Last week, we told you about an academic study that showed hedge funds weren’t just net providers of market liquidity, but that some strategies (e.g. emerging markets) demanded more liquidity that they provided to markets. For these funds (which, according to the study, exist in various degrees across all strategies) the “illiquidity premium” may be a critical source of returns. In fact, hedge fund skeptics often say these kinds of funds are “short volatility” since they profit consistently when volatility is low, but have to pay the piper when volatility spikes – and liquidity dries up.
Much of the discussion about this “illiquidity premium” has focused on the amount of returns that must be ascribed to it – and therefore not to manager skill (see previous posts: “Is the vaunted “illiquidity premium” partially an illusion?“, “…3.96% per annum“, “Study finds many hedge funds simply hold back liquidity to power returns“, “Apples to Apples: the case for liquidity adjustments to hedge fund of funds returns“, “Research puts price on hedge fund illiquidity premium”, “Liquidity Insurance”, & “Liquidity Alpha”)
But a new white paper by analytics firm Risk Metrics and fund administrator RBC Dexia proposes method of valuing liquidity that is not based on ex poste returns, but is instead based on the immediate impact of illiquidity on fund managers and their investors.
The firms define two types of liquidity. The first is funding liquidity or “the ability to raise funds in order to support its normal business activities.” For fund managers, funding liquidity comes in two forms: investor allocations and margin provision. (Regular readers may recall the Hedge Fund Fame and Fortune post discussing a research paper that explores the implications of this two-headed beast).
The second form of liquidity described in the Risk Metrics/RBC Dexia white paper is asset liquidity. For fund managers and non-managers alike, this refers to the market depth of their holdings.
When a fund administrator like RBC Dexia values a portfolio, it can be liberal and mark it to market, or it can be ultra conservative and use a “mark-to-exit” valuation (i.e. value the portfolio as if the manager was forced to liquidate the whole thing). Rather than pursuing either of these poles, the firms advocate “a valuation framework that fits between the two extremes and more accurately reflects the investor’s actual constraints.”
The first step in finding this happy medium is to understand the market liquidity of holdings using the “marginal supply-demand curve (MSDCs)” for each stock (see example below from paper). By knowing the market impact of buying and selling significant blocks of stock, the fund managers is then able to calculate the “mark-to-liquidity” value of the portfolio.
But how much “distress” should the manager factor in? How far down the demand curve should they envision going? Risk Metrics and RBC Dexia propose that funds have a well-articulated “liquidity policy (LP)” to answer this question. The LP might say, for example, “Be prepared to raise 1M in cash while maintaining the desired sector allocations and limiting risk to an acceptable level.”
The LP essentially defines the extent to which the manager needs to be prepared to take a haircut if the articulated liquidity-demanding event were to occur (e.g. a major redemption). Satisfying the LP would obviously incur a cost in the form of downward price pressure on holdings. That “liquidation cost” is essentially treated as a handicap – as if you were saying “My portfolio is worth $10m, but I would take a $100k hit if I were to liquidate $1m within one day.” The mark-to-liquidity value would then be $9.9m.
While a manger could always dump his/her highly liquid large cap holdings in a pinch, this would pervert the existing sector allocations and thus contravene the stated LP. But as the manager dumps holdings overboard in the requisite equal proportions, s/he is likely to take a bigger haircut on the less-liquid holdings (e.g. emerging market stocks) than s/he is on, say, US large caps.
The paper walks through a practical example of a real 36-stock portfolio (with unreported MSDCs for each stock). The liquidation cost is negligible when the LP is somewhat toothless. But when the LP demands a high level of liquidity, the cost escalates dramatically (as you can see from the example fund used in the white paper below).
The paper concludes by making the point that in the absence of such a “liquidity handicapped NAV” (our term), the earliest investors to redeem are at an advantage since they shift the burden of generating liquidity to remaining investors. Buy marking down the value of the portfolio right off the top, all investors share the burden in a sort of “insurance policy” (our term) in case of mass redemptions. Risk Metrics and RBC Dexia say that this is not only fair, but it also reduces the incentive for investors to start a stampede for the exits:
“An investor with otherwise no need or desire to redeem may in fact choose to redeem because of the fear of other investors’ actions. This dynamic can lead ultimately to the “run on the fund” scenario, with investors rushing away from the fund simply because they do not want to be last.”
This makes a lot of sense to us. In fact, we wonder if this might have helped to mitigate at least some of the panic of 2008. The key questions are whether the MSDC curves adequately factor in a market-wide liquidity crisis (they are, it would appear, based on “normal” market circumstances) and whether the liquidity policies of individual funds would be realistically aggressive (assuming a positive illiquidity premium and a hefty incentive fee, manager may not choose to set particularly aggressive LPs).
It also seems to have implications for the generally subjective art of gating. If the circumstances requiring gates could be more objectively articulated to investors, maybe those investors wouldn’t feel hoodwinked by their managers when they hear the news that they’ll be sticking around for a while.