Liquidity Provision in the Secondary Market for Private Equity Fund Stakes

Once marginal due to contractual restrictions on transfers, the secondary market for Private Equity Fund (PEF) stakes took off soon after the 2008 financial crisis to an annual turnover of over $30 billion per year.

Behind this growth are the liquidity needs experienced by some investors since the crisis and the increased need of investors to rebalance more actively their private equity exposures as these are now larger fractions of their total asset holdings. While the demand for PEF stakes on the secondary market as grown in lock step, transactions occur on average at a discount to the Net Asset Value (NAV) of a fund. These discounts vary significantly over time and across funds, leading potentially to a large liquidity risk faced by the Limited Partners (LPs) of specific PE funds who may have to sell their stake in the secondary market when discounts are also large.

In this paper, we study the determinants of these discounts. One main reason found in the literature, echoed by practitioners, is that secondary market discounts are a compensation for liquidity provision: when funding liquidity is low, LPs may be forced to sell their stakes for cash while potential buyers are also strapped of cash. Another explanation is that secondary market discounts are a compensation for asymmetric information: expecting that incumbent LPs will only accept bids when they are higher than the privately know true NAV, bidders may respond with discounted bids. The goal of this paper is to identify the impact of the liquidity compensation channel on bid discounts and quantify the proportion of the discount attributed to variations in observable measures of aggregate liquidity. The main challenge is that the asymmetric information and liquidity provision theories have predictions in common.

Insufficient liquidity provision means that a seller will have to accept a low price for a PEF stake. Likewise, because LPs may have an information advantage over the value of a PEF, buyers will submit discounted bids on average. To distinguish between the two hypotheses, we follow a tradition in the microstructure literature, and exploit the fact that the liquidity and private information motivations for trading have different dynamic implications. In particular, if selling were truly the cause of liquidity shocks, then the fund’s expected cash flows will not fundamentally change and as a result bid levels will eventually revert from their pre-shock values. Moreover, given that reported NAVs cannot incorporate liquidity discounting under U.S. and European fair value rules (SFAS 157 and IFRS 13, respectively), NAVs will also eventually decrease to a permanent lower value once all information is fully revealed. If, on the other hand, demand was responding to the possibility of informed selling, then bids would adjust downwards gradually as they increasingly reflect the private information. Moreover, current demand flows to a given fund would predict its future NAV performance.

We study the demand flows to different private equity funds with a unique data set consisting of 4,365 bids on 497 LBO funds by 144 bidders submitted to a London-based secondary market intermediary of private equity stakes between September 2009 and December 2016. Our data provider is an important sell-side agent for LPs wishing to exit via the secondary market. Its bid book is comprehensive and representative of the global market.

We find a large heterogeneity in the response of demands for different types of funds and by different types of bidders to aggregate liquidity shocks. For example, the overall number of bids decreases in months where the yield curve steepens, yet this total effect comprises a decrease in bidding for the old and middle-aged funds but a small increase for the younger funds. Similarly, an increase in bond market illiquidity is also associated with an overall decrease in demand, especially for old funds, but is associated with an increase in the demand for young funds. We observe similarly heterogeneous patterns across fund sizes, where demand flows out of large funds towards small ones in response to liquidity shocks.

Our demand model produces a measure of liquidity-driven demand, i.e., the component of the demand of a given fund that is explained only by liquidity shocks. In consistency with the liquidity provision hypothesis, we find a significant negative correlation between bid levels and the liquidity-driven demand. Rejecting the asymmetric information hypothesis, we show that the relation between liquidity-driven demand and bids is contemporaneous and not dynamic: past demand shocks are not correlated with current bids and liquidity-driven demand flows do not predict future fund performance. Moreover, the negative correlation is strongest for the youngest and the smallest funds, whose LPs are expected to face the highest liquidity risk.

To identify the liquidity providers, we estimate the liquidity-driven demand by type of investor: Secondary Funds, Funds-of-funds, and other asset owners (pension funds, endowments, banks). Secondary Funds are responsible for three quarters of the bids in our sample and their demand responds the strongest to most liquidity drivers. Yet we find that its asset owners, not Secondary Funds, who increase their bidding in response to some liquidity shocks. Consistent with liquidity provision, asset owners submit bids that are on average 1.5 percentage points lower when their liquidity-driven demand increases by one sample standard deviation. One possible explanation for this effect is that, because they are not constrained to invest mostly or even exclusively in the secondary market, these investors have the flexibility to bid in this market when capital call risk or portfolio rebalancing puts pressure on other constrained investors.

Download the complete
article PDF

Comments

There are no comments for this entry yet. Be the first to add your thoughts!

Add Your Comment