Tuesday, 25 April 2017
Last updated 20 hours ago
Feb 8 2017 | 3:22pm ET
Editor’s note: Due diligence in the hedge fund world has long involved some combination of the four P’s – people, process, performance and philosophy – customized to suit each investor’s particular style or requirements. However, due diligence practitioners are often heavily influenced by four commonly-held fallacies, writes PivotalPath managing principal Jonathan Caplis, that can impact proper fund selection and evaluation, as well as unnecessarily skew conclusions.
Debunking Conventional Investment Wisdom
By Jonathan Caplis
The basics tenets of hedge fund investment due diligence have long been termed as a series of “P’s”, referring to some iteration of people, process, performance and philosophy. While those are certainly the right areas to evaluate, issues arise from the actual research practices due diligence practitioners employ to arrive at conclusions for these “P’s.”
Hedge funds are an asset class that I should avoid as they haven’t performed well recently.
Simply put, hedge funds are actively managed strategies within a vehicle and fee structure that usually incorporate a management fee in addition to a performance fee. Due to reduced regulation around limited partnership (LP) investments relative to their mutual fund cousins, hedge funds have the flexibility to employ significant leverage while trading any asset class or their derivatives, both long and short, inclusive of traditional long only equity and credit strategies.
Accordingly, given hedge funds’ broad investment mandate, lumping them into one all-encompassing benchmark such as an HFRI Index and making sweeping conclusions about them as an asset class, may cause investors to miss significant opportunities.
Nobody would dispute that New York offers some of the best restaurants in the world. However, extending the logic above, if investors averaged all Zagat ratings on New York Restaurants, they may avoid Per Se based on the conclusion that New York restaurants are just average. What’s true in the aggregate tells you nothing about the distribution.
Generic benchmarks often include thousands of funds both of institutional quality and many of which institutional investors would never consider. The myth that there are 10,000 hedge funds for institutional investors to consider is just that.
Any investment decision surrounding active managers, including hedge funds, should be predicated upon whether the investor has the tools to select active managers that meet investment objectives, risk profile, and most importantly, justify their fees, regardless of how high or low they are.
Quantitative performance over a recent time period is a good screen to filter which managers I spend time on and ultimately invest in.
While the hedge fund universe for institutional investors is a small percentage of the ten thousand often reported by the media, the number of opportunities is still relatively large. Accordingly, investors usually choose to sort funds by returns, Sharpe Ratios and other quantitative metrics to help them pare down a large universe to a more manageable figure. However, none of these metrics alone tells you anything about the risks taken to generate returns or the potential for the performance to persist in the future. Screening on those metrics may lead you down the wrong path.
Most investors are familiar with the concept of “quality of earnings” when evaluating a company, but fail to explicitly consider the “quality of performance” when evaluating a manager.
Consider a simple strategy that selects a company at random and invests in a high yield straight cash bond that pays a quarterly coupon of 3%. As long as the company makes its coupon payments, the strategy will generate a 12% per year with a standard deviation of 0, producing an infinitely high Sharpe Ratio. While this strategy would show up at the top of any Sharpe Ratio sort, there are significant risks embedded yet little insight and skill involved.
In a 2015 white paper, PIMCO offered empirical evidence that selecting hedge funds based on historical Sharpe Ratios has no predictive ability on future performance. Just as importantly, investors may also dismiss funds farther down the list that could be additive to portfolios on a forward-looking basis.
It is also important to note that Sharpe ratios are not adjusted for liquidity. Just because volatility is not observed, does not mean it doesn’t exist when there is reduced price discovery, often the case in less liquid strategies like credit. This is exactly the reason why an MBS strategy trading off-the-run credit generates very high Sharpe ratios while more liquid exchange traded strategies like CTAs are unjustly penalized for producing lower ones. Only if the portfolio were liquidated would one observe the true volatility.
As significant, and just as commonly misunderstood, is alpha and how it is affected by liquidity. What shows up as alpha in a univariate linear regression is often just liquidity premium or some other unspecified factor in disguise. Similar to the effect on volatility, reduced price discovery in the underlying securities relative to the benchmark, or simply a mismatch in liquidity between the benchmark and portfolio, can create the allusion of alpha when additional factors might easily explain it.
Even if the model is properly specified, investors usually only observe the level of alpha but fail to consider whether that alpha is statistically significant. PIMCO points out in the same paper that alpha must have some level of statistical significance for it to be meaningful or it may just be spurious.
Accordingly, investors benefit when they understand the quality of performance in addition to standard performance metrics. When implemented properly, quantitative analysis can provide insight to differentiate fortunate market timing from skilled security selection, real alpha from unspecified beta, adjust for differences in liquidity, identify whether portfolios are properly marked and most importantly, help determine whether the investor should pay for a market risk that may be easily and cheaply replicated.
At a minimum, sorting on quality of performance will help investors filter the universe of funds in a more forward looking and valuable way, and increase the probability of focusing on the right opportunities.
A.I avoid investing in funds that are in an existing drawdown.
B.I only invest in funds that are in an existing drawdown.
Both of these philosophies are faulty if the drawdowns are not considered in the right context. Without understanding the investor composition and its overall “happiness”, investors fail to grasp the implications of a drawdown.
While every investor and manager worries about the implications of investor concentration, it is mostly considered anecdotally, and rarely in the right context. Even fewer effectively quantify and monitor it. PivotalPath has captured this dominant risk in a two-dimensional graph it calls the Liquidation Risk Matrix (LRM). The x-axis captures the liquidity of the underlying securities, while the y-axis quantifies the stability of the investor base. It is important to note that the stability of the investor base incorporates investor diversity, and estimates the overall “happiness” utilizing prospect theory, i.e., what is the probability of a large redemption in the near term? When that probability is considered with the underlying liquidity and redemption terms, the position on the graph visually quantifies and captures the number one risk to the investor and fund manager.
As assets grow and portfolio liquidity changes, redemption terms remain static. Accordingly, when poor performance or external factors cause an investor rush for the exit, well known funds such as Claren Road, Third Avenue, Luxor and Lioneye, among many others, have been forced to suspend redemptions, create side pockets, SPVs or even shutter when unable to meet these requests. This all could have been predicted well before the event.
Consultants advise that it takes 9-12 months to complete due diligence a manager from scratch. Prior to beginning the process, it is beneficial to “watch” a manager for a similar period of time.
The typical investment consultant often takes nine to twelve months (or more) to complete due diligence on a fund. This can be quite frustrating for both institutional investors as well as the manager of interest. Additionally, the length of time required makes tactical allocations challenging, leading capital to flow into a strategy long after the bulk of the opportunity has passed.
Consultants tend to rely on analysts who have experience with the manager or have seen a manager in action before. This familiarity is often seen as a way to truncate the due diligence process, which may seem appealing given the length of time quoted above. However, this “experience” is often problematic from the beginning as the analyst may be primed with an opinion based on their own network or limited information which may no longer be relevant.
The typical due diligence process begins with a short meeting or call where the analyst subsequently concludes whether the manager is “good” or “bad.” If bad, they stop the process and move on. If good, they continue to “watch” the manager to build more confidence surrounding their initial conclusion.
As the process continues, and more time is invested at the expense of due diligence on other opportunities, pressure builds in the form of “sunk cost,” to reach a positive conclusion. Over the 9+ months, the analyst continues to have further calls to confirm their original conclusion or bias, often downplaying any evidence to the contrary.
While experience can be invaluable when utilized correctly, it can be a hindrance if biases are not acknowledged and corrected for. Top-tier research actually employs the opposite methodology. Due diligence should be conducted within a rigorous and repeatable process featuring multiple participants, independent data sources and analytical methods that create checks and balances. Biases are mitigated and the time necessary to reach a robust conclusion is dramatically reduced.