Benefits
- It's fast. Obviously, this goes without saying. HFT takes out human error, time delays, and anything else to complete the one task it's assigned: trade stocks.
- It increases liquidity. With so many trades being made per second and algorithms trying to constantly figure out future prices, HFT puts a more accurate value on each stock with every trade.
- It cuts costs. HFT helps investors by reducing the costs necessary to make a transaction.
- Computers can crash. The complex computer algorithms that make up most of HFT are susceptible to computer errors or sudden changes in the market; numerous examples have occurred in the past few years and I will talk about them in the coming weeks.
- It can have far-fledged effects. One author calls it the "butterfly effect" - a huge storm in the U.S. can be caused by a butterfly flapping its wings in Asia. HFT makes all markets more connected and therefore more respondent to events happening around the world.
- There's no safety net. In a number of recent reports on HFT algorithms, industry pros revealed that profits were more important than safety; often times firms would fix bad algorithms just by "tweaking old code" and then reimplementing the algorithm into the system.
Of course, there's many different kinds of HFT and the algorithms that make it up. This list is in no way a complete, closed book on high frequency trading. With that being said, what do you think? Is the evolution of HFT a good thing for Wall Street and the financial world, or not? Is it profitable only for investment bankers, not us everyday consumers or occasional individual investors?
I think it's safe to say that for now, at least, we'll have to live with HFT. Unless another market crash or significant government regulation is near, these algorithms will always be a part of Wall Street. Pretty soon, we could even see algorithms that can process trades in microseconds. It's getting faster than ever before. Does the machine ever stop?
Sources:
[1] "Are Computers Bringing Down The Stock Market?", Forbes Magazine, August 15, 2011, http://www.forbes.com/sites/investopedia/2011/08/15/are-computers-bringing-down-the-stock-market/2/
[2] Jeff Cox, "High-Frequency Trading: It's Worse Than You Thought", CNBC, September 20, 2012, http://www.cnbc.com/id/49102808/High_Frequency_Trading_It_s_Worse_Than_You_Thought
[3] Fabrizio Goria, "Easy Money", The European Magazine, October 15, 2012, http://www.theeuropean-magazine.com/861-goria-fabrizio/862-high-frequency-trading
[4] Bruno J. Navarro, "Vanguard CIO: High-Frequency Trading Cuts Costs", CNBC, October 18, 2012, http://www.cnbc.com/id/49434073/Vanguard_CIO_High_Frequency_Trading_Cuts_Costs
[5] "The Fast and the Furious", The Economist, February 25, 2012, http://www.economist.com/node/21547988
I'd like to address some of your benefits and disadvantages listed.
ReplyDeleteFirst of all, being fast isn't necessarily an advantage for the market. It's better for a trader seeking to exploit others' algorithms, but what is fast is not superior to more deliberate purchasing of stocks.
Second of all, you claim that HFT increases frequency of trades and therefore more accurately represents the value of the company. This might be true if HFT were based off evaluating the value of a company and subsequently purchasing stocks. What HFT actually does is try to predict how the market, specifically other HFTer's predict how the market will behave in a loop where stock value becomes increasingly distant from reality. Even under a conventional paradigm, volume does not necessarily correlate with an accurate valuation of a company. I also don't quite see the connection between an accurate stock value and liquidity. More trades make a stock more volatile and actually decrease liquidity.
Finally, I think you should eventually analyze the type of risks taken by a computer versus those taken by a human. It might be interesting to see how much more or less risky HFT is when weighing its consequences.
Something you should definitely comment about is the impact of high frequency trading on arbitrage. Arbitrage trading used to be a huge component of the market, but nowadays with the millisecond buys and sells of algorithms, traditional arbitrageurs have no inefficiencies to capitalize on. What's your thought on arbitrage becoming pretty much a completely electronic thing now?
ReplyDeleteGood overview of the benefits and drawbacks of high frequency trading.
ReplyDeleteRather than saying it eliminates human error, I might say that it shifts human error from the traders to the programmers.
As you point out, “Everyone uses them - from small hedge funds to the big investment banks like Goldman Sachs.” This of course means everyone who is a big company with lots of complex computer systems. Where does that leave the individual trader (or is that a dying breed, and we’ll all just have accounts managed by the big guys?)
ReplyDeleteYou promise to offer some examples of computerized trading glitches. Those will be a great addition to the blog (along with analysis). I like the references you have for this post.
It might be interesting to take an even broader perspective on the nature of highly-coupled transhumanly-fast systems, which go beyond just stock trading. Starting long ago people have argued that getting dependent on these systems opens us up to unpredictable disasters in areas such as military systems (missile defense was the big example at the time), power and transportation, etc.
Good work. I’m enjoying reading this.