High Frequency Trading: Wall Street's New Rent-Seeking Trick

Goldman Sachs Group Inc. (NYSE: GS) disclosed recently that it had 46 “$100 million trading days” in the second quarter of 2009. That was a record number, even for one of the biggest players on Wall Street.

When the U.S. economy is facing collapse and merger and acquisition volume is way down, it seems odd that investment banks like Goldman had record quarters.

Well, here’s the secret: They’ve found a new way to skim more of the cream off the top of U.S. economic activity. It’s called “High-Frequency Trading” (HFT).

High-frequency trading uses the speed of supercomputers to trade faster than a human trader ever could. Human owners of the supercomputers program them to take advantage of information milliseconds faster than other computers, and whole seconds faster than ordinary human traders.  This is not a minor development; HFTs now represent about 70% of the trading volume in the U.S. equity market.

HFT computer servers are able to beat other computers because they are located at the exchanges. They take crucial advantage of the finite speed of light and switching systems to front-run the market. They also gain information on orders and market movements more quickly than the market as a whole. They operate not only on the New York Stock Exchange (NYSE), but also on the electronic trading exchanges such as the NYSE hybrid market.

According to a paper “Toxic equity trading order flow on Wall Street” by the brokerage Themis Trading LLC, there are a number of different types of HFT. Liquidity rebate traders take advantage of volume rebates of about 0.25 cents per share offered by exchanges to brokers who post orders, providing liquidity to the market. When they spot a large order they fill parts of it, then re-offer the shares at the same price, collecting the exchange fee for providing liquidity to the market.

Predatory algorithmic traders take advantage of the institutional computers that chop up large orders into many small ones. They make the institutional trader that wants to buy bid up the price of shares by fooling its computer, placing small buy orders that they withdraw. Eventually the “predatory algo” shorts the stock at the higher price it has reached, making the institution pay up for its shares.

Automated market makers “ping” stocks to identify large reserve book orders by issuing an order very quickly, then withdrawing it. By doing this, they obtain information on a large buyer’s limits. They use this to buy shares elsewhere and on-sell them to the institution.

Program traders buy large numbers of stocks at the same time to fool institutional computers into triggering large orders. By doing this, they trigger sharp market moves.

Finally, flash traders expose an order to only one exchange. They execute it only if it can be carried out on that exchange without going through the “best price” procedure intended to give sellers on all exchanges a chance at best price execution. The Securities and Exchange Commission (SEC) has now promised to ban this technique, and flash trading on the Nasdaq will stop on September 1.

This toxic trading has caused volume to explode, especially in NYSE listed stocks. The number of quote changes has also exploded and short-term volatility has shot up. NYSE specialists now account for only around 25% of trading volume, instead of 80% as in the past.

The bottom line for us ordinary market participants is that insiders are using computers to game the system, extracting billions of dollars from the rest of the market. While it is illegal to trade on insider knowledge about company financials, these people are trading on insider knowledge about market order flow. That’s how Goldman Sachs and the other biggest houses make so much from trading. By doing so they are rent-seeking, not providing value to the market.

There are two ways to stop this: Ideally, the SEC will employ both. First, they can introduce a rule that all orders must be exposed for a full second. That will reduce the volume of HFT, but still doesn’t truly protect non-computerized outsiders.

The second, and better, solution is to introduce a small “Tobin tax” on all share transactions. It could be tiny; maybe 0.1 cents per share. (The SEC would also need to ban “exchange rebates” to traders.) Such a tax would make the worst HFT types unprofitable without imposing significant costs on retail investors.  It would also provide funds to help run the vast apparatus of regulation and control that seems to be necessary to run a modern financial system.

Goldman Sachs, and other financial institutions of its ilk, have imposed huge costs on the U.S. public with their “too big to fail” status. Now they are adding to the problem by scooping out money from the stock market through HFT. It’s about time the government imposed some taxes to stop the worst of these scams and recover the public some of its money.

 News and Related Story Links: