Queue Position And Algo Slices: Like Taking Candy From A Child (Order)


An exclusive discussion with Haim Bodek, Managing Principal, Decimus Capital Markets. By Will Haskins, GlobalTrading

Every buy-side trader has been there. You see the bid, you try to hit it and it disappears in front of you. Haim Bodek explains how it’s done.

High frequency trading (HFT) strategies modelled on the Chicago Futures pits and enabled by bespoke order types allow firms using them to exit trades moving against them and rejoin the queue at the top to wait for favourable market conditions. And the key to the whole strategy? A ready supply of orders behind them in the queue – child orders created by the slicing algos used to execute many buy-side orders.

How does this work?
For this HFT strategy, knowing queue position is alpha. Firms using this strategy calculate their queue position in real-time, Bodek told a room of buy-side traders. Another key are hide-and-light order types, which allow an order to sit at the top of the queue in a market’s order book.

An HFT options trader and SEC whistleblower, and Managing Principal of Decimus Capital Markets, Bodek was following up on his recent keynote presentation at the 14th Asia Pacific Trading Summit, organized by the FIX Trading Community.

Against a rising bid, Bodek explained, these HFT firms’ high queue position allows them to execute first and capture the spread. When the bid is falling, their high queue position acts as risk management, allowing them to flip out – change from a liquidity provider to a liquidity taker – and exit the position against the order behind them in the queue, paying only the transaction fee. Such strategies are rarely economically viable except on trading venues where transaction fees are low or subsidised by liquidity rebates.

Haim BodekThe buy-side’s dilemma
Buy-side traders’ complaint is often that it looks like they are being front-run – as soon as they trade, the market moves away from them. More than one buy-side desk has shared evidence with regulators, but the HFT firms using this strategy typically escape punishment. From a regulatory perspective, the firms using these strategies are not front-running because they have simply changed from being long on the market to short.

In a market environment where the buy-side have increasingly sought to take more ownership of the execution process, Bodek believes trading desks at asset managers and hedge funds should consider what lengths they are willing to go to in search of a remedy.

One option is to avoid venues that provide supercharged order types catering to HFT. However, in the US, where the use of such order types is concentrated, orders have to go to the venue with the National Best Bid or Offer (NBBO). When a buy-side firm attempts to route an order to a preferred venue when the market is moving the risk of an inferior queue position is not the only concern. Often the price has already moved away and the preferred venue returns an error message to the buy-side trader’s desk indicating that this venue does not have the NBBO, according to the SIP.

The SIP or Securities Information Processor is an official public reference price, but it is just one option allowed by the SEC to meet best execution mandates. Firms are also allowed to recreate the NBBO from their own data feeds, which creates an information arbitrage between HFT firms aggregating and analysing low-latency data feeds and firms trading based on data in the SIP.

The other option for buy-sides tired of paying the spread to HFT firms is to trade at comparable speeds and levels of sophistication, for example, calculating queue position in real-time. For algos that slice large orders into smaller child orders to minimise market impact, child orders might utilise direct market connectivity enhanced by an FPGA layer and low latency network interfaces rather than executing through an EMS run over software.

The question is who pays for this expensive execution system? Effectively stripped of proprietary trading desks, most brokers would have to amortise the cost of development through higher execution fees for buy-side flow. This would also shift control of execution back to the brokers contrary to the recent trend.

Until buy-side firms are clear which option is the most cost effective and achieves the best execution outcome for their portfolio managers and end clients, a so-called HFT tax will continue to be levied on child slice orders.

We’d love to hear your feedback on this article. Please click here


recommend to friends
  • gplus
  • pinterest