Insights

FlexAlgoWheel’s Best Kept Secrets

August 8, 2022 | By: FlexTrade Insights

By Sharat Kumar

Every institutional trader knows the term Algo Wheel but may not be aware of the advances made by FlexAlgoWheel which go far beyond what the term Algo Wheel implies. Over the past five years, FlexAlgoWheel has become an essential decision support and workflow automation module used by a rapidly growing number of FlexTrade EMS and OEMS users on the buy-side and sell-side.

With feedback and suggestions from our client base, the product is continuously evolving, progressing from automating the easy orders to managing the more complex, less liquid names for algorithmic execution. As FlexAlgoWheel has been integrated with historical and real-time data and pre-trade cost analytics, it has become more sophisticated. This article will review why the algo wheel has gained traction on the buy-and-sell-side desks and examine some of the latest developments in FlexAlgoWheel such as switching between algo strategies or tuning algo parameters in response to real-time market conditions; expansion to handle basket and portfolio trades; and integration with dark pool aggregators. Looking ahead, we will offer our perspective on where the FlexAlgoWheel is headed in terms of its augmented intelligence and expansion into other asset classes.

Why Algo Wheel

Based on the Pareto principle, which can apply to trading complexity, about 20% of the trades take around 80% of the time and attention of the traders. The initial objective of algo wheel was to delegate the simplest and least sophisticated orders through automation so that the traders can focus on hard-to-trade names; but now, with augmented intelligence, FlexAlgoWheel has become the default choice for almost all kinds of order routing ranging from the least liquid names to orders with specific characteristics or special instructions. FlexAlgoWheel leverages order attributes, static, historical, and real-time market data along with pre-trade cost and market impact models. These advances have enabled FlexAlgoWheel to make the point-in-time decision about the selection of broker/route and the algorithm along with their respective parameters.

Growing Trends: Switching Algos in Real Time

Adoption of algo wheel started with buy-side asset managers and hedge funds. However, for the past few years, sell-side broker-dealers have been increasingly looking for such a solution to present the best performing strategies and customized workflow automation as a differentiator to their clients. A key innovation is the ability to dynamically switch between algos and/or tweak algo parameters based on the current market conditions and algo performance. The concept of algo switching is not new; however, since FlexAlgoWheel focuses on order routing, we offer dynamic switching as an integral feature. Dynamic re-evaluation, an advanced feature of FlexAlgoWheel, continues to assess the real-time market conditions, order state, and execution performance for the lifecycle of the order on the street. With dynamic re-evaluation, adaptive rules are created to monitor the fill quantity, quality, and destination, to facilitate the modification of the order parameters and/or routes avoiding sub-optimal execution.

Traders want automation with control and predictability. FlexAlgoWheel sources all interval analytics, snapshots, benchmarks, and slippages calculated by FlexTrade’s real-time benchmark server (BASS). BASS (Benchmark, Analytics and Snapshot Server) is a real-time analytics engine, which monitors incoming parent orders, outgoing child orders, fills, live market data, and proprietary data sources from clients. BASS uses these data points in real-time to record snapshots, calculate interval analytics, benchmarks with slippages and various reversion numbers and makes them available in the blotter and FlexAlgoWheel to make optimum routing decisions.

Several sell-side clients executing baskets require functionality within FlexAlgoWheel that can schedule slices of the batch for open auction, intraday, and close auction at the exchanges, without any manual intervention. Portfolio level analytics are now being considered to make the routing decisions as well. . For example, a simple use case is to use net notional value of the basket to decide which broker to route to. Another interesting usage is to keep the basket atomic, intact as one unit, while FlexAlgoWheel evaluates the potential results of a market model impact model to decide when and where to route the child orders.

There are many mid-size and boutique brokers with their own quants/developers building custom models and execution algos inside FlexTrade’s proprietary algo and analytics engine. These firms use FlexAlgoWheel to offer their proprietary algos alongside the downstream offerings of other white-labelled algo providers. FlexAlgoWheel seamlessly offers the ability to perform A/B testing before a firm introduces their own customized algos to the street.

Laws of Attraction: Liquidity-Based Routing

Given that institutions use algo wheels to improve execution performance, we’ve added more intelligence to FlexAlgoWheel to search for liquidity based on the fill quantity and quality. Buy-side traders prefer to route orders to several venues contingent on the available liquidity (especially through dark aggregators). As a result, FlexAlgoWheel adopted a liquidity-based routing mechanism in search of potential liquidity and fill ratios to reload to the same destination, thereby maximizing the execution while reducing the duration and market impact. Traders need to create a schedule of distribution for a set of venues (typically dark algos), allowing FlexAlgoWheel to carefully search each venue sequentially or simultaneously. Searching sequentially gives preference to the order of the venue and can potentially reduce information leakage. Simultaneous search mode is most useful when unusual or illiquid stocks need to be scanned at multiple venues. Buy-side firms also integrate actionable IOIs as a liquidity source in FlexAlgoWheel.

Incorporating a PM’s Alpha Profiling Strategy

To improve trading performance through Algo Wheel, developing an optimal execution schedule that factors in a portfolio manager’s (PM) strategy or PM’s alpha trading profile can be critical.

For example, traders care about the PM’s trading horizon, which individual stocks and sectors the PM decides to buy or sell, and what happens with the order from the decision point forward.  The PM’s alpha profile has an investment and trading horizon component, where the investment theme is typically weeks to months and the trading horizon is generally from a day to multiple days. Using transaction cost analysis (TCA), we can calculate the PM’s alpha profile historically based on the trades from start-to-end across all orders. Without requiring any programming knowledge, traders can select various order attributes and parameters through FlexAlgoWheel’s drag-and-drop or point-and-click functionality, factoring in PM specific Alpha Profile. As real-time market conditions change, traders can adjust factors such as time horizon, aggressiveness levels, and order specific attributes.

Emulating the trader’s decision process, the algo wheel’s logic automatically adapts in real time, with aggressiveness adjustments based on factors such as arrival slippage, relative volatility, liquidity, real-time trade cost estimates, and other analytics. Ultimately, the successful strategy is one that achieves lower trading costs, automates the trader’s workflow, and frees up time to handle more complex orders.

Going forward: Other Asset Classes

As financial markets evolve, FlexAlgoWheel will respond to the changing dynamics of the markets, market structure and needs of buy-and-sell-side clients. Recent innovations have included expansion into different asset classes. Having experienced the benefits of FlexAlgoWheel for equities trading, many of our clients are actively working on leveraging the algo wheel for equity derivatives, FX, and fixed income.  Some clients have already deployed FlexAlgoWheel for listed futures, options, and FX.

With increasing sophistication, FlexAlgoWheel has become an indispensable tool for decision support and workflow automation, liquidity seeking, and optimal execution, within multiple asset classes.

Sharat Kumar

Sharat Kumar is Vice President, Head of Algorithmic and Analytical Solutions at FlexTrade. Sharat heads the development of a suite of Algorithmic & Analytical solutions that includes FlexTrade flagship FlexAlgoWheel, trading algorithms, real time analytics, alerts and Transaction Cost Analysis (FlexTCA).

Explore more

Insights

Europe’s Fragmented Liquidity Challenge

Insights

Driving the evolution of FlexOMS in EMEA

Explore All Posts