We need to talk about capture ratios

Someone once said that the first rule of investment is “don’t lose money”; the second rule is “don’t forget the first rule”. This may sound facile, yet it contains a profound truth; one that is central to how investors should think about risk and assessing the relative performance of portfolio managers over time. 

Of course, in addition to preserving client capital, successful fund management is about generating outperformance. The best portfolio managers are able to make more than their peers in strong market conditions and lose less money when markets dip. 

There is a simple, but effective, tool to measure this. Capture ratio separates a historic return series into two buckets - months when the market went up and months when the market went down - and looks at how an investment performed in each. Taking the average of the two populations results in the investment’s capture of the upside, downside capture, and dividing one by the other gives the overall capture ratio. 

In recent years, however, capture ratio has often been overlooked. This may be due in part to a longstanding obsession within the investment industry for metrics that narrowly equate risk with volatility, such as Sharpe ratio, or the extent to which a fund outperforms without deviating from its benchmark, the information ratio.  

The latter entails using tracking error, which has particular limitations. Being sufficiently different from the index does not have to be an ‘error’ or something fund managers and their clients should be afraid of. If you want a better result than the index, you have to do something different. It is surely better that a manager invests with conviction on their assessment of risk and reward rather than allowing an outside index to dictate portfolio construction decisions. 

The other possible explanation why capture ratios have been underused is that stock markets have generally enjoyed a sustained period of strong returns in the last decade. So, while it has been easy enough to see who has made hay while the sun shone, there hasn’t been a persistent down period on which to judge a manager’s ability to preserve capital. 

Years of strong positive market returns have meant funds with high upside captures do well, but people have perhaps forgotten the high downside aspects this often comes with. The same outperformance number can be produced with wildly different capture ratios. 

Arithmetic versus geometric returns

The reason lies in the distinction between the calculation of arithmetic average versus geometric average returns and the fact aggregate portfolio returns are compounded over time. 

Imagine a coin flip where each outcome is equally likely, but heads pays 60 per cent, while tails loses 50 per cent. This sounds like an attractive bet since the expected return is the arithmetic average of five per cent. Indeed, some active fund managers would describe this return profile as asymmetric, offering greater upside than downside. But let’s roll the exercise forwards: a positive 60 per cent return, followed by a negative 50 per cent loss, followed by a positive 60 per cent gain, then a negative 50 per cent loss. Despite the arithmetic average remaining at five per cent, the geometric average multiplies each outcome such that our errant coin flipper has lost 36 per cent of their starting capital.

It is hard to argue with maths. What was presented as a seemingly attractive bet overlooks the fact that a 100 per cent return is required to return to break even after being down 50 per cent. When assessing a prospective investment, it is not enough to merely quantify the upside because the mathematics of compounding returns make it difficult to recover from catastrophic losses. 

This logic also holds true for a portfolio. An investor who receives 16 per cent annually for a decade ends up better off than an investor who earns 20 per cent a year for nine years and then loses 15 per cent in the tenth year, as Figure 1 shows. Again, the mathematics of geometric compounding are responsible for the deleterious effect of one bad year. This is what makes capture ratio a suitable metric for assessing relative performance on a risk-adjusted basis.

Figure 1: Why the journey can be just as important as the destination

The COVID-19 crash in March (which unwound five years of market returns in a heartbeat) gave us a brief, but important, reminder of how fragile equity markets can be: stairs up, elevator down. The next sell off in markets could be more persistent and painful: it only takes one bad year to undo years of outperformance and reveal that some managers have been carrying a lot more risk than they or their clients appreciate. 

Paying full attention to how a fund performs during negative periods for the market can show otherwise positive returns in a different light. For example, imagine a market that returns 16 per cent for nine years and loses 16 per cent in the tenth year. One fund unexcitingly performs in-line with the market during the up years but falls by half as much in the final year. 

Overall, it beats the market by one per cent annualised with a capture ratio of 2.00. Another fund excitingly beats the market by four per cent during the up years but falls by twice as much in the down year. While it has outperformed by a similar amount, the far inferior capture ratio of 0.63 speaks to the very different levels of risk taken to achieve these returns. 

Downside risk - the impact of loss - effectively hurts more than the equivalent percentage increase helps a portfolio’s return. Going back to our original example, you could argue the asymmetry in outcomes now works the other way around: played out through time, which all investments are, it can pay to place more weight on the downside than the upside. If you fail to lose money, then all your other outcomes are probably good. 

This is why we should “never forget the first rule”, and why focusing on capture ratios is such an important tool when assessing manager skill. No metric will ever be perfect, but looking beyond raw performance numbers and ascertaining whether a manager captures a decent amount of upside while protecting well on the downside is vital for assessing whether an investment manager has sufficient defensive qualities to take advantage of the mathematic quirks explained above.