What happens when you take humans completely out of the market?
By C. Scott Garliss
Anything a human can do, a machine can do better. Anything a human can screw up, a machine can screw up worse.
If you’re using machines in the stock market, that means algorithms. Algorithm-based investing and trading is an increasing piece of the market.
And it means that market swings are going to get worse, not better.
It’s simple economies of scale and power. Any time human effort is multiplied, the good and bad effects are multiplied as well. The effect becomes progressively greater and potentially worse when humans are removed from the equation and no longer understand either the technology or principles involved.
That last part is key. After all, both mutual funds and exchange-traded funds (ETFs), consist of capital gathered by investors that allow fund managers to invest vast quantities of money. In both cases, investors are no longer buying shares in a specific company, but shares in an idea. Then the fund proceeds to buy assets that meet the criteria of that idea.
But even though mutual funds and ETFs move the investor a step away from direct ownership, there’s still transparency and human oversight. The investor knows the criteria that the fund or ETF is using to buy assets and can see what assets were bought and why. Should a fund or ETF go belly up, it’s relatively easy to sort out the wreckage. While funds and ETFs hold enormous sway in the markets, the transparency and self-imposed limits make them a wise use of economies of scale.
In the chart below, we’ve marked the growth of these exchange-traded products (ETPs)…
The real problems come when you have zero transparency and zero humans.
One of the major causes of the financial crisis were the derivatives which Wall Street brokers and insurance companies were selling to investment managers and corporations as ways to play investment ideas and “hedge themselves.”
But these derivatives were so complex no one fully understood what they were owning. And the hedges didn’t work like they expected. Many of the derivatives were baskets of different ideas lumped together. And unlike ETFs, they contained far more exotic products than garden variety stocks or bonds. Because these exotic components were blended together, when one part (subprime mortgages) went bad, the whole thing had to be thrown out.
Imagine if there was a nationwide recall of flour… If you had separated and labeled all your ingredients in your kitchen, all you have to do is throw out your bag of flour. But if you’ve already baked a cake, you must throw out the whole cake.
Because ownership of derivatives was vast, everyone was trying to unload them at the same time, making the financial crisis worse in their quest for liquidity.
The next crisis will add algorithms and programmatic trading into the mix, with vast sums of pooled investor money and less-than-ideal transparency.
Increasingly, algorithms are a bigger part of the market, both in asset allocation programs and trading execution. For example, commodity trading advisors (CTAs), which are essentially algo-driven funds, have seen assets swell by 36% to $360 billion over the past 10 years.
Without a human with a hand on the throttle or brake, these programs can intensify the havoc on days with big moves up or down.
CTAs are to market swings what warmer waters are to hurricane formations. Stocks go up, the models say buy, stocks go down, the models say sell – in either case intensifying the move. Designed for small moves, without a human with a hand on the throttle or brake, these programs can intensify the havoc on days with big moves up or down.
So, for now, we benefit as CTAs propel stocks up further than they would have otherwise… But as rising ocean temperatures are increasing the intensity of the hurricanes we’re seeing, we can expect more frequent and more intense volatility as the effects of CTAs and other momentum investors continue to exercise outsized effects on the markets with little oversight or transparency.
That worries me. It should worry you, too.
After all, before any of this force-multiplying technology came into being, one of the most damaging things ever done in the markets was caused by a single person in the 1990s. Nick Leeson, a Singapore-based trader, attempted to cover prior losses by making bigger and bigger unauthorized trades. It didn’t work. Instead, his losses ballooned enough to cause the insolvency of Barings Bank, the oldest U.K. merchant bank.
Then consider that the second-largest one-day swing between the high and the low in the Dow Jones Index was the so-called “flash crash” on May 6, 2010. In less than 45 minutes, algorithmic trading caused the Dow Jones Industrial average to drop a total of 998.5 points at the lows before snapping back.
Imagine what someone like Nick Leeson could do with an algorithm.
C. Scott Garliss is one of Wall Street’s most connected analysts. He spent 20 years trading for some of the top investment banks in the country – including First Union Securities, Wachovia Securities, Stifel Nicolaus, and FBR Capital Markets – for clients like SAC, Viking Global, Discovery Capital, UBS O’Connor, T Rowe Price, and Fidelity. Today at Stansberry NewsWire, he uses his expertise to reveal what’s happening behind-the-scenes on Wall Street so you can profit on Main Street.