Part 2: Portfolio Monitoring

This blog will be considered Part 2 of a 3-Part series pertaining to Portfolio Monitoring. Please view the Part 1 blog.

Effective Performance Analysis Requires Qualitative Insights

The phrase stamped on every investment fund prospectus is something along the lines of: past performance is no indication of future performance. While a regulatory document is not the usual source for nuggets of investment wisdom, too many in the industry continue to behave in a way that does not reflect an awareness of this investment truism.

That said, one of the richest data sets available when evaluating any investment manager or strategy is historical performance (returns and risk). We cannot ignore historical returns; but nor can we find future top performers simply by assuming that historical outperformers will repeat. We assess investment managers or strategies by applying our Ten-Ps due diligence framework. Performance is one of the ten P-factors. This blog, the second instalment of a three-part series by my colleague Joe Modica, looks at how we use performance in the selection and monitoring of investment strategies and mandates.

Performance Persistence

The opening sentence touches on the notion of performance persistence. If past out- and under- performance persisted into the future, we could easily find future outperformers simply by looking at managers that outperformed in the past. But many academic studies have confirmed that this persistence does not exist. Examples include Mark Carhart’s 1997 paper titled “On Persistence in Mutual Fund Performance”. Carhart’s conclusions have since been validated by many studies, including by Morey and Blake (1999) and Sah, Tidwell, and Ziobrowsky (2010).

That said, a few studies have found that chronic underperformers of the past tend to remain cellar-dwellers in the future.

Past Performance

The assumption that strong past performance persists into the future is embedded in some widely-used industry practices – none of which we employ.

Many institutional investment policy statements include clauses stating that a manager will be placed under review (the investment consultant equivalent of sitting in the penalty box) if they underperform their benchmark for six consecutive quarters. Another common practice is to evaluate a manager’s performance based on four-year risk-adjusted returns. Often these metrics prompt the termination of an underperformer in favour of a manager that has outperformed over the same period.

These common industry practices are not rooted in empirical evidence and they all hinge on the existence of performance persistence. Moreover, such practices are likely to worsen performance over time. (See this 2017 blog post on firing underperformers.)

Performance Behaviour

Past performance can be meaningful if we can gain good insight into how it was achieved. This requires us to study the nature of past investment environments and gather intel on (for starters) the investment manager’s process, genesis, and history. While this is subjective, the resulting insights help to distinguish luck from skill; key to determining if we can expect future performance to resemble the past.

For example, sometimes a strategy boasts strong returns and benchmark outperformance in its early years but more middling performance in later years. This pattern would prompt us to dig into why this may have occurred; starting with looking for significant changes in assets managed, the possibility that early returns were a stroke of luck, or changes in the investment team. Key to this analysis is understanding how the manager thinks about and analyzes businesses – and how investment decisions are made.

Bull & Bear Market Performance

Rather than look at total returns over the past five or ten years, we examine performance in specific bull and bear market periods. Since most strategies have not been around long enough to experience more than a couple of bear markets – and this year is the first bond bear market since 1994 – this also requires us to apply some of the same ‘softer’ analysis as described above.

Extreme market moves – up or down – can serve as a very handy test. For example, the dot-com boom (late 1990s) and bust (2000-02) allowed us to quickly discern who paid attention to the prices paid for businesses and those that would buy growth at any price (GAAP crowd). Managers that underperformed in the late 1990s and outperformed in the early 2000s almost universally were those that would never chase trendy stocks.

The behaviour of markets over the past five years have served up a similar – albeit less extreme – period that will be useful in evaluating managers for some time. While this is part of what constitutes “past performance” it can yield quick but very telling insights because it offers insight into decision-making and valuation.

Whether speaking of performance alone or the other nine components of our due diligence process, the factors we analyze when initially selecting a manager are the same ones used to monitor managers after we hire them. This work happens behind the scenes, the highlights of which are reported to clients – the focus of Mr. Modica’s next instalment.

Please view the Part 1 blog.