Tradeworx CEO  Manoj Narang Defends HFT

Manoj Narang has a vigorous defense of HFT in this post. Obviously he’s biased, given where he works, but he also has some key data that appear to show increases in market volatility are biggest during after-hours trading when high-frequency traders are not active. It doesn’t automatically follow from that that volatility during the trading day is NOT due to HFT, but it’s a pretty good prima facie case. What do others think?

highfrequencytradingreview.com

This is a Guest post by Manoj Narang, Founder and CEO of Tradeworx Inc In recent months, a plethora of articles have remarked on elevated levels of market volatility, while simultaneously pointing out that computers generate an…

When discussing issues concerning HFT, there are really two things that come to mind for me, volatility and efficiency. First, HFT allows for faster incorporation of new information into prices. In other words, the shorter the timescale of the transactions the more informationally efficient the market becomes. So it can be argued that HFT produces greater market efficiency. With regard to volatility, there are good arguments for both sides, some say that there is increased liquidity provided by HFT which has some smoothing effect and there are others that have concerns that HFT algorithms can run amok and cause wild swings in volatility due to a lack of human oversight. The first argument does have merit, and the second has been seen in a different form from wild product pricing on Amazon due to unsupervised algorithms (seehttp://www.michaeleisen.org/blog/?p=358).

Personally, I am in favor of HFT and do not believe that what I consider “wild” volatility is solely the product of HFT gone awry.

—

Thanks for the response. I think you may be confusing a couple of concepts, or using words that I’d use differently. For example, what do you mean by “informationally efficient?” The HFT that I’m familiar with does not take in traditional market information.

In addition, I’m not sure it’s a good apples-to-apples comparison to talk about Amazon’s pricing algorithms. I see the analogy, but I’m not aware of market pricing problems caused by “unsupervised algorithms”. It may be (and really, always has been) the case that a large trade sent to market will affect the price. Part of the purpose of a well-designed algo is to eliminate or at least minimize the impact on the market price of the trade the algo itself is executing. There are certainly situations in which two or more strategies may interact in unexpected and negative ways, but the markets are generally open to all licensed or sponsored traders, whereas Amazon bears sole responsibility for the misbehavior of its pricing strategies.

I apologize for any misunderstanding, please allow me to clarify. When I
talk about informational efficiency, I am speaking of the speed with which
new information is reflected in the market price. Using this definition,
the shorter the time between information availability and price adjustment,
the smaller the opportunity to profit from having knowledge of that
information (whether human or machine has said knowledge). HFT comes into
the picture because a system is presumably based on some sort of trading
rules, otherwise all you have is an incredibly fast dartboard-style stock
trading machine. These systems inevitably react to market activity and
whether those movements are initiated by the HFT systems or human traders,
my conclusion is that HFT can assist in the rapid adjustment of prices. In
other words, no matter who gets the ball rolling, the price will adjust
faster with HFT presence. There is disagreement on this point, but there
isn’t any empirical data that currently exists which demonstrates a
negative effect on the price adjustment process by HFT (see page 27 of
report linked in next paragraph).

There have certainly been cases of erratic market behavior and pricing
problems due to HFT, as noted on page 11 of the report “Regulatory Issues
Raised by the Impact of Technological Changes on Market Integrity and
Efficiency” released by the Technical Committee of the International
Organization of Securities Commissions in July of this year (see
http://www.iosco.org/library/pubdocs/pdf/IOSCOPD354.pdf
<http://www.iosco.org/library/pubdocs/pdf/IOSCOPD354.pdf> ). In chapter 5
of the report, which begins on page 41, it is mentioned that “…whilst
algorithms and HFT technology have been used by market participants to
manage their trading and risk, their usage was also clearly a contributing
factor in the flash crash event of May 6, 2010.”

With regard to the Amazon example, I am not sure whether the algorithmic
pricing error was the fault of Amazon. The problem may have been related
to a third-party price management system used by sellers. There have been
such tools available to manipulate auction results on eBay for many years
so that kind of situation is not unusual. Whether the pricing problem was
the result of Amazon algorithms or third party software, it’s a relevant
case where automated mechanisms involving price or trades can have
unexpected results if appropriate safeguards are not in place.

I hope the above clarifies my earlier post and I apologize for any
confusion. If that didn’t help please let me know and I’ll try to explain
myself using some different examples or terminology.