Academic research on hedge funds can be tremendously valuable to investors. But with increasing complexity, comes a greater chance that research conclusions can be interpreted in many different unintended ways. For example, a widely publicized study released last week by Andrew Ang and Nicolas Bollen was presented by some as evidence that gating provisions themselves have a certain calculable cost.
As Reuters reported:
“In a paper titled “Locked Up by a Lockup: Valuing Liquidity as a Real Option,” Mr. Bollen and Columbia Business School’s Andrew Ang show that a manager’s right to block redemption requests “generates an implied cost of between 5% and 15% of the initial investment.”
This claim was likely based on the following statement in the introduction to the paper:
“…we show that a manager’s discretion to block redemption requests using gate restrictions or suspension clauses generates an implied cost of between 5% and 15% of the initial investment.”
Gate Provisions vs Gate Closures
Like Reuters, we read this to mean that the manager’s right (option) to cease redemptions was worth the equivalent of 5-15% of the value of the fund. But a more detailed reading of the paper left us with the understanding that the manager’s actual decision to halt redemptions – not simply their option to do so – had the effect of immediately decreasing the value of the fund by 5-15%.
Ang and Bollen put a price on the investor’s option to redeem at their freedom. The following table shows the value of a fund with a given volatility, a given likelihood of failure, and various expected mean returns (6%-14% per annum) under 5 separate liquidity regimes (columns from left to right: no liquidity, 2 yr lock-up + 3 month notice, 2 yr lock-up only, 3 month notice only, no restrictions at all). The numbers represent the fair value (in dollars) for a fund with a $100 NAV on the day of its launch (i.e. at “Age=0”).
As the authors explain:
“Panel A shows that for relatively low expected returns, such as 8%, the potential cost of redemption suspension can be enormous, with $85.37 [for the no-liquidity scenario] $100.26 [for the 2 yr lock-up, + 3 month notice scenario]. This implies that the investor is receiving an asset worth about 15% less than NAV [when gates are closed], rather than one worth about par when the liquidity option is honored.” [our notation]
So it would appear that a newly-gated (or always-been-gated) fund should suffer a 15% decrease in expected value vs. a fully-liquid one. But this does not necessarily mean that a fund with a mere gating provision should be discounted by this amount.
We called Bollen to confirm this and he explained that in situations where investor redemptions coincide 100% with the imposition of gates, then an initial investment in a fully-liquid fund is, for all intensta nd purposes, locked-up. In other words, liquidity is only there when you don’t need it – and disappears when you do need it.
Of course, hedge fund investors redeem for all types of idiosyncratic reasons. So the value of the possibility of a gate being imposed falls somewhere between the extremes of a 100% liquid fund and 100% illiquid fund.
Still, an instant drop in expected value of a fund once a gate has been closed is still a tough nut for investors to swallow. This immediate discount is based on a number of assumptions regarding the fund’s volatility, mean return, the likelihood of the fund liquidating given its strategy and age, and the expected recovery rate in the event of such a liquidation.
As Ang & Bollen explain, their analysis assumes that if a gated fund liquidates, investors will receive only 50% of their money back:
“Upon failure, we assume as a base case that investors receive a payoff of 50% of the prevailing NAV of the fund, reflecting additional loss of asset value during liquidation. The 50% liquidation cost is based on results reported in Ramadorai (2008), who analyzes a sample of transactions on a secondary market for hedge fund investments conducted on Hedgebay. During 66 “disaster” transactions, involving fraud or collapse, the average discount of transaction price to NAV is 49.6%.”
But the final discount to NAV is only a proxy for the actual recovery rate for a liquidated hedge fund. Regular readers may remember Ramadorai’s paper from this AllAboutAlpha.com post. In his paper, Tarun Ramadorai of the University of Oxford finds that investors trying to redeem out of funds that eventually end in “disaster” do so at a 49.63% discount to last reported NAV (note: “disaster funds” are define as “funds that suffered heavy and publicly reported losses and are either liquidated, or likely candidates for liquidation, or have been implicated in the press for fraud.”)
But other researchers say that the actual amount recovered when funds liquidate may actually be far higher than 50% (i.e. the loss may be much lower).
In a paper published around the same time as Ang & Bollen’s, James Hodder, Jens Carsten Jackwerth, and Olga Kolokolova studied “dead” hedge funds (defined as those who stopped reporting to a database). Hodder, Jackwerth and Kolokolova argue that common assumptions about “dead” funds may be overstated:
“…we find that the estimated average delisting return is fairly small and nowhere near values of -50%… [our research] provides rather strong evidence that on average, delisting returns are far from disaster scenarios with exit returns of -50% or worse.”
In fairness, Ramadorai’s 50% recovery rate is based on funds that suffered “heavy losses” prior to liquidation and Hodder, Jackwerth & Kolokolova base their findings on all funds that simply ceased reporting to a database.
But as Ang & Bollen show, the recovery rate still has a significant impact on the value of the investor’s option to liquidate at their freedom. The following chart from the paper shows that the investor’s liquidity option on a fund with an 8% expected return has a significant value when recovery rate (“l“) is 50%, but no value at all when the recovery rate is, say, 81%.
So, the cost of the “no liquidity” (already-gated) regime in influenced by recovery rate assumptions. But the recovery rate also impacts the cost of the plain vanilla “2 yr lock-up + 3 month notice” liquidity regime. In the chart below from the paper, “restriction cost” refers to the cost of these standard liquidity restrictions (vs. the full liquidity regime).
Note that even if you assume a low 6% annual return and a 50% recovery rate in a potential liquidation, the value of the standard (2 yr lock-up + 3 month notice) liquidity restrictions is still less than 5% of the fund value.
This is a very interesting paper. But as usual, we need to be careful when interpreting it. The study is about the value of being able to redeem – not about the value of a manager’s option to cease redemptions.
Perhaps in an acknowledgment of the complexity of these issues, the authors draw only general lessons in their interview with Reuters:
“This paper was written for academics, but I think it has practical applications…The legalese that described the vague rights that managers have and that a lot of people may have skimmed over in the past could turn out to be very important.”
“Very important”, yes. But contrary to media reports, the study does not actually value “the manager’s to block redemption requests.”