Non-Equities Algos on the March

If an instrument trades electronically, it's only a matter of time before it will be traded algorithmically. Their fungible nature and listed market structure have made equities prime candidates for algorithmic trading. But now Wall Street has algo aspirations for futures, foreign exchange (FX) and, in some cases, even credit instruments. Sell-Side Technology sits down with Greg Wood, a vice president with Advanced Execution Services (AES) at Credit Suisse, to discuss how his firm is leveraging its algorithmic development across the various electronically traded asset classes.

SST: What is the status of algorithmic trading outside the equities markets? Are the non-equities algorithms being developed in lockstep with the equities markets or does each market’s micro-structure demand its own algorithm?
Greg Wood, Credit Suisse: It depends on the type of algorithm. There are algorithms that are used to generate alpha and those designed to trade efficiently to preserve alpha.

At Credit Suisse AES, we provide execution algorithms that can be used to trade various asset classes, such as futures, options and foreign exchange (FX), on top of the existing algorithms for the equities market.

All of these algorithms are developed in parallel. We have made a concerted effort to push the multi-asset space over the last three years. It's taken us out of our equities background and into those other asset classes.

Generally, the aims of execution algorithms are the same regardless of the financial instrument traded. Traders look to minimize impact of their orders, execute them efficiently without showing their hand, and avoid paying the spread wherever possible. The difference between the asset classes comes down to market structure and the subtleties on how those asset classes trade.

SST: After working in this space for three years, is it still a nascent market or have these algorithms matured?
Wood: We are seeing significant organic growth in non-equities algorithms driven by buy-side trading desk consolidation. Traders are doing far more these days. In the past, a buy-side equities trader would call across the room to a futures trader to hedge a position. Now, traders are looking for tools to let them trade futures, options and FX using the same concepts that they are used to for equities.

SST: Is the industry seeing the same algorithms implemented across the various asset classes?
Wood: There are similar algorithms being used across asset classes. Often, an algorithm doesn't regularly lend itself to another asset class mainly because of differences in market structure, but quite often they do translate across markets—particularly those for common benchmarks or styles of trading. One example would be implementation shortfall, which is used to minimize an order's impact on the market by referencing the price at the time the order is submitted. Traders want to get their executions as close as possible to that arrival price.

Scheduled algorithms like time-weighted average price (TWAP), where traders want to stretch out their executions over a period of time, also translate well across different markets. As do algorithms that minimize signaling risk.

Other algorithms do not necessarily translate, and you will find new types of algorithms that meet a need for a particular nuance of that asset class. Our options algorithms are a good example of where the tactics are different from those used for equities.

SST: To get the most benefit out of multi-asset-class algorithms, do firms need a silo-less trading infrastructure?
Wood: Sell-side firms that offer these services to clients need to have infrastructure in place that can be readily leveraged across the various asset classes. It also requires cooperation between the individual silos of the institutions to harness the technical and business knowledge of each unit to provide the best service possible.

SST: What type of capabilities are clients requesting from these algorithms in the next 12 to 24 months?
Wood: People are getting comfortable with the tools that we have available. As I mentioned, we've seen good organic growth in our multi-asset space for the past three years.

What people are saying now is that we should take it one step further. Rather than trading an asset class individually, how about we trade a couple of asset classes together—for example, an exchange-traded fund (ETF) versus a futures contract, or using futures as hedges for an options strategy? There is interest to bring these tools together across the asset classes and creating a true cross-asset product rather than a multi-asset product that handles each product individually.

Another request that we are seeing is to extend our product to help clients execute more complex trades. We've already done this in the options space where we've moved into multi-leg options as opposed to just single-leg options.

In the futures space, a very important part of trading is that the contracts have a finite life; they are designed to expire. People want to be able to roll the expiring contracts into the next maturity efficiently with minimal cost, and at the moment there are various manual techniques to accomplish that. Having become confident in using the algorithms to trade the individual futures contracts, clients are actively asking for a range of algorithms to do exactly the same for the futures rolls.

SST: Are you seeing client demand for one-click execution strategies or are clients asking for a tool kit so they can create their own trading strategies without having to wait for Credit Suisse to develop them?
Wood: People like to have a tool kit, so that's why we try to offer a variety of off-the-shelf algorithms for the asset classes we cover. We also provide customization on our side. We don't give people the ability to change the algos themselves, but instead a client comes to us and says that they like the performance of a certain algo, but they would like it to do this or that instead. We have a team of people who work on those requirements and tailor something that is specific to the client's requirements. We've been doing that for years in equities and now across the other asset classes.

A very important key to this is simplicity of use, so we never want to make anything too complicated. The more buttons you have to press to select the parameters to execute your order, the harder it becomes to use and the more likely that you can press the wrong button and have it execute in a way you don’t want it to. That's why we try to understand the client's individual requirements and adjust our tools to meet them.

SST: Where are you seeing the greatest adoption across the various asset classes? Is FX outstripping futures adoption?
Wood: There's a healthy growth across all asset classes. Each class is in its own phase of adoption and maturity. You find that algorithms in asset classes like FX have gained a considerable amount of traction because they are radically different from what has been previously available in that space. We will see the same in fixed income over the next few years. The futures market has a very different market structure since products trade on a single exchange, so there is a subtly different appreciation on how to use the algorithms. However, adoption still has a lot to do with education. As part of that education, we introduce concepts such as pre-trade and post-trade analysis outside of the equities space. These are important quantitative tools that help the adoption of execution algorithms since they allow the client to measure the benefits rather than just have blind faith.

SST: Does it take the same amount of time to develop this set of algorithms as a single-asset-class algorithm?
Wood: There will be a lot of analysis done before our algorithms are rolled out. We look at how they will behave in the market structure of a particular asset class. We test extensively before we release them to clients and closely monitor them when they're in production. The lesson we have learned over the last three years is that you cannot just take something from equities and expect it to work in another market; it needs to be groomed for success.