Explaining quant fund performance as illiquidity risk

There has been a growing amount of grave dancing on quant funds by traditional stock pickers. These quant funds have performed well over the last five years but have seen significant declines in performance over the last month. It may take some time before we fully understand the return declines of these funds. Some would suggest that herd behavior was the cause. Too many managers were using the same factors to squeeze out extra returns. Once some tried to get out of their trades, there was a liquidity problem. Certainly that could be a contributing factor to the decline in performance, but there are other explanations which are similar but provide a better framework for discussion.

There is a growing body of research on illquidity as another factor that can be used to describe the return behavior of stocks. Liquidity risk, as a missing factor, can explain the conditional behavior of stocks and can also provide a prescription for when this type of factor will contribute to performance. This story is well presented in the research work of Dimitri Vayanos in his paper, "Flight to Quality, Flight to Liquidity, and the Pricing of Risk". See http://papers.ssrn.com/sol3/papers.cfm?abstract_id=509858. Vayanos is one of the rising stars in finance. His work is a thoughtful presentation on the liquidity issue that provides a good explanation for the under-performance of some quant funds.

The framework is not dissimilar to work on bank runs. There is a need for liquidity by money managers because clients will demand redemptions during times of financial stress. This will often be associated with higher volatility or uncertainty. It will also be periods of transition in asset returns. These will also be times when there is increasing risk aversion or a reduction in risk appetite. Assets will become more negatively correlated with volatility as compensation is required for holding these assets. If this is happening across all assets, there will also be an increase in the correlation of stocks. We have seen this increase in correlation whenever there has been a downturn in the equity market.

If there was not a threat of redemptions, there would be no need for immediate liquidity and this effect will be less. These redemptions can, of course, come at any time but there is a higher probability during those periods of financial stress.The conclusion from this work is that unconditional analysis will not capture this type of risk, so when this type of event occurs, there will be a breakdown of the relationships that usually occur in the market. The problem for a quant shop is that this type of conditional factor does not occur very often, so it is hard to capture. We are thankful that the threat of redemption runs occur infrequently, but the very fact that these events are rare means that it can have a important impact during concentrated periods.