The evolution of risk strategies for transition management

The transition management business is a very dynamic industry; seeing changes in form, function and process at a rapid rate for the last decade. Our industry has shaped itself over a remarkably compressed time period, which means that looking back at "the way it was" doesn't require one to go too far into history to notice very significant changes. 

Early beginnings

The genesis of our business came with the recognition that there 

were savings to be had by more efficiently project managing changes to investment portfolios. At this time, (pre-2000) trading was largely a by-product of the project management activity. The first priority was to 

coordinate movements to reduce exposure; the second, to attack tangible elements of cost: i.e. market spreads and impact. With that, trading tended to focus on these two factors alone. Crossing (in many forms) was popular as it was easy to point to half bid/offer spread cost savings and the lack of market impact with such transactions. The industry mantra became, "cross, cross, cross and then clean up on the open market"; the more one crossed, the more one "saved". 

 A New Era - bringing risk to book

True to market forces, the industry witnessed a rapid influx of new transition providers which forced all participants to refine and improve their offering. Now, the industry began to ask, "How much are we risking, relative to what we are saving"? This was a fundamental change. Providers were now not only quantifying expectations of cost, but also the degree of certainty around those expectations (ie. opportunity cost dispersions). To compete, providers had to devise ways to reduce cost, while at the same time minimise the dispersion (increase the certainty of the estimates being put forth). 

However, it's one thing to state that a particular strategy is more or less risky, but quite another to quantify that risk. 

So how much is at risk with a crossing strategy? One way 

J.P. Morgan looks at this is to take a single market and monitor the daily average tracking error of one industry sector relative to another. The outcome from the most recent 30 days (August 2010) is 62 basis points. This is significant when one considers this approach is typically taken precisely in order to avoid much lower levels of commission, spread and impact. 

Risk management comes of age 

A new theme arose - risk management. This focus was important as it helped to evolve the product from process management to one that also concentrated on trade strategy development and execution management. The variants of ‘Risk Management' ran the gamut - ie. from maintaining a cash balanced trading strategy to introducing trade optimisation. 

Through trade optimisation, for example, J.P. Morgan began to measure tracking error between the aggregate legacy and target portfolios, and more importantly, drill into the marginal contribution of tracking error at a name by name level. This facilitated more sophisticated strategies - ie., identifying which names were the highest marginal contributors of risk and which names we would prefer to hold in the portfolio rather than simply selling with market liquidity (low marginal contributors of risk).

These developments also coincided with the proliferation of algorithmic trading. This, along with technology enhancements in many of the larger investment banks, pooling trade flow volumes from various businesses (internalisation of order flow) and 

electronically harnessing market fragmentation so that transactions could be executed across many venues (smart order routing), meant transition managers could now deliver significant advances in trade strategy development and execution. 

Many providers also introduced hedging into the mix, through index futures or Exchange Traded Funds (ETFs). While adding yet more tools, these products sometimes add as much confusion as they improve risk management. It's particularly important to measure the "fit" of the hedge relative to the exposure. For example, if implementing a hedging strategy will reduce the tracking error estimate from 9% to 2%, the hedging strategy would be worthwhile. However, if hedging only reduces the tracking error estimate from 9% to 7%, then taking into account the short duration of the hedge and increased transaction costs, this may not generate sufficient savings to warrant its use.

Staying ahead of the curve

Technology continues to create opportunities for enhancement. We can easily filter names in any transition event according to those that "live well" within an algorithm, and focus high touch trading on other names that are the largest contributors of risk, the largest drivers of performance and which do not trade well in an automated fashion. We can reach multiple execution venues simultaneously, trading names (or even substitutable names) more cheaply and with greater speed (and lower opportunity risk) than ever before. Incremental improvements to computing memory and speed have allowed us to take an iterative approach to trade optimisation by measuring risk intra-day, creating an optimised trade basket and executing, and then re-measuring. We can measure the success of the approach and repeat the exercise, or we can note the lack of success, and adapt to new market conditions. 

Looking ahead, questions focus now on how to manage risk in "real-time", or whether advances in trading mean that operational risk exceeds execution risk. The transition manager has no choice but to move with the markets. The really excellent transition management providers, however, look to leverage technology to move ahead of the market, and I, for one, am excited to see what the next decade has in store.