Reporting: a better performance measure

Past performance is no guarantee of future returns. RiskMetrics’ John Matwey says more and more investors will therefore inevitably demand third-party reporting of risk positions taken by hedge funds

A track record of positive returns is a good starting point for anyone looking for a good fund manager. Nobody wants to invest with a manager who either doesn’t have a track record or who has already demonstrated a propensity for losing money. Past performance may not guarantee future returns, but good performance is certainly better than unknown or bad performance when it comes to attracting and retaining investment capital.

So, for lack of any better measure, investors have continued to chase good returns and shun bad, while rationalising their reliance on past performance as a corollary for future return.

But working out the precise link between past performance and future returns is less easy. And maybe there are more telling measures of a manager’s proficiency.

A recent study of past performance presented by David Gordon of Glenwood Capital* showed that there was no statistically significant correlation between past manager performance and future returns.

Volatility, however, is another story. The study also found that there is a statistically significant correlation between past and future volatility. The findings suggest that while past performance is nothing more than a lagging indicator, volatility of returns is a better indicator of what’s to come.

Intuitively this makes sense: highly volatile strategies tend to remain volatile because the manager is investing in inherently risky securities. When the strategy works in their favour the returns are quite good. But as soon as things go wrong, performance deteriorates. We can see this at work time and again: value stocks go out of favour and the value manager suffers. Technology stocks suffer, and so do the tech fund manager’s returns. After months and months of positive returns, convertible arbitrage strategies begin losing money when equities rise and volatility drops.

When these things happen, who is in better shape – the investor who knew the risk profile of the fund and knew he was in for potentially wide swings in value, or the investor who was betting on the past two or three years of good returns?

Assume for a moment that fund selection is limited to managers with proven past performance that meets or exceeds a certain hurdle. If we accept that there is little correlation between past and future returns, where should we be focusing our attention for the initial and future allocations of capital? If past volatility is significantly correlated with future volatility, surely we should be looking more closely at the risk profiles of hedge funds and fund of funds? As alternative investment markets grow more popular, attracting investment from pension funds and mass-affluent individuals, it is critical that hedge fund risk is clearly and accurately reported. While we don’t need to know everything about a fund, not knowing its risk profile is like throwing darts blindfolded and hoping for a bull’s-eye.

This should not come as any great surprise. Fund investors do study performance attribution, the volatility of returns and the correlation between different strategies when constructing a portfolio of various managers. However, without knowing what is in the current portfolio it is difficult to judge what risks may lie ahead. The strategy may have changed materially from that used in the past, or the market conditions may have changed. Only position-based risk reporting can tell us if the strategy changed or if the market conditions that affect the strategy have changed.

The chart to the right depicts the typical presentation of one-day RiskMetrics VAR forecasts (90% two-tail confidence interval) along with the daily P&L of a sample portfolio consisting of 215 cashflows that include foreign exchange (22), money market deposits (22), zero coupon government bonds (121), equities (12) and commodities (33). It demonstrates how, by seeing what’s in a portfolio, we are able to asses volatility movements.

Consider the behaviour of each player in a poker game. Most players will occasionally experience a winning streak where they are dealt one great hand of cards after another. At some point, the winning player appears invincible as the stack of chips grows higher and higher, with the player’s confidence growing with each pot they rake. A skilled player recognises what is happening, treating each new set of cards as separate events and betting accordingly. But the novice will let the excitement overtake their emotions, believing they can’t lose, and will continue making big bets even when they hold the weaker hand, with disastrous results.

It is obviously valuable to know what cards someone is betting with. When we know the odds of each hand, we can then judge if the player is making skilled bets and outplaying his opponent or simply getting lucky. While luck plays a role, skilled players are the consistent winners.

The same balance of luck against judgment applies to asset management. How do you know if a manager is on a lucky streak or making a series of well-calculated bets?

There is a big difference from a risk perspective regardless of whether the strategy has been historically profitable. Looking at the actual bets being made, rather than the results, tells us a great deal about the risks being taken by a fund manager, and from that we can make a judgment as to whether or not they are making good decisions or just getting lucky, or perhaps unlucky.

Investors are obviously in tune with the need to look at a fund manager’s risk, and have taken steps to get a grip on what risks are inherent in each portfolio they invest in, hence all the buzz about transparency. But for the most part, results have been mixed.

In theory, one could know everything about the bets a risk manager is taking by examining the positions held in the fund - much like looking at a player’s cards. Some investors go as far as demanding full disclosure of the actual holdings from their managers. Unfortunately, getting the positions is only part of the battle. Many hedge fund strategies have complex portfolios that are nearly impossible to analyse without the use of sophisticated risk management applications that require very expensive data sets. Since most investors either do not have such a risk system, let alone a method for importing the terms and conditions of complex OTC derivatives, position transparency falls well short of the goal of meaningful risk analysis, and may even provide a false sense of security.

Internal approach

Another solution favoured by fund managers has been to provide investors with summary level exposure and sensitivity analysis on their portfolio using their own internal system. Although this allows managers to avoid disclosing actual holdings, it is almost impossible to make apples to apples comparisons of any two funds, because risk statistics are rarely consistent. To get some understanding of how difficult this is, consider the ‘exposure’ of an equity versus an equity option versus an equity future. For the equity we need the market value, for the equity option we need the delta equivalent and for the future we need the underlying market value of the futures contracts. This gets even more complicated as we move into swaps, swaptions and cross-currency swaps, but you get the point. Since there is no standard method of reporting exposure in the hedge fund industry, investors are left to interpret and make relative comparisons of inconsistent methods - a daunting task.

Third-party aggregation is an attractive solution to the risk transparency challenge for three reasons. First and foremost, it provides a fair and consistent method for comparing risk statistics including exposure, sensitivities, value-at-risk and stress testing. While we may not all agree on the methodology, at least we know what it is, and this gets us much closer to standardising such an important part of the investment process. With a risk report in hand, the investor can begin to assess the risk each manager is taking and the combined risk of many different strategies, which has been all but impossible without third-party aggregation.

Second, this is an economically attractive solution because it is scalable. A single dataset, security master database and analytics engine can be used for an unlimited number of investors and funds.

Finally, the funds do not sacrifice any competitive advantage that might be harmed by disclosing their positions to every investor. The aggregator can keep position-level information confidential while disclosing more meaningful summary information to the investor.

While this type of reporting service is in the early stages, it doesn’t take much imagination to see how valuable it will become over time. Rather than wait for the performance to suffer, investors can now detect when historically uncorrelated strategies are suddenly highly correlated - and therefore likely to experience losses under the same conditions - undermining all the work that went into creating a diversified portfolio of managers. This can happen even when managers stay true to their disciplined strategies because correlations between underlying risk factors change over time.

Analysing past returns simply cannot provide this kind of early warning system. Some investors are taking this quite seriously and starting to insist on getting risk reports before making an allocation of capital. And that puts to rest the notion that risk reports are useless because the lock-up period prevents investors from getting out of the fund. Fund managers are also beginning to recognise the opportunity available by addressing investor concerns about risk.

John Matwey is manager of HedgePlatform, a third-party reporting service provided by New York-based risk analytics firm, RiskMetrics Group.

* The results of this research were reported in a presentation entitled ‘Risk By Any Other Name’, made by David Gordon, CFA, Glenwood Capital Investments on October 16, 2003