Strategies to fit the benchmark

Mark Maloney, director of autobahn Equity® sales at Deutsche Securities in Tokyo, looks at strategies traders can deploy to optimise IS-benchmarked orders.

Boom and bust. Fear and greed. Though famed investor Jesse Livermore once said, “Wall Street never changes,” trading microclimates do change. As markets become volatile, bid/ask spreads widen, and posted quantities drop. Average trade sizes shrink, but prints occur faster. Volume rises, though crossing opportunities are more elusive. Trading costs grow, sometimes exponentially. In volatile markets, rigorous control of execution risk is increasingly valuable. Wellcrafted algorithms are at the forefront of risk control, offering a choice of methods to employ agility and efficiency in high-velocity trading.

 Implementation Shortfall (IS) is a familiar benchmark, although there is little consensus on how to transact against it, or even how to define the approaches. Practitioners often swap the terms Arrival Price, In-Line and Implementation Shortfall without distinction, creating confusion regarding the behaviour of discrete strategies designed to minimise shortfall from a benchmark price. Algorithmic constructs differ significantly in their risk profiles.

Scaling and scheduling Two basic approaches to trading versus the Implementation Shortfall benchmark involve scaling and scheduling. Scaling approaches vary participation rates as securities move around a benchmark price: generally users choose ‘mean reverting’ style (participation increases as price moves favourably and falls as price moves away), though there are some ‘trend following’ fans that increase participation as price moves away. Scheduling strategies construct trading trajectories without specific reference to price – the best of these will optimise between criteria specific to the security and market environment (spread, volatility, order size). Each approach has its strengths. Scaling strategies will react appropriately to intraday movements while an optimised scheduling strategy should reduce variability of shortfall costs.

Trading trajectory – managing your risk While practitioners often use some variation of a Percent of Volume algorithm for IS-benchmarked orders, research¹ suggests an IS-benchmarked order should be traded differently depending on one’s risk profile. In general, the less risk one wants to assume, the more the execution schedule should be frontloaded rather than constant (Figure 1.).

Deutsche empirical analysis reinforces the conclusion that over the long term, a true optimised IS strategy outperforms a percent- participation or VWAP. This is a function of optimising the benefits of trading rapidly (lower exposure to market risk) against the advantages of trading slowly (lower market impact costs.)

Hybrid and portfolio approaches
Recent development is blurring the distinction between scaling and scheduling. Aggressive-in-the- Money scaling parameters or other liquidity-seeking functions, added to a core scheduling strategy, allow for uniquely hybrid execution styles. Risk models dynamically update schedules when scaling or manual interventions take advantage of favourable price changes. We believe the value increases with order size, and so can be particularly helpful with the largest/ most difficult orders in a trader’s book.

A new strategy on the horizon in Asia is the portfolio- level Implementation Shortfall algorithm, which is becoming known on trading desks in Europe and the US. While a single-security algorithm can only consider the execution risk of one name, a portfolio-level IS algorithm is able to optimise a single stock’s execution risk with its hedging effect to the entire portfolio. What’s the benefit of completing a liquid order quickly if completion will harm the risk characteristics of remaining orders? None – or negative! As securities are purchased and sold, impact and risk models rebalance trading trajectories on the fly to reflect the changed conditions.

As tick sizes reduce (scheduled for Japan in Q3 2008²) and execution venues proliferate, leading to harder-to-manage trading conditions, the use of robust, adaptable and empirically proven strategies will be vital for consistently and optimally trading against the Implementation Shortfall benchmark.