by Frederic Blanc-Brude, Director, EDHECinfra
Asset owners are winning the argument to lower private equity manager fees, their next battle will be about the valuation of private assets. Private equity funds have long been characterised by high fees. Even as the number of asset managers (general partners or GPs) offering PE funds increased steadily over the past decades, competition for the attention of limited partners (LPs) did not lead to an immediate shift in the cost of mandating specialist managers to buy and sell private companies.
Recently however, as reported here by Sarah Rundell (2016), years of pressure from asset owners have led to a seemingly ineluctable trend towards fewer and lower private equity management fees.
The fact that increasing competition did not immediately lead to lower costs, when these were high to begin with, is an interesting puzzle.
Economists would see a case of strong information asymmetry between buyers and sellers, combined with a case of “type pooling” i.e. in a market where some managers are capable of delivering a high quality service at a fair price (type A) but most are not (Type B), both types of manager can tend to “pool” together and offer the same low quality product at a high price.
Competition fails. This common phenomenon (think “finding a good plumber”) is the result of information asymmetry: clients cannot tell beforehand which service providers are of type A or type B. Fee levels are only a consequence, or a symptom, of information asymmetry between GPs and LPs. High fees had so far been acceptable to LPs thanks to high reported returns. But as the recent decisions by CalPERS or NYCERS to pull out of hedge funds altogether illustrate, lower returns in recent years have made fee levels much harder to justify in alternative strategies.
Reported excess returns in hedge funds or private real estate are lower in part because they are better measured: valuation methodologies and available data have improved sufficiently to allow quasi-market valuations to be reported. Hence, out-performance measurement has become more accurate, and also lower because a greater proportion of hedge fund or private real estate performance can now be understood as a combination of market betas. The evolution of asset pricing techniques and data availability have led to more accurate risk-adjusted performance measurement and to a re-evaluation of the benefits received by LPs from delegating investment decisions to a hedge fund or private real estate manager. In other areas of the alternative universe, private asset valuation remains a source of significant information asymmetry between GPs and LPs.
For very illiquid and thinly traded assets, such as infrastructure, investments returns continue for the most part to be reported on the basis of quarterly or annual appraisals, which leads to a number of well-known issues: stale pricing, return smoothing and the impossibility to estimate the true diversification contribution of such assets in the absence of reliable measures of co-variance. When assets are valued by discounting 25 years of future cash flows, the choice of cash flow scenario and discount factors explains most of the reported performance. In the absence of robust cash flow forecasting and discounting techniques grounded in financial theory, the entire reporting exercise amounts to a set of ad hoc and often opaque assumptions. Nor is this news to asset owners.
In a new EDHEC Infrastructure Institute/ Global Infrastructure Hub survey of institutional investors’ perceptions and expectations of infrastructure investment (link), less than half of respondents declare trusting the valuation reported by private infrastructure asset managers. A quarter declared not knowing whether or not they can trust them, and a little more than a quarter bluntly reports not trusting reported NAVs in infrastructure PE funds (see Blanc-Brude et al 2016). Today, in the infrastructure and other very illiquid asset classes, reported returns are also somewhat lower because new investments are being made at higher prices, and the pressure from LPs to lower fees is on.
But bludgeoning service providers into charging lower fees (sometimes with the help of the regulator) without asking them to provide better service does not solve the market failure identified above. Those who pull out of delegated investment can either abandon an asset class altogether (no more hedge funds!) or try and internalise the skill set. But in both cases this creates opportunity costs from lower diversification with and within the asset class (you cannot replace twenty asset managers with one internal team). In the latter case it also creates direct costs. Furthermore, private investment teams operating within asset owners still mostly report the same stale NAVs and IRRs, while those who will stick with private asset managers who can only charge low fees (because the regulator said so) may now be faced with 100 percent of type Bs.
So are asset owners condemned to maneuvering between the Scylla of DIY private investing and the Charybdis of opaque and expensive delegated investment in private assets? There are solutions to minimise the effect of information asymmetry in market dynamics. To avoid the pooling of manager types, market participants can create “sorting devices” (Spence 1973, Rothschild 1992) or “revelation mechanisms” (Laffont and Martimort 2002) to facilitate the processing of information from uninformed to informed participants. In economics, this problem is typically modelled as a “market with adverse selection and competitive search”, where some agents post terms of trade (term sheets) and others aim to screen the other side of the trade by agent type (Guerrieri 2010).
In such models, the informed side of the trade (here the asset manager) can move first and signal to the market what terms they can offer, or the uninformed side (asset owners) can move first and request bids for a discriminating “menu” of potential products or investment solutions, which includes items that only type A managers can deliver, prompting type Bs to exit this segment of the market. Hence, beyond the debate on fees, the need to better discriminate between private asset managers should lead asset owners to require better reporting, better valuation techniques and better risk-adjusted performance measurement.
Their next battle will logically be to improve the valuation framework of private assets and move away from reporting single IRRs and towards a fully-fledged factor decomposition of returns. In fact, CalPERS did keep one hedge fund manager on its books, the only one that could offer a hedge fund strategy defined in terms of alternative betas. Still, it should be noted that solutions to problems of information asymmetry always involve the uninformed side paying an “information rent” to the informed side for revealing its type.
In other words, asset owners should not expect to get type A managers on the cheap. It remains in their best interest to try and discriminate between them and the rest, because it can lead to the cheapest outcome for a given level of risk-adjusted performance. Finally, given the recent trend on fee levels in private equity, it is also in the best interest of type A managers to signal very clearly that they can deliver better measured out-performance by adopting more transparent and useful performance reporting standards, and valuation methodologies that are better adapted the nature of the information available about private assets.
– Blanc-Brude, Frederic, Grace Chen and Tim Whittaker (2016). Towards Better Infrastructure Investment Products, EDHEC Infrastructure Institute-Singapore
– Guerrieri, V., R. Shimer, and R. Wright (2010). Adverse selection in competitive search equilibrium. Econometrica 78(6), 1823-1862.
– Laffont, J. J. and D. Martimort (2002). The Theory of Incentives: The Principal Agent Model. Princeton University Press.
– Rothschild, M. and J. Stiglitz (1992). Equilibrium in competitive insurance markets: An essay on the economics of imperfect information. Springer.
– Rundell, Sarah (2016). Investor Pressure on Fees, top1000funds.com, 20 April 2016
– Spence, M. (1973). Job market signalling. The quarterly journal of Economics 87(3), 355-374.
Read the article online here