Skip to main content

Keeping Pace With Pairs

Since equity pairs trading was first introduced to the market in the 1980s, it has become immensely popular amongst investment managers, and has evolved from its original incarnation of matching a long position with a short position in two correlated stocks (still the most common type of pairs trade) to encompass a much broader landscape of strategies. Pairs trading can now be conducted cross-region, cross-currency, cross-asset, and crossbroker, with strategies including not only stock components, but also futures/options, ETFs, baskets, bonds, FX, crypto/digital assets, and more.

However, despite the wide variety of pairs strategies now being traded, traditional PMS, OMS and EMS systems have been slow to adapt, particularly from an algo trading perspective. This means that many firms have had to adjust their pairs trading workflow to accommodate system deficiencies rather than the other way around, i.e. having the ability to configure their systems to match their preferred workflow.

Cross-market and cross-asset challenges

In today’s markets, firms are increasingly seeking alpha from pairs strategies that go across markets, often in different regions, and across asset classes, with instruments that might be denominated in different currencies. Whether it’s stock versus ETF, ETF versus basket, stock versus futures, or any other cross-asset pairs strategy, it is essential to be able execute each leg of the transaction simultaneously, and to hedge any currency exposure at the same time.

This can present a challenge to some firms, particularly if they run separate OMS & EMS systems for each asset class. Another hurdle they may face is if their brokers are also siloed, offering pairs trading algos that work for only one asset class (e.g. equities), but not for others. For instance, a futures focused tool and an equities-focused tool may not communicate well with each other. This  extends to strategies involving futures against ETFs, ETFs against baskets, any FX hedging component, and any strategy that involves pairs of pairs.

Also, from a global perspective, firms with international trading desks often look for cross-exchange arbitrage opportunities such as between US & Asia, or US and Europe. To trade such strategies, they need to be able to execute their pairs orders seamlessly on different exchanges, considering market closes and opens, as well as any pauses within the trading sessions.

Therefore, for any firm that is serious about pairs trading, it is essential to have the appropriate tools available, regardless of what their brokers provide.

Why broker neutrality is important

With many trading platforms, the pairs functionality links directly to a broker’s own algo trading engine. This can be restrictive if the firm needs to send one leg of a pairs trade to one broker and the other leg to another.

It’s important that the firm controls the pairs algo rather than the broker, meaning that the pairs trade can be executed through two separate counterparties, with each leg being routed individually as a single DMA order. As the broker only receives the DMA trade for that leg and might not even be aware that it’s part of a pairs trade, the firm retains much greater anonymity around its pairs strategy. Even if the firm is using the same counterparty for both legs of the pair, that counterparty only sees the individual legs, not the pairs trade itself.

This is important because it protects the firm from any information leakage around when the pairs trade is entered, how much size has been done, what the ratio is, and when it is unwound.

Broker neutrality means that firms can continue to trade with the specialist brokers of their choice and retain control and flexibility over how their pairs orders are routed to the market, rather than being constrained by the functional limitations of an EMS linked directly to a broker’s pairs algo.

Functional requirements

Key functionality in any pairs suite should include a comprehensive list of market-neutral strategies, such as classic spread and ratio (or a combination of both), rights arbitrage, merger arbitrage and daily percent return.

The OEMS of choice should also offer various execution styles to control the aggressiveness of each leg, the ability to manage volume participation at multiple price levels, along with a range of options around floating on one or both legs and managing slippage when one leg cannot be hedged within its limit price.

Traders shouldn’t have to constantly watch the screen. It should also be possible to automate what happens with a hung leg, by pre-defining aggression levels that determine whether to trade at market, to stay on the bid or offer, or to go to mid-tick, for example.

Linking up to all major prime broker stock borrowing desks allows traders to log on once to access all broker platforms across the market, enabling them to access the borrow and trade it directly without having to switch to different platforms and perform multiple logins.

Advanced TWAP and partial hedging

Another key requirement when using a TWAP algorithm to reduce the market impact of a pairs order, is to be able to define start and end times, and to fine tune parameters to send out orders at specific intervals, with pre-defined instructions on when and where to cross the spread.

For example, a trader may wish to input a passive catchup strategy in cases where the pair gets out of limit, inputting the catchup percentage for the unmarketable period (i.e. the period when the pair is out of limit).

To minimise slippage in such cases, traders should be able to drip feed a pairs trade into the market using its own TWAP algorithm. By using what we call “partial hedging”, the order can be sliced into continuous waves, to execute based upon the aggression level set within the algo.

However, a common issue when using TWAP algos for pairs trades is when markets are moving fast, causing spreads to widen. In such cases, the algo can be sat waiting for one of the waves to execute the lead leg and miss an opportunity if there is a sudden spike in the market. Partial hedging functionality avoids this situation by enabling the wave size to be automatically reduced in fast markets.

FX hedging

When trading cross currency FX pairs, firms either need to be able to feed in their own FX rate, use an FX rate that the client wants to use, or use automated FX conversion.

It is important to look for a provider that can support all three methods of FX hedging and use real-time FX rates. This provides a dynamic way to auto-hedge the FX component of a cross-currency pair, in effect adding a third leg to the pair, which can also be routed to a separate broker. Ultimately, this ensures that the pairs strategy is FX neutral throughout the whole trade.

Buy side firms have long understood the benefits of pairs trading, and pairs strategies will continue to provide a rich source of alpha for firms who can use them in new and unique ways. However, to do that, firms will need to have the right tools available – ones that give them the control, flexibility, and rich functionality they need to stay competitive a fast-evolving market.

They also need to be able to operate at scale without being restricted by the capacity of single broker engines. The solution is a scalable, cloud based, broker neutral platform that provides firms with everything they need to support their current and future pairs trading activities.

Chris Jenkins is Managing Director at Tora


© The Sortino Group Ltd

All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or scanning or otherwise, except under the terms of the Copyright, Designs and Patents Act 1988 or under the terms of a licence issued by the Copyright Licensing Agency or other Reprographic Rights Organisation, without the written permission of the publisher. For more information about reprints from AlphaWeek, click here.