Absolute return is seen as the holy grail of asset management: pure performance without market risk. Theoretically, any source of outperformance can be turned into a source of absolute return by shorting the benchmark. In practice, however, it is not that simple because risk can come in many forms, not just market beta. So how can an approach featuring smart beta provide ‘purer’ and more liquid alpha that is less subject to drawdowns and can offer real portfolio diversification?
When interest rates are low, as they are now, investors tend to chase absolute return in the knowledge that relative return can only be expected to be modest, at best. The usual measure of efficiency for absolute-return strategies is the Sharpe ratio which reflects the excess returns over risk-free rates (in practice, the excess you can earn over money market fund returns) divided by the volatility of those excess returns. For example, a Sharpe ratio of 0.5 means that a fund manages to generate half as much performance as volatility on average over time.
It’s smart to opt for uncorrelated assets
This gives a first hint on how to improve an absolute-return strategy because, while return is notoriously difficult to predict, this is easier for portfolio volatility – as we have shown in our academic papers here and here. Since Markowitz, we have known that volatility can be reduced by diversification. If two strategies have the same Sharpe ratio, then the combination of the two must necessarily have a higher Sharpe ratio that will rise as the correlation between the returns from these two strategies falls. If the correlation falls far enough to be negative, then the overall Sharpe ratio is even higher as one strategy typically works when the other has a tendency to fail. This is why absolute-return strategies usually involve multi-asset or multi-strategy portfolios and are based on wide universes of different securities.
However, most investors are less sensitive to volatility per se than they are to ‘downside volatility’, or drawdowns. Drawdowns reflect the depth of the drop from the highest valuation of the fund and the ‘maximum drawdown’ (jargon for the largest drawdown) reflects what you would have lost had you been particularly unlucky in timing your investment in a given fund over a given timespan.
Balancing drawdown risk and returns
An alternative measure of the quality of absolute-return funds is the Managed Account Ratio (MAR) which equals the annual rate of return of the investment divided by its maximum drawdown, observed historically. The inverse of the MAR (1/MAR) can be expressed as duration: it measures how long it takes to recover from the worst drawdown while earning the expected annual rate of return.
For instance, investing in global equities (represented by the MSCI World index) from January 2003 to June 2017 had a 1/MAR of almost eight years. In our view, this is quite poor, but it is obviously because the global financial crisis of 2008 created a drawdown of 52% for an average annual rate of return over cash in those roughly 13 years of 6.9%. This indicates that the equity premium is a poor source of absolute return. By contrast, non-risky assets have a 1/MAR that is close to one day, but at the cost of very low returns. In between, reaching a worst drawdown of only one year of average return is a classical target.
To the rescue: smart beta (aka factor investing)
This is where smart-beta investing can help, in particular its more sophisticated form known as factor investing. This is because the objective of factor investing is to change the allocation in a portfolio so as to add targeted exposures to systematic factors that not only represent various styles, but also are persistently remunerated over the medium to long term within a given asset class. The excess returns generated by the changes in allocation of securities in a portfolio that is constructed to target these factor exposures can be explained by different cognitive biases that underlie the statistical mispricing of securities.
Such factors are typically value, momentum, quality, low volatility or, within fixed income, carry. Factor investing is usually associated with long-only strategies, but it can be turned into an absolute-return approach by shorting the exposure to the underlying benchmark. These factors can generate excess returns that show low correlation with each other, in different market conditions or with their counterparts in other asset classes. They are typically complementary in a portfolio.
Exhibit 1: illustrates the very low correlations for purified (beta-neutral, sector and region-neutral, and target volatility) long-short factors in the main asset classes: equities, government bonds and foreign exchange (weekly data from January 2003 to December 2016):
Source: THEAM, BNP Paribas Asset Management, as of 31/07/2017
Of course, to turn these theoretical factors into actual strategies, an investor must ensure that the beta of the strategy is neutralised, in each asset class and on currency risk, because being exposed to currency risk could destroy most of the benefit of diversification. However, a portfolio of multiple, diversified factor-based alpha sources can generate a good MAR, i.e. an attractive average return over the investment period with a limited drawdown.
Written on 01/09/2017
Smart beta – defines a set of investment strategies that emphasise the use of alternative index construction rules to those employed in constructing traditional indices based on market capitalisation
Factor investing – an investment strategy in which the selection of securities is based on characteristics that are associated with higher returns
Absolute return – the return that an asset achieves over a certain period of time