The way we trade today with technology is real progress, not the new boogeyman.
Now that the dust has settled from the release of “Flash Boys,” I wanted to add my own perspective to the high-frequency trading debate, from the standpoint of a former index arbitrageur and ETF market maker.
I started my career in index trading in 2001, and I spent most of it trying to deal with, and adapt to, the high-frequency robots. Anyone on a dealer desk during that period of time is intimately familiar with the technology arms race.
During that time period, I was a liquidity provider. As an index arbitrage trader, I would simultaneously bid and offer stocks and stock index futures to bring the futures back closer to fair value.
It was pretty primitive. I would be on the phone with the futures pit in Chicago, with my finger hovering over a mouse, ready to send a basket of stocks into the marketplace. The profit on these trades was small, often under $1,000. And it seemed like a lot of work and a lot of effort to make that money, generating a lot of trading activity in the process.
But I was performing an important function; two functions, actually. I was ensuring that securities were accurately priced, while providing liquidity to the marketplace. Liquidity, as we will see, is very important.
But over time, people began to program computers to do what I was doing by hand. These computers would bid or offer the e-mini futures and simultaneously bid or offer the basket or stock.
Actually, they were smarter than that. They could trade the stock algorithmically and they didn’t have to trade the whole basket of stocks. They could trade some now, and work into the rest later, with slight tracking error. These computers were programmed by people a lot smarter than me.
Soon the computers began to compete with me, and win, making futures trade closer and closer to fair value. And then they began to compete with each other.
Pretty soon, futures hardly moved away from fair value at all, and there was tons of liquidity in the index—both in the futures and the stock—all because of superfast robots conspiring to limit each other’s profits.
So it goes without saying that my career in index arbitrage didn’t last long, and in fact, by 2004, it was totally done. I was moved over to ETF trading, where I basically performed the same function. A customer wanted to sell 100,000 shares of XYZ ETF, so I would show him a bid, get hit, then sell a perfect basket of the stocks in that ETF to hedge.
Over time, I would redeem the shares of the ETF and get the stock in return, flattening my position. Once again: I was ensuring that securities were accurately priced, and I was providing liquidity to the marketplace. But you already know where this story is going.
People started to program computers to do the same thing, and sure enough, there they were—bidding and offering for ETFs. Over time, those bids and offers got bigger and bigger, and tighter and tighter.
For example, back when the iShares MSCI Emerging Markets (EEM | B-99) was in its infancy, the bid/offer spread was at least $1 for any size order. Today EEM’s spread is a penny wide, with hundreds of thousands of shares on the bid or offer.
Think about that. We, as investors, can gain exposure to an index of stocks listed in obscure exchanges all around the world—some open, some closed, all with different currencies and different settlement procedures. Just think of the operational hassle and that you can trade tens of millions of dollars practically for free.
All of this because of high-frequency trading.