If hiring levels are an indication, the buyside is taking the bull by the horns, as it were, by in-sourcing trade execution. Many of the new hires come from the sell-side where trading desks have been shrinking for years. As sell-side brokers retrench due to regulations that have increased their cost of capital, there has been a decline in principal trading which was the most effective way to find a counterparty for a large trade. At the same time, the unbundling of sell-side services has shone a light on execution quality-agency trading is less likely to be accepted as a commodity in exchange for research, the value of which is being increasingly questioned by portfolio managers.
Disintermediation enabled by technology has affected asset management no less than any other field. To what extent has or can buyside trading desks elevate technology to the level of strategy to capture alpha? To a large extent, as it turns out-the ability to minimize market impact costs is a competitive advantage of asset managers that persistently outperform. According to one study, the gap in performance between the top- and bottom-ranked institutional trading desks is 62 basis points, a differential that cannot be explained by market conditions, the particulars of the order or investment style. Given the fact that 40 basis points can move a mutual fund manager 1 up or down 10 places in a peer group ranking, trading skill is a major contributor to alpha.
Whats more, top tier desks actually manage to achieve negative trading costs by providing liquidity to the market-posting limit orders or responding to order imbalances. Dimensional Fund Advisors, for one, has been a market maker for a good part of its history-portfolio managers at DFA see trade execution as an integral part of the portfolio construction process, delegating timing decisions to traders who assess momentum indicators.
For buyside desks that are not prepared to cross the divide between liquidity taking and liquidity making, what can they learn from top ranked institutional desks? What is trading skill? It is theexpert use of a decision matrix to optimize the configuration of venue, means-high-touch or low-touch, type of algo–and timing, among a myriad of other variables in achieving best execution. Skill lies in the decision-making process, and in this regard technology is merely a competitive requirement, it does not in itself confer an advantage, as institutional desks are quickly learning. Greenwich Associates estimates that algorithms are now used to handle up to 80% of buyside order flow, but institutions are increasingly dissatisfied with algos that cannot be customized.
While technology will doubtless improve, the choice of algo should be more deliberative and less reflexive. In this regard, there is a disconnect between the algos commonly used and those that are most effective. For instance, scheduled algos are the tools of choice by buyside traders for large orders, whereas implementation shortfall and dark algos are best given the opportunity to find blocks in dark pools. On the other hand, fixed participation rate algos have consistently delivered better results in the execution of large cap orders with low participation rates, whereas traders too often use VWAP algos for these orders, unnecessarily increasing opportunity costs. A decision matrix should take account of the trade-offs in each configuration of means, time and place; transaction cost analysis can be used to measure these trade-offs in real-time, enabling traders to modify orders as required by market conditions, thereby increasing control over the trade process.
More importantly, a feedback loop-testing how the market reacts to the order and refining it as information is received-help traders hone their skills. There is a tendency in any profession to form habits that are hard to dislodge. Similarly, there is a risk that buyside-only venues are becoming the Maginot line of trading-liquidity needs to trade, not just pool, as one commentator has stated. There is an opportunity for the buyside to use these consortiums as an incubator-carrying out market tests of trading strategies in order to refine approaches as market structure changes. As I stated in a previous post, its high time that we use technology in trading to create online communities of interested parties, as is done in other areas, rather than simply using technology to obfuscate and evade detection.
