High Speed Risk

Is the era of high-frequency trading over? 

While you ponder whether “High Speed Risk” might be a good name for your garage rock band, let’s reflect on stocks. We said last week: “Our Sentiment is negative for the first time since the election. It’s a weather forecast.  No need for panic, only preparation.”

We measure the short-term movement of money with a 10-point scale. It was about 5.0 or higher from the election until Mar 9, 80 trading days. Last week it dipped below 4.5.

And weather arrived yesterday before today’s VIX expirations. It’s not news about the Trump administration.  It’s the end of a long, leveraged run. Monthly options and futures expired Friday the 17th.  New options traded Monday, Mar 20.

Yesterday was what we call Counterparty Tuesday. If counterparties have estimated demand Monday for new options incorrectly, they true up on Tuesday. Since markets fell, counterparties overshot demand.  

Derivatives have featured prominently in gains since the election. Investors have been buying both stocks and rights to more of them in the future. That additional implied future demand breeds higher current stock prices.

For the first time since the election, investors didn’t buy more future rights.  Does this mark an end to that pattern?  Certainly for the moment.  And it dovetails with the state of high-frequency trading.

For you new readers, let’s canvass high-frequency trading.  In 2007 after Regulation National Market System, a firm calling itself Octeg splashed through the data. In Intel alone, Octeg was driving 35% of monthly volume, crushing Goldman Sachs.

Who is Octeg, we wondered? The firm defied what we knew about brokers, which always wanted to hang a sign out, advertise that they had products for sale. We couldn’t find even a phone number for Octeg.  It was like stumbling on an unmarked warehouse in the suburbs packed to the ceiling with all the stuff you tried to buy at the mall.

While rooting through regulatory filings we found an address in Chicago and then another firm in the same suite called Global Electronic Trading Co (GETCO). 

And then we got it.  Octeg was GETCO spelled backward. The two were the same firm.

Getco dominated trading through the financial crisis, profiting on two ideas. First, exchanges began paying traders to sell shares on their markets. Think of it like a store coupon: Do business with us and we’ll give you a discount. Getco cashed coupons. In gargantuan manner. Exchanges paid them in coupons for relentless volume.

Second, Getco realized that it could be first to set price. So why not set as many prices as possible, forcing big institutions to chop their trades into smaller pieces?

Volume exploded. 

But it wasn’t investment.  Getco had no customers. It was using computers and mathematical calculations to continuously set prices in the stock market, getting paid to buy and sell stocks while simultaneously changing the price ever faster to force big investors into chopping up stock orders into smaller pieces so Getco and its burgeoning ilk could sit in the middle buying low and selling high in fractions of seconds.

At the pinnacle in 2009, we pegged this behavior, high frequency trading, at 70% of volume. Now high-frequency trading by our measures is less than 40% of volume.

The entire market the past decade is built on it. On the floor of the NYSE, four big high-speed firms price all NYSE stocks at the open. At the Nasdaq, a larger number does the same, trading prices for coupons.

The problem is high-frequency traders don’t have customers. They aren’t “working orders” for investors. They are buying low and selling high in fleeting fashion, for profit. Mistake these prices for ones from investors, and you mentally misprice stocks.   

You read that high-frequency traders are “market makers.” They’re “furnishing liquidity.” Traders with no customers can’t make markets. They can only exploit what others in the market don’t know. In 2007, it was easy. Now it’s not.

That’s because big stock brokers are doing the same thing with Exchange Traded Funds, rapidly repricing them, and index funds, and the stocks comprising them, and the options and futures derived from them. The big brokers are better at it than high-frequency traders because they have customers and can make longer directional plays by reading what customers are doing.

In a market without high-frequency trading, all stocks would trade like Berkshire Hathaway Class A shares.  About 400 shares daily.  It would be better for investors. But all the exchanges would go broke. Ironic, isn’t it?

High-frequency trading isn’t done. But with the market we’ve got, the harder it is for high-speed machines to price stocks, the greater the risk of big moves.