How to Quantify Liquidity Risk

1183935_scout_icon_setIn the great market tumble of 2008-2009, many pension funds were caught between a rock and a hard place – or rather an intractable liability and an unyielding market. To fund ongoing pensions, particularly when a plan is mature and in payout mode – i.e.,  no new contributions or contributions that are insufficient to offset outflows – they have to cash in assets. But cashing in assets would have meant liquidating at close to the bottom of the market – except for Canada bonds. So liquidity has become another aspect of the risk apparatus.

A new paper by RiskMetrics and RBC Dexia, “The Value of Liquidity: Can it be measured?” written by Carlo Acerbi and Christopher Finger, both of RiskMetrics examines the role of liquidity. They distinguish between two forms of liquidity: funding liquidity and asset liquidity. “One element of funding comes from the investors themselves; the risk being that investors redeem in greater numbers or more quickly than the funds’ assets can support. Indeed, just as banks may suffer a ‘run’ when their depositors withdraw quickly, a fund can suffer adverse shocks to its NAV in trying to meet heightened redemptions…. Asset liquidity, on the other hand, relates to the depth of financial markets and the ease with which a security or portfolio may be converted to cash. Under most circumstances, a large cap equity or a large, widely held bond issue may be sold relatively quickly and large orders are likely to attract roughly the same price as small ones. For more lightly traded securities, the market may support only relatively small trades, or large trades may only be possible at a significantly discounted price.”

Finger and Acerbi propose a framework that incorporates both elements: funding and asset liquidity. They begin with a dichotomized model: the investor who faces no liquidity restraints and the investor who does. For the investor who is constrained,  “it would be irresponsible to value this portfolio according to the best market quotes, knowing that the investor will shortly be liquidating, and likely not realising these best prices. Rather, the investor should value the portfolio according to the expected proceeds from the forthcoming liquidation.”

They call this “mark-to-exit” or “mark-to-liquidation.” This is the first element of the framework. “Beyond simply the best bid or offer in the market, we require information (or at least a hypothesis) for the price we can expect to realise for transactions of any given size….Our approach utilises marginal supply/demand curves (MSDCs). In the middle of the curves are the best bid and offer prices–the prices we assume we can realise for small transactions–and further out on the curves is information about the price at which we could liquidate our entire holding.”

The second element requires an explicit liquidity policy. “The LP accounts for the endogenous constraints to which a particular portfolio owner is subject. A simple LP might be ‘Be prepared to raise 1M in cash within one week’s time’.”

By means of these two elements, an exogenous MCDC and an endogenous liquidity policy, investors can calculate the “liquidity-adjusted value of the portfolio.” In turn, that may lead to no action, or it could mean bringing a portfolio back in line with its liquidity policy. If the latter, then there are a variety of procedures such that the “mark-to-liquidity value of the portfolio then is the mark-to-market less the optimal cost of achieving the LP. This optimal cost is referred to as the portfolio’s liquidity impact.”

Establishing liquidity impacts, however, is a reactive policy, Acerbi and Finger note. “The mark-to-liquidity framework does not truly change investors’ behavior in the face of liquidity constraints. The optimisation gives us a means to react more efficiently to the need to liquidate, but we are only reacting and not planning. This brings us back to the notion of liquidity policies and our emphasis that in all our examples, those policies began with the words ‘Be prepared’.”