Investors who used portable alpha experienced disappointment with its implementation during the 2008 market downturn. In the US, the strategy’s usefulness came under question following steep losses by the Pennsylvania State Employees Retirement System (SERS), which had committed 23% of its portfolio to portable alpha at the end of 2007. In August 2009, the Massachusetts Pension Reserves Investment Management voted to end the pension fund’s portable alpha program. For the 12 months through June 2009, the program was down 46%. Despite these setbacks, proponents of portable alpha point out that it has delivered above market returns over the long run. Pennsylvania’s SERS reportedly lost $1.5bn using the strategy, however, even when accounting for last year’s losses, portable alpha produced $317m in above market returns since 2002 when it adopted the program, notes Robert Gentzel, SERS Director of Policy. SERS eliminated its portable alpha program earlier this year.

In the last nine months investors have started to ask whether the losses suffered by SERS amount to an indictment of the portable alpha architecture. Before answering the question, it may be useful to define the investment strategy which aims to produce above market returns. One of the fundamental principles of portable alpha is that active managers produce two types of returns. The first is beta, which comes from broad market movements. The second is alpha, which comes from the portfolio manager’s active bets. Portable alpha is a way to separate the active return from the market return of the portfolio. By isolating alpha from beta, the excess return can be enhanced while the beta return can be neutralised. That is the basic theoretical idea.

In practice, a portable alpha strategy may work in a number of ways. In one scenario, a pension fund selects the S&P 500 as its benchmark. It then invests in the index using passive managers in order to gain market or beta exposure. Separately, the pension fund hires an active small cap manager benchmarked to the Russell 2000 small stock index. The pension fund then tries to remove its market exposure to the Russell 2000 index by shorting Russell 2000 stock index futures. What is left from the combination of the small cap portfolio and the short futures is the alpha return. By removing the Russell 2000 index, the pension fund is left with a return on the S&P in addition to the alpha generated by the active manager on the Russell 2000 Index.

Portable alpha can also be used as part of an overlay strategy. In this case, the pension fund invests in an alternative asset class such as a hedge fund or FOHF with little correlation to the broader market. The pension fund also buys equity index futures such as the S&P 500. The investment in the alternative assets is supposed to generate an alpha return uncorrelated to the index, and very little or no beta return. The risks inherent with both of these approaches are that the investor is exposed to the skill of the active manager and his ability to produce sufficient de-correlation with the broader market. In the case of SERS, a number of its active strategies were closely correlated with the S&P 500, offering lower than expected de-correlation and resulting in larger than expected losses.

Short-extension during downturn

Last year also saw a number of other active strategies such as market neutral and short-extension, which use the tools of alpha/beta separation, failing to provide sufficient de-correlation from broader indexes. There were, however, positive exceptions, which highlight the importance of active manager selection when implementing a portable alpha strategy. As of June 2009, Gottex market neutral was up 0.50% for the year. While it posted a 20.85% loss in 2008, its performance was stronger than global and developed equity indexes such as the S&P 500 which plunged 38.5%. Gottex’s performance was in line with broad hedge fund indices. A portable alpha program with exposure to Highbridge Statistical Market Neutral, which buys shares in companies and also sells them short, would have produced alpha separate from the market’s beta returns. Highbridge had a total return of 10% in 2008 and an expense ratio of 1.95%. Such cases of extreme out-performance were rare in 2008 and absent from the portable alpha programs at SERS and the Massachusetts Pension Reserves.

In December 2008, the portable alpha strategy at SERS was trailing the market by about 15% behind the S&P 500 partly due to the under-performance of its actively-managed hedge fund strategies which are part of its portable alpha program. “Many hedge fund strategies, including FOHF used in this particular portfolio, seemed to mimic equity markets in the third quarter of 2008,” notes AllAboutAlpha. “They seemed to have contained a hidden market exposure that was only revealed in times of distress. The result was that SERS seemed to have an excess exposure to markets: 100% exposure via swaps and an additional x % exposure via the supposedly uncorrelated FOHFs.”

The implementation of a portable alpha program also presents the challenge of selecting the right beta universe. “The problem is fundamentally one of defining what the beta should be for a particular manager’s strategy,” note Robert Whitelaw, Salvatore Bruno and Anthony Davidow, in an essay titled “Alpha/Beta Separation,” which appeared in the Journal of Indexes in April 2009. “If a manager claims his benchmark is the S&P 500, it would be convenient to simply call the S&P 500 Index return his beta and consider everything else alpha. But if in fact that manager is consistently selecting from, say, the Russell 3000, then extracting the appropriate market benchmark is problematic.”

There are a number of risks investors need to be aware of when adopting a portable alpha strategy, among them alpha volatility, failure to diversify alpha sources, presence of beta within the alpha source, insufficient liquidity to feed futures contracts, active management risk, counterparty risk, the use derivatives/leverage to hedge out market risk, and the impact of fat tails on alpha strategies. Key decisions must be made whether to hire separate alpha and beta managers and how high to set the alpha target.

Alpha/Beta separation

Despite last year’s setbacks, portable alpha has a number of proponents. Sweden’s AP7 is a leading advocate of the approach. A national buffer fund, AP7 uses a holistic approach to alpha/beta separation, awarding alpha briefs in recent years that are essentially notional overlays applied to the fund’s passive portfolio. During last year’s downturn, the fund’s traditional long-only mandates underperformed their pure alpha briefs, notes Svante Linder, Head of Fund Administration at AP7. “Four out of our six portable alpha mandates had positive returns, while five out of six long-only mandates underperformed their benchmarks. “The main advantage of alpha/beta separation is better performance,” says Linder. “We also have a better overview and understanding of what parts of the overall performance comes from the strategic asset allocation (beta) and what parts come from active management.”

AP7’s aim is to establish a framework to assess how market returns (beta) and attempts to add value (alpha) contribute to returns. “Over time, we expect investors to push active managers to adopt more transparent and cost efficient alpha-beta separation,” notes Richard Grötheim, CIO of AP7 and co-author of Alpha-Beta Separation: From Theory to Practice. “Attempts to add value can be achieved in the form of pure alpha strategies where manager’s positions reflect only relative value insights. For example, an active manager specialising in stock selection might hold a collection of long and short positions that reflect their valuation insights, but represent no net market exposure.” The board of AP7 decided to pursue a strategy of alpha/beta separation in 2005.  “The main change was a complete specialisation approach in the daily investment operation. Beta specialists were already in place and the new regime focused on finding pure alpha specialists.”

Born at PIMCO

Portable alpha was conceived more than two decades ago at PIMCO, a unit of German insurer Allianz SE, during a time of relative market stability. Although PIMCO is widely known as a fixed-income manager, it was interested in developing a strategy “presumably because of the potential to both increase returns and improve diversification. In certain cases, portable alpha strategies may even result in a simultaneous increase in expected return and decrease in downside risk relative to more traditional investment strategies,” notes the company. This did not prove to be the case in 2008 when a PIMCO stock mutual fund using portable alpha trailed the market by 10%. In a report titled “Birth, Death, Rebirth of Portable Alpha,” Dr. Rob Brown cites several shortcomings the strategy faced last year. Its implementation entails several key elements which Brown describes as the hiring of active managers (the alpha generators); the evaluation of each manager’s market exposures (betas); defining the structure of each manager’s liquidity needs and structuring the technical terms of derivatives contracts. “Some portable alpha adopters experienced disappointment in 2008 because they had implementation gaps in one or more of these four areas…Some misunderstood how positively correlated their different alpha sources were,” he notes.

Despite these challenges, the death of portable alpha is exaggerated. “Portable alpha came into existence because a new organisational structure was needed to make better use of alpha opportunities and to a somewhat lesser extent to take advantage of derivative’s advantages over physicals,” says Brown. “We believe that forward-looking alpha opportunities are likely to be the largest in our professional careers. Moreover, the high and long-lasting uncertainty that has been injected into the global capital market and macro economy establishes an environment where the rewards to future betas are more uncertain than the historic norm. This coincidence of factors significantly increases the need for optimal alpha capture and lower cost beta delivery.” Pension funds may be taking heed: a bfinance survey in March 2009 showed a somewhat higher percentage of pension funds who expect an increase in their portable alpha allocation in the coming years compared to the results of a similar survey six months earlier. Not dead yet, but for now, the strategy remains in intensive care.

VB