I read Mark Cuban's blog post the other day and was immediately struck by a couple things. First, I appreciate his candor, and his invitation for commentary and corrections. Second, he begins by clarifying that he is "going to be wrong in several ways." Indeed, he is. I wanted to provide some of my notes and thoughts on his post, and perhaps point some of his inaccuracies out.
I suppose before I begin, I too have a disclaimer: I have worked in the electronic trading industry as (mostly) a software engineer. My experience has been at a small equity options trading firm; I have never worked at a high frequency trading firm. I have performed several different jobs within the industry; at times I would classify myself as one of Mr. Cubans "algorithms writers." That's not all I did, but it was part of the job description[1]. I have worked at the firm for some time, but I am by no means the world's foremost expert in either electronic trading or high frequency trading. The firm had a direct relationship with only one of the several derivatives exchanges in the US (so I don't have very much knowledge of the cost/incentive programs offered throughout the many exchanges); we were by no means "high frequency" or truly "low latency"--those firms collected their pennies from our traders, too, more often than not.
1. Electronic trading is part of HFT, but not all electronic trading is high frequency trading.
Agreed, and this point is lost on a lot of people because it involves a fair bit of nuance. As an aside, I often feel as if there is this perception that electronic trading is all done by artificial-intelligence-driven robots and that algorithmic trading cannot possibly have any human element whatsoever. It's been my experience (in the equity options world, at least) that these black-box robots don't exist in most places, if at all substantially. More often, small parts of the trading workflow are automated, or algorithms take care of the minutia on the trader's behalf. Large positions are not meant to be acquired by artificial intelligence, but rather by traders' decisions.
The fact that there is a matching engine which runs on a server somewhere that everyone submits orders to is not a bad thing, it is in fact probably a good thing. Do you have any data to back up your attribution of the narrowing of the markets to the electronic age versus the high frequency trading phenomenon? If not, I think it is a sensible enough bit of guesswork, but if you do I think you should share!
2. Speed is not a problem.
You're right in that it doesn't create problems in the market, per say....but ultimately I think you're wrong. More technical minutia: there are two types of speed here. The first is the amount of time it takes the exchange to go from receiving my order to filling it with available liquidity. A decrease in this time is a good thing and seems harmless enough to me. I think this is the speed you are talking about.
The second kind of speed is trading speed. This is the typical time a share is held; as it decreases, the rate of the marketplace increases. This contributes to the fragilization of the market and exposes us to more of the systematic risk you mention below. Some of my thoughts on HFT and it's systematic risks can be found here.
3. There has always been a delta in speed of trading.
Agreed again.
4. So what has changed ? What is the problem?
Couple points here. I think the main point you are making is valid: the real issue is the time people are holding positions in major equities. The marketplace has accelerated and as a result is more fragile. Also, not all algorithms are so bad!
However, I take a bit of an exception to your definition of a rigged market place. Just because someone is arbitraging something in the market doesn't make that market rigged. Just because there is a way to continually arbitrage things in the market, doesn't make it rigged either. For instance, I am sure there are somewhat similar instances in the equity options world--some instances of traders being able to execute several trades at once or in rapid succession to lock in a guaranteed profit...I would not call that rigging the game, just taking advantage of an opportunity you spotted. The fact that you can only spot that opportunity if you make (semi-insane) investments in technological and intellectual capital doesn't bother me, either. Somebody is going to make an investment to take advantage of arbitragable scenarios, and those people are going to be professional traders, not everyday investors.
Also, I don't know the exchange-by-exchange particulars here, but this notion of "jumping in front of the line" seems really flawed to me. Do you have some reference material of an exchange which allows some orders to take precedent over others simply because those participants have a (pricey) appointment with the exchange? It was always my understanding, and it has been my experience, that it is generally first come first serve for incoming orders, and any preferential treatment is given to quotes. Preferential treatment of some market makers relative to other market makers doesn't seem to be too big of a deal if their preference applies only to quotes. If they can all of a sudden take precedence in the order flow, then that seems a bit whacky. I'd love to learn more.
6. Is this bad for individual investors?
a. Either the profit HFT makes will float back into wider spreads, or electronic traders will help fill the gap. I would hazard that HFT does narrow spreads, thus the profit they are scalping off you may not be as big of a cost as one would think, but again, I don't have any data to back up that guess.
b. I think the blanket rule that HFT don't touch small stocks may be a bit flawed, too. What do you mean by small? Small daily volume? small stock price? I am sure some HFT trades happen on smaller dollar stocks so long as volume is there. They may even happen on smaller volume stocks, too. Who knows, it is all pretty opaque.
c. Given my understanding of HFT, it does not seem unethical to me. There is a big externality to their activity in terms of systematic risks, though. Those issues are not theirs to solve; they are up to us to solve.
7. Are There Systemic Risks That Result From All of This?
Agree with your conclusion, but I provide a slightly different look at where those systematic risks are coming from here.
8. So Why are some of the Big Banks and Funds not screaming bloody murder?
This analysis makes a lot of sense to me, but I have literally no experience at or with big banks, so I can't really provide much insight.
9. So My Conclusion?
I completely agree: you can craft arguments which paint HFT as spread-narrowers, however this misses the point! Maybe they narrow spreads, maybe they don't; in either case they are DEFINITELY fragilizing our capital markets via creating unquantifiable systematic risks.
- As is often the case at smaller firms, I performed a bunch of different jobs at once. I was mostly a software engineer, but I also wrote trading algorithms, option pricing mechanisms, provided systems support, handled most of the technical communication with the exchanges, analyzed trade data, etc.