Eric Kirzner – Professor of Finance, Rotman School of Management, University of Toronto

The polar extremes of investment strategy are index investing and absolute return approaches.

Index investing is based on the view that the market for securities is efficient. The resulting strategy attempts to match the performance of an underlying index. The opposite is absolute return investing, in which the returns are not tied to a specific benchmark but are instead expected to be positive most of the time. Losses are assumed to be rare, associated with unexpected occurrences or unusual events.

The root of absolute return investing is in the partitioning of risk. A basic paradigm of investment finance is that in a diversified portfolio, the unique risk associated with a security is diversified away, leaving the systematic or economy-wide risk associated with a market index. For years, most took this road.

The alternative path of portfolio building was set out in 1949 by Alfred Jones, who described how to isolate the market exposure of individual securities and eliminate the associated market risk by buying securities and shorting the market in specified proportions. This was the genesis of the hedge fund and the broader absolute return fund.

There are a host of investment strategies employed under the name of absolute return funds (ARF). Most funds are highly specialized, their objectives hard to determine. We can view the ARF universe on a spectrum ranging from zero or near-zero market exposure to considerable market exposure.

At the extreme left are funds that are engineered precisely as market-neutral funds. Using delta hedging and other analytic techniques, the funds performance is expected to bear absolutely no correlation to bond and stock markets. Market-neutral funds typically invest in both long and short equity portfolios, usually in the same sectors of the market, and by pairing up positions.

In the next stage of ARFs, the funds take hedged positions but are only partially market-neutral. These would include ARFs that have long/short positions but have a bias in a particular direction (usually bullish). This category also includes event-driven isolation hedge funds, focused on special situations, and takeover arbitrage funds, which buy target firms and short-sell takeover firms. On the right side of the ARF spectrum are those funds that take directional positions in the market. These tend to be opportunistic or directional, and may use momentum or trend-based strategies. However, they may have relatively low correlation with conventional bonds and stocks.

Useful in understanding ARFs is the distinction between convergent and divergent traders. Convergent traders believe the world is stable and knowable, and that they can apply suitable yardsticks to value companies. As such, their trading decisions are based on finding deviations from assessed values and exploiting such anomalies. Such traders will fit squarely into the non-directional camp, will buy and sell believing in mean reversion and looking to avoid fat tail events. Convergent traders are theoretically harmonized to non-directional market- neutral funds.

Divergent traders believe that markets are not always stable, that determining fair value for securities is difficult, and that trends exist as markets adjust or learn from new information. They look to profit from fat tail or unusual events and are congruent with momentum or directional traders.

ARFs should be defensive, if they are structured properly. Given the long/short positions in stocks, they should underperform in a strong equity market since the shorts (even if the fund manager has done a good job of finding overvalued ones) will increase in value due to the systematic pull of the market. In a down market, a hedge fund should outperform since the short positions should at least offset the losses on the long positions. Also, the variability of the fund, given the long/short positions, should be less than a market index.

ARFs are an asset class with specific expected absolute returns and standard deviations and low or zero correlation with conventional asset classes. *