Free Site Registration

Q&A with Tradebot's Dave Cummings

Traders Magazine Online News, June 3, 2010

Traders Magazine Staff

Days after the so-called flash crash on May 6, veteran high-frequency trader and exchange entrepreneur Dave Cummings sat down with Traders Magazine to discuss a bevy of issues facing the equities markets. Cummings started the pioneering high-frequency trading firm Tradebot Systems out of a spare bedroom in his house 11 years ago. He remains its owner and chairman. In 2005, Cummings launched the electronic communications network BATS Trading. Today it's the third-largest equities exchange. Here is part one of the interview.

Dave Cummings, Tradebot

Traders Magazine: So, does a prop trading firm ever have a day when it loses in the HFT game?
Dave Cummings: It depends on what your models are and how they work. Our goal is to create a very large number of very small trades, and to get statistical significance in a short window by being spread out. That's our way of doing it. Other people would take much more directional bets than we would. We're trading typically close to 3,000 stocks every day. And so, on any given stock we may make or lose money. But when you look at the whole basket, the good ones and the bad ones tend to average out. Most days we make money.

TM: They say that if you're right 55 percent of the time in trading, you can make money?
DC: I don't know what the exact number is, but yes. Do you have the edge, or not? I teach my traders: trade pretty small until you have your profit/loss chart going in the right direction. And when you like the shape of that, you start trading bigger. And when you get nervous, you start trading smaller. It's really varying the trade size according to whether you think you've got the edge, or not. And I mean the statistical edge (not the edge on every given trade).

TM: So, whatever formula you have and how the market reacts to the way you're reading it ...
DC: Yes, there are a lot of different models out there. And certain days, our models seem to be doing well (so we should be a major player in that stock). And then there are days when you struggle with a stock, and you'll need to back off a little bit. And some other algorithm is there to fill that gap. That's what makes a robust market: having thousands of different people playing for all kinds of different reasons and all kinds of different time frames and all kinds of different correlations and models. Markets are all about transferring risk. And the more people you can pass that risk around to, the more value is created by the whole system. And that sounds like markets 101, but people forget that. It's more than just a big poker game.
...
I think it's unproductive to label kinds of trades as morally better or worse than others. All investors and traders are in the market to make money. Some people do that in seconds and others do it in decades. But you set up a destructive process when you try to start to characterize trades; there are all these different nuances of why people do what they do. As long as they're making legal trades, I don't think you should try to second-guess the motives of a certain group of traders.
TM: Do you think that's going on right now?
DC: There's clearly a witch hunt. It's driven by the political process.
I think if you're talking about the events of [May 6], the market definitely needs better circuit breakers. There's a thoughtful discussion underway to get to better circuit breakers. I've always believed that any time you break trades that's really, really bad. It's bad for the markets and the public's confidence in the markets. The answer is not to go back and break the trades that somebody after the fact on some committee (without any transparency) decided these must be errors and these must be good. The answer is to build logic into the matching engines that don't let errant trades happen in the first place--limits, collars, whatever you call it (that are thoughtful, transparent, known by participants up front) rather than having messes like we had last Thursday.

For more information on related topics, visit the following channels:

Advertisement

Advertisement