So you’ve written and rigorously tested your forecasting strategy and built a risk framework around it. Now it’s time to put the theory into practice. What do you need to consider in order to effectively execute your strategy in the markets?
Trade execution at CTAs has come a long way in the past 20 years. Early in the period, all the business was done via the phone with a combination of desk brokers, bank traders and floor brokers. As volumes migrated off the floors and onto the screens, execution desks were presented with a new set of challenges as technology received greater and greater focus. Today, execution is predominantly either fully automated (strategies interact directly with the markets with no human intervention; also known as box to box) or follows a hybrid model where part of the flow is automated and part is manual.
Despite this evolution, fundamentally, executing your strategy has always been about three things:
- Controlling costs
- Reducing risk
- Limiting the visibility and impact of your trades, especially as your assets
With that in mind, here are eight areas of focus to help you in your pursuit of excellence in trade execution. Note: Although this guide spawned from our experience in executing futures, FX and cash equity strategies, hopefully its content has broader application too.
1) COUNTERPARTY MANAGEMENT
Execution counterparties (typically banks or brokers) facilitate a trade either by providing a route to the exchanges, by matching your business with another client or by taking on the risk of the trade themselves i.e. making a market. They are fundamental to getting your business done well and crucial partners in market dislocations. In order to build a strong counterparty line-up, you should consider the following:-
Counterparty Assessment Criteria – compile and document a key set of attributes and metrics that allow you to judge the performance of a bank or broker versus its peers. This applies to both choosing new counterparties and monitoring the existing selection. The criteria used may be different according to the asset class (e.g. commodities vs financials), the market type (e.g. OTC vs exchange traded) or the execution method (e.g. voice vs Direct Market Access). A side benefit of these criteria is their utility in demonstrating best execution to regulators or when the firm is the subject of operational due diligence.
Redundancy – quite simply, don’t rely on a single route to market. For each instrument traded, have a choice of counterparties and where feasible, a choice of channels i.e. voice, manual DMA and algorithmic DMA. Don’t neglect the importance of voice relationships - if the markets are in meltdown you want your line to be picked up first. Ensure the requirement for counterparty choice is balanced against the need to pay them meaningful revenue – there’s no point in having ten choices when they each receive negligible revenue.
Monitoring – keep track of the volume and revenue received by each counterparty and use the numbers to ensure business is being awarded according to their performance (see first bullet).
Commissions – don’t assume the cheapest is the best for your organisation.
Although implicit in the first point, it’s worth stressing that when moving into new markets, put the time in to find the counterparties who are specialists in that field.
Note: Some of the above points are equally applicable to clearers/prime brokers.
2) EXECUTION PERFORMANCE METRICS
In the CTA world, slippage is a universal term but it can vary in its interpretation from firm to firm. To help in the construction of a suite of metrics to assess your trade execution function, consider the following:-
Where does the signal generation stop and the trade execution start? In smaller funds that trade once a day per instrument this is straightforward, but for larger funds smearing their trading across the day, it isn’t so clear cut.
Include metrics that describe the quality of the execution i.e. incorporate opportunity available and the market conditions. For instance, the system generates a buy of 100 March corn when the market is trading 366 and you have the rest of the day to complete it. The trade is filled at 364; better than benchmark and on the face of it, great. What if the range in March corn after that trade was generated was a high of 367 down to a low of 357. That fill doesn’t look so great now. Looking at just the fill vs the trade generation price in isolation does not consider the opportunity afforded for that trade.
For implementation shortfall type metrics consider carefully whether the benchmark is the last traded price, the mid-price (of the bid/offer spread) or the weighted mid (incorporating displayed size).
If altering your execution methodology, use metrics to objectively assess new versus old.
Some metrics become meaningless if aggregated above instrument level e.g. by sector. Equally, some metrics are best suited to time series analysis (how is execution performance in an instrument vs history), rather than in comparing one market with another.
Create metrics for all trade types i.e. daily signal trades, re-gears and rolls. For slower models, rolls could constitute in excess of 50% of total turnover so it’s important to give them adequate focus.
3) EXECUTION TECHNOLOGY
Technology’s role in trade execution has progressively increased. We effectively have an arms race where you have to steadily increase sophistication just to stand still relative to your peers. With that in mind, it’s of paramount importance to know the technology available and what the current market standard is. As well as using it to reduce costs, control risk and limit visibility, also give attention to streamlining routine workflow; making it as low-touch as possible. Some elements in your technology arsenal could be:-
DMA Execution Environment
- Redundancy in exchange connectivity e. connect via more than one counterparty.
- Execution Algorithms – build in house and/or use counterparty supplied as Generally they fall into liquidity harvesting or price improvement categories. Select according to the market type and the current environment.
- Smart Order Routers – can be used in seeking the most cost-effective execution for multi-venue markets (e.g. FX, equities) and less commonly to obfuscate flows amongst
- Visualisation Tools – if you trade 150 markets algorithmically, you can’t physically watch Use these tools to quickly process events.
- Exception Management System (EMS) – in conjunction with visualisation tools, the EMS can be set up to flag events of interest in the workflow in real time or even interrupt an execution
- Circuit Breakers – no one wants to be on the front page of the FT after their execution algorithm behaves Think of the things that scare you and then limit their impact by introducing appropriate circuit breakers.
- Analytics Suite – ability to generate real time and historical reports regarding the quality of execution, counterparties, strategies, risk Where possible, these should be tailorable rather than flat reports.
Try and architect a trading chain that is modular allowing the removal of old elements and the introduction of new ones with relatively little disruption.
4) MARKET AND STRATEGY KNOWLEDGE
If the people executing the strategy are different to those that designed it, they should understand the models inside out to allow quick isolation of any erroneous trade generation. This is particularly important after new strategy releases. Additionally, effective execution demands a detailed knowledge of the markets traded. This includes:-
Fundamentals – what influences price e.g. which economic releases are crucial?
Who are the different types of market participants?
Intraday Dynamics – what are the intraday volatility and liquidity profiles? This may be difficult to isolate for the OTC markets.
Roll versus outright dynamics e.g. the Dax roll market behaves very differently to the outrights.
Correlations – know how markets relate i.e. will actions in one execution detrimentally affect another?
Contract Specifications for Futures
- Cash or physically
- Dates – distinguish between spot date, first notice date and last trading
- Position limits – reporting, spot, single month and any How does your strategy handle hitting these? Hard stop or allocate elsewhere?
Futures Curve Dynamics
- How many contract months are liquid enough to trade?
- When does liquidity typically shift to the next month?
- Is the market in contango or backwardation?
- How active are intra-market spreads?
Exchange Conventions (it pays to have a dialogue with the exchanges)
- Opening/closing price
- Fill algorithms g. FIFO, pro-rata or hybrid.
Which markets have a history of interventions or are currently pegged?
For OTC markets, which are the primary platforms?
Ready to find out more?
I hope you enjoyed/learned a lot from the points above. There are, in fact, 4 more steps you Must take, in pursuit of Excellence in Futures Execution, and they can be found in our FREE Ebook, which you can download right now.