigger and faster and could “back-test” increasingly voluminous historical trades for increasingly subtle patterns. The patterns need not be valid under all circumstances or even very persistent. If the patterns worked for a period — and the periods got shorter and shorter — so be it. The algorithms would look for the next pattern and the cash register would keep ringing.
The pure quant traders work hard for a living, but it has little to do with investing. The same goes for most other functions performed by machines on the street, such as tracking and rebalancing Exchange Traded Funds (the mutual funds of the quant era, but assembled at vastly lower costs than the old human-managed funds).
Quants are statisticians. And as the great University of Chicago economist Frank Knight pointed out, statisticians perform the remarkable service of turning uncertainty (will lightning strike my house?) into “risk” — the statistical probability that someday lightning will strike somebody’s house, maybe mine. This makes insurance possible. Rationalizing uncertainty can reduce its impact from catastrophic to manageable by spreading the cost across time and helping you to get on the right side of the law of large numbers.
But here is what the statisticians cannot do: banish the lightning. For that we need an entrepreneur who as a result of his careful study of electricity is able to invent the lighting rod. (Thank you Benjamin Franklin) All economic progress is the result of replacing unpredictable hazards: drought, disease, famine, with, you guessed it, information about the world. We can insure every house against lightning but this does not banish the lightning. For that we need an entrepreneur, who, by the laborious study of electricity, invents the lightning rod., replacing one more hazard of life with — information. (Thank you Benjamin Franklin.)
This is the source of all economic progress. All our great entrepreneurs are hailed as risk takers. But they did not accept risk. They fought fiercely to eliminate risk by replacing it with information, the only source of wealth. This Wall Street quants cannot do.
I plan to dive into this topic more in upcoming Prophecies.
Editor, Gilder’s Daily Prophecy