Investment
Finding alpha
BY  |  

Seasoned investors understand that returns are a function of both beta and alpha. Beta being the component of returns derived from the benchmark return and alpha being returns generated over and above (or below) the benchmark return. The difficulty with generating alpha is that it takes more than just manager skill. It also requires a manager to recognise that alpha is in limited supply and where a manager has a genuine edge (skill) they need to give careful consideration to the level of funds the strategy might be able to accommodate before returns become diluted (capacity).

In addition, managers looking to generate meaningful alpha must design portfolios that maximise the opportunity set by creating exposures that are materially different to that of the benchmark (active share).

In an environment where both bond and equity market valuations are extended, pure 'beta' exposure may leave portfolios vulnerable to market corrections. Traditional asset allocation strategies relying on negative correlation between bonds and equities may no longer offer the same diversification benefits that they have historically. In this environment, alternative strategies designed to preserve capital in the event of a market downturn and offering low correlation to traditional asset classes have the potential to provide meaningful diversification benefits to investor portfolios.

This paper discusses the intersection of capacity, active share and manager skill in efforts to exploit market inefficiencies in generating alpha, and how an uncorrelated strategy can provide portfolio diversification and capital preservation in the event of a market downturn.

Capacity

Active managers need to carefully consider their ability to generate alpha and how this potential is eroded as fund size increases. Perversely, managers are typically rewarded on funds under management and not always on excess return, implying manager incentives are not necessarily always aligned with investors.

Capacity is the amount of assets under management invested in an active strategy at which it is no longer possible to make additional investments that generate marginal alpha in excess of a minimum threshold (Vangelisti, M 2006, 'The capacity of an equity strategy', The Journal of Portfolio Management, winter 2016).

Chen, Hong, Huang and Kubik ('Does fund size erode mutual fund performance? The role of liquidity and organisation', The American Economic Review, December 2005) demonstrated that fund returns are inversely correlated to fund size. This prevalence was more acute with small-cap-focused funds, indicating that the relationship may be driven by liquidity constraints. Trading in a limited number of names becomes more difficult due to liquidity constraints and market impact.

O'Neill, Schmid and Warren ('Capacity analysis for equity funds', The Journal of Portfolio Management, spring 2018) demonstrated that capacity is also determined by the number of opportunities available; the market segments from which opportunities are sourced; the cost of executing trades; and any constraints on stock holdings or ability to participate in trades.

As a strategy's size increases, managers are forced to either diversify the portfolio into a larger number of smaller positions or limit capacity to retain any edge they might have in stock selection. Managers that demonstrate real skill and achieve excess returns will typically experience increased net inflows. This, in turn, increases fund size and eventually erodes any edge the manager may have exhibited. With smaller-company managers, the liquidity constraints are more acute and capacity becomes an issue at lower fund size.

The trade-off between capacity and the ability to generate alpha is clear. Different approaches to this trade-off lead to the separation of managers into the 'alpha hunters' and 'beta grazers' so eloquently described by Martin Leibowitz ('Alpha hunters and beta grazers', Financial Analysts Journal, September/October 2005).

Managers that primarily seek to accumulate assets will by definition produce portfolios that increasingly resemble the benchmark. They will therefore tend to produce returns that are reflective of the benchmark return (beta grazers). Of course, this is before costs and fees. The more assets under management, the harder it becomes for the manager to provide investors with an outcome that betters the benchmark after fees and costs. This is evidenced in the S&P Dow Jones research SPIVA Australian Scorecard Mid-Year 2019, where managers show a consistency in underperformance across multiple timeframes.

There will also be reputational and career risk aspects to consider. As fund size increases, managers may become more risk averse, avoiding portfolios that differ significantly from a benchmark. Agreed, a manager might underperform a benchmark after fees and costs, but a little bit of underperformance - consistent with peers - may be a more palatable outcome for the manager, notwithstanding the interests of the investor.

Investors are therefore faced with a conundrum.

Managers that demonstrate real skill attract fund flow, thus increasing funds under management and eroding their edge if capacity is not constrained. Large-cap funds tend to have unconstrained capacity and show an inability to exceed performance benchmarks over the long term. Small-cap managers, where capacity is frequently constrained due to liquidity, demonstrate a better ability to exceed performance benchmarks over the long term.

Active share

The ability to generate alpha requires a manager to construct a portfolio that varies meaningfully from that of a benchmark or index. The greater the variation in relative weights, the greater the expected variance in returns relative to the benchmark.

Active share (Cremers, M & Petajisto, A 2006, 'How active is your fund manager?' A new measure that predicts performance', The Review of Financial Studies, September 2009) is a measure of the percentage of stock holdings in a portfolio that differs from that of the benchmark. Portfolios that hold a limited number of concentrated exposures and where those weights differ from index weight, typically have a high active share.

Alternatively, portfolios that hold a large number of highly diversified positions that resemble benchmark weights typically have low active share. For example, an index fund, designed to replicate the index exactly, would have an active share of zero.

Importantly, active share also provides an indication of whether a manager is a closet 'index hugger' or a true-to-label active stock picker.

Low active share indicates a manager's stock position sizes fail to deviate significantly from the benchmark. As an investor, one would question why one pays active fees for a portfolio that has a high likelihood of producing index-like returns, and on an after-fees and after-costs basis is likely to underperform a benchmark.

An active stock picker typically has a high active share, indicating what is likely to be a more concentrated portfolio with position sizes that deviate substantially from benchmark weights. As a result, the return outcomes have a greater chance of being significantly different from that of the index, and subject to the manager's ability to generate alpha, justify higher fees.

Patrick O'Shaughnessy ('Alpha or assets', The Investor's Field Guide, April 2016) demonstrated the importance of active share in alpha generation. The premise was that active strategies should 'build for alpha, not scale'. This is not typically reflected in the funds management industry where large managers dominate. This tendency is evident in the local funds management industry where large providers of product dominate the industry, encompassing both active, index and smart beta strategies.

O'Shaughnessy's study is reproduced using Australian data.

As an exercise, let us assume as an investor we had perfect foresight on stock returns for the next 12 months and built portfolios to achieve the best forward annual return at various levels of active share. We have used the S&P/ASX 200 index as the basis for the study with data from June 2000 to September 2019. To avoid an exposure to just a handful of positions that have the potential to skew outcomes, we have limited the exposure to any single position to 5%, unless the stock has an index weight greater than 5%, in which case its weight is limited to no more than its index weight.

 

Link to something 0jjw1Rj4