It is a well-established statistical fact that a collection of unknowns is less variable than any one of its constituents. As more stocks are added to a portfolio, the total return tends towards the weighted average return of all stocks and the volatility of outcomes reduces, eventually reaching that of the market overall. The same applies when selecting investment managers: the expected variability of returns from a combination of carefully selected investment managers is likely to be far lower than hiring a single manager.

A common dilemma in a multi-manager portfolio is how many managers to appoint. Fewer managers may result in higher returns (if chosen well) but also higher volatility because not many managers can maintain high returns all of the time. In contrast, hiring too many managers may result in a combined portfolio that essentially resembles the market overall, resulting in benchmark-like performance. 

The benefit of diversification really depends on the distribution of returns of investment managers, and the relationship between them. The benefit of adding a third manager will be reduced if the first two managers are already well diversified. In contrast, if managers have significant volatilities and outsized returns (both positive and negative) then there should be a greater benefit from having multiple managers especially if they have different drivers of outperformance.

Many investment managers in emerging markets tend to run concentrated portfolios. Our intuition led us to expect that the benefit of combining multiple managers should be far higher in China than in Europe.

Taking the available manager returns in eVestment, a large manager database, for China A specialists and Europe ex-UK portfolios over the period Jan 2014 – Dec 2019, we noted that:

·       There were 83 investment products in the China A space that had a track record covering the sample period

·       There were 66 products in the Europe ex UK space

·       The mean outperformance for China A managers was 0.5% per month

·       The mean outperformance for Europe ex-UK managers was 0.1% per month

·       The volatility of a typical manager’s return stream as measured by the standard deviation was 3.3% for China A vs 1.5% for Europe ex UK.

·       Perhaps not surprisingly there was a far greater occurrence of extreme returns in China than in Europe.

Source: eVestment, RisCura calculations

While it was nice to note the greater outperformance in China A, the substantially higher volatility was less desirable. If investors want to avoid significant underperformance over shorter time periods then it is clear from the graph that they should hire more managers in China than in Europe. The entire European manager universe rarely underperformed by more than 3% in any one month. Therefore, an investor could have experienced acceptable levels of volatility by hiring two or three managers, but not necessarily in China.

A counterthought may arise where an allocator considers risk reduction to be useful but not overriding in the investment objectives. In particular, an allocator may feel they have the skill to select the best managers in each region and enjoy the superior returns these should deliver. Why bother with risk reduction or risk-adjusted performance if one can pick outright winners?

To investigate this, we assumed that we were blessed with perfect foresight and repeated the exercise with the ten best performing managers over this period. The result was remarkably similar. As expected, the mean return for these subgroups was higher than the full population (they were the best managers after all) but their volatility was much the same.

Source: eVestment, RisCura calculations

From the graph above we can see that ‘talented’ European managers reduced bad results and were far better at delivering moderate levels of outperformance consistently. Based on these results, an investor could have experienced a decent result by selecting a single manager if they had perfect foresight. However, this is not the case for the Chinese managers. There, top managers had both extremely negative and positive relative returns. They delivered outperformance by taking much higher levels of risk. Therefore, multiple managers (with different return drivers) were still required in China to manage consistency of relative returns even with perfect foresight. 

This makes perfect sense to us. The number of companies in China is only second to the USA, information is less reliable and research is more time intensive. Therefore, the most talented mangers tend to specialise in specific niches, are concentrated and biased to their areas of competence. They are expected to have higher volatility of outperformance.

Our conclusion from this is that a risk-aware investor needs many more managers for Chinese equities than for a developed market like continental Europe all else equal. This remains the case even if the investor has perfect ability to select skilful managers. In a region like China where outperformance is volatile there is substantial risk reduction benefit by selecting several managers instead of only one or two. In fact, one may as well allocate to all 10 of them.

Lars Hagenbuch, Consultant, RisCura