How Are the UK and Australia Able to Avoid HFT Pitfalls Ravaging the US Market Scene?
In one of my first blog posts for CFA Institute, I described problems that arise when high-frequency traders (HFTs) are able to use their speed advantage to effectively “front-run” some kinds of order flow. The basic idea is that in a fragmented equity market, HFTs can observe part of a block trade executing on one venue and then race (and beat) the other parts of the block trade to other venues, which moves the price against the investor. Some solutions to this problem have been proposed, including batch processing, transmission zoning, and of course, the time-delay of IEX.
The UK’s Financial Conduct Authority has just released a report about its attempts to determine whether HFTs are anticipating order flow across markets. Using proprietary, non-anonymous data collected via their monitoring and enforcement role, the broad result is that the authors did not find evidence that HFTs systematically observe marketable orders on venue A and then trade in advance on venue B.
The authors caution that although it is possible that HFTs simply do not pursue these kinds of strategies in the UK (US readers will likely raise a cynical eyebrow at this point), there are some market structure issues in the UK that may simply make it impossible to import these strategies from the US. First, there is no quote protection in the UK, which makes it harder for algorithms to predict where child orders will be routed in the immediate future after having observed one on a given exchange. In the US, orders (including unfilled portions) have to be routed to the venue with the best price, whereas in the UK, best execution obligations are more nuanced. Rule 611, which requires routing to venues quoting at the National Best Bid and Offer (NBBO), means that the US market is effectively more predictable on a micro-level compared with the UK market.
The authors also raise geographical issues. The UK market infrastructure is located in a concentrated space in the southeast of the country, so there may simply not be enough distance for high-speed microwave signals to outrun their proletarian fiber comrades (although attempts to ruin the Kent countryside in this manner are underway). However, it should be noted that a lot of market infrastructure in the US is located in New Jersey, which is also relatively compact and firms fight over which cupboard to be located in at the point-of-presence. So, it seems unlikely that geography plays a role. Even one microsecond is enough to be first in line.
In some ways, the results of this report mirror our findings on the quote roll phenomenon, which is when electronic market makers lean on the lit quote on dark venues (i.e., offer nominal price improvement to execute in the dark) before executing lit orders just before quotes roll down. This strategy is a way to ensure that lit orders are executed only when they are on the wrong side of the trade — that is, adverse selection. In that study, we found evidence for this behaviour in the US but not in the UK, and we theorized that the difference may be related to the lesser degree of market fragmentation and lack of quote protection in the UK.
It is difficult to argue that the UK has a “better” equity market given the lower spreads and higher liquidity in the US. But the less interconnected market structure and multifaceted best execution obligations (i.e., no trade-through protection) do seem to allow the UK (and also Australia) to avoid many of the controversies that have been ravaging the US market microstructure scene.
If you liked this post, consider subscribing to Market Integrity Insights.
Image Credit: iStockphoto.com: DavidLeshem