Primary navigation:

QFINANCE Quick Links
QFINANCE Reference
Add the QFINANCE search widget to your website

Home > Blogs > Anthony Harrington > The HFT challenge to global equity markets

The HFT challenge to global equity markets

The HFT challenge to global equity markets Anthony Harrington

Two excellent recent commentaries on the High Frequency Trading (HFT) dilemma have shed new light on the problem that is perplexing securities regulators around the world. The fear is that ever faster algorithms and more powerful computers not only will distort, but are already distorting global equity markets. Larry Tabb, found and CEO of the market analysis house, TABB Group, has written an illuminating piece on how markets are being fragmented across multiple dimensions , while Pragma Systems, an electronic trading boutique specialising in algorithms, has produced a paper arguing that a good chunk of HFT profits are paid for by ordinary investors.

Tabb's article, entitled: "Fragmentation is redlining the markets", sees three dimensions along which markets can fragment. The first and most obvious is the proliferation of trading venues. When a seller wants to sell a security, whatever trading platform they are using will be programmed to check all possible trading venues available to it in order to find the best price to match the trade against. Tabb points out that the US alone has 13 recognized trading venues and 50 "dark pools" (institutional trading volumes that are unavailable to the public). Adding to the number of trading venues adds to the compute task for price matching.

Time and price are two other dimensions on which markets can fragment. Price fragmentation comes as trades speed up and quotes start adding decimals to the price. Similarly if the quoting frequency changes, as Tabb points out, "from seconds, to tenths, to tens of milliseconds, and then down to milliseconds," the quoting rate climbs proportionately. The more quotes that have to be checked the more the market fragments, tying up programming time simply trying to scan all possible quotes to find the best price at which to do a trade. If you multiply the fragmentation across multiple venues, multiple price points, and multiple fractions of a second, and then multiply again across 5,000 stocks, you have a massive technology and message infrastructure that you require simply in order to trade at "best practice", ie matching potential buyers and sellers at the best price.  As Tabb notes, what all this means is that the quote-to-trade ratios have gone through the roof, with vast numbers of quotes having to be scanned to produce a single trade.

Given the problems with the Flash Crash, Facebook and now Knight, has our market become too fragile? Has the technology surpassed our management capabilities? Have the regulators been left in the dust and investors on the sidelines? Increasingly, I believe that, yes, our fragmented infrastructure needs to be simplified. No matter how technology proficient the firm, these systemic problems hurt investor confidence and trust."

The Pragma Securities paper, "HFT and the Cost of Deep Liquidity" looks at the idea that HFT has a beneficial impact on markets because HFT trading provides massive liquidity to markets. This idea is based on the obvious truth that the more buyers and sellers there are in a market, the easier it is for everyone to buy and sell. HFT trading brings vast sums to both sides of the market so everyone benefits, or at least so a number of academic studies have argued. The Pragma authors argue that HFTs account for some 50% of market volume and that most HFT trading is "fundamentally a form of market making with very tight inventory control and very short-lived positions." As such, HFTs are beneficial rather than harmful. HFT's compete among themselves to provide liquidity (in that they are buying or selling, which narrows spreads and therefore reduces trading costs for ordinary traders.

The problem with evaluating whether HFTs are benign or malevolent is that, from first principles, it seems as if their profits must come at the expense of ordinary investors. If HFTs are pulling billions out of the market in trading profits this must be at the expense of investors, yet analysing the only empirical evidence suggests the opposite. In other words, it suggests that the liquidity provision that HFTs are giving to the market offsets the profits being taken.  It seems that by bringing in very large in aggregate, and very persistent orders the HFTs are "benign", the point being that even though each trade an HFT places may well be a micro trade, small in substance and fleeting in duration, they place so many of them that investors can always find buyers and sellers to take the other side of the trade.

As Pragma puts it, "High volume is associated with high liquidity, and liquidity by definition means cheap, easy trading." But the fly in the oitment is that Pragma's research shows that when you look at the highest volume stocks in the world, such as BACs or Microsoft, they are actually costlier to trade than their low volume peers. The reason, Pragma suggests, is the disproportionately long order queues with these stocks, which is a result of the huge number of orders competing to provide liquidity, compared with the rate at which aggressive orders arrive to take that liquidity. Technically this increases trading costs instead of decreasing trading costs and the difference amounts to several hundredths of a percent. That might not sound like much but aggregated over a year across the high volume stocks it amounts to billions of dollars a year in spread costs (the spread being the difference between the buying and the selling price).  Market regulators are going to have to look more closely at the "crowding" effect that is going on as a result of HFT's jostling each other to capture the bid/ask spread and to effectively stand between directional traders (real investors, institutions and individuals, who are taking a view that a stock will go up or down). No one can put the tech genie back in the bottle, but there may be ways to enable directional traders to deal more directly with each other (hence capturing the bid ask spread for themselves) instead of being forced to cede it to HFTs.  The Pragma research note is very thoughtful and provides an interesting basis for regulators and stock exchange authorities to contemplate.

Further reading on HFTs

Tags: Facebook , Flash Crash , HFT , market makers , regulators , systemic problems , trading venues
  • Bookmark and Share
  • Mail to a friend


or register to post your comments.

Back to QFINANCE Blogs

Share this page

  • Facebook
  • Twitter
  • LinkedIn
  • RSS
  • Bookmark and Share

Blog Contributors