The last thing your client wants when markets sell off is a group of funds that fall together. As a result, finding funds that do not move together is key to improving risk-adjusted returns.
Added to this, a lot of investors are not keen on the idea of paying active management charges for a fund that is highly dependent on the index, meaning that advisers need to be able to show the value their chosen active fund can provide.
Data from FE Analytics shows that some of the best-performing funds have a relatively low correlation to their stated benchmark.
Our data can reveal the correlation of a fund’s returns to an index over any timeframe, and produce a figure called r-squared which shows how much of a fund’s returns can be explained by the movements in that benchmark.
Essentially, a lower r-squared score means that the relationship between the fund’s returns and the index’s is weaker.
This can be to the positive or negative side of course, but it does show that managers are making their money elsewhere, either through good stockpicking or fishing in a different pool.
A low correlation could help a fund outperform if the market crashes, and could make it less susceptible to the events that cause markets to wobble.
Buying funds in the same area with a low correlation to the index is wiser than the opposite, as it should smooth the risks investors are exposed to and the volatility of their returns.
Data from FE Analytics shows that of the 67 funds in the IMA UK All Companies sector with an r-squared to their benchmark in the bottom quartile of significance over three years, 28 have also produced annualised returns in the top quartile over the period.
R-squared of funds to benchmark and annualised returns over 3yrs
| Name | r2 | Return |
|---|---|---|
| Unicorn - Free Spirit | 0.3 | 18.66 |
| MFM - Slater Growth | 0.47 | 16.48 |
| Unicorn - Outstanding British Companies | 0.55 | 17.91 |
| Cazenove - The Capital | 0.57 | 25.08 |
| Lindsell Train - CF Lindsell Train UK Equity | 0.66 | 21.6 |
| Baillie Gifford - UK Equity Alpha | 0.68 | 15.43 |
| Cavendish - Opportunities | 0.69 | 21.21 |
| Liontrust - Special Situations | 0.7 | 22.15 |
| SVM - UK Growth | 0.7 | 17.29 |
| Barclays - UK Lower Cap | 0.71 | 23.56 |
| Franklin - UK Managers' Focus | 0.72 | 18.95 |
| S&W - Revera UK Dynamic | 0.73 | 15.68 |
| Schroder - Recovery | 0.73 | 17.44 |
| Artemis - UK Growth | 0.74 | 16.57 |
| JOHCM - UK Growth | 0.74 | 15.45 |
| Liontrust - UK Growth | 0.74 | 15.79 |
| Majedie - UK Focus | 0.74 | 16.81 |
| Neptune - UK Mid Cap | 0.74 | 24.65 |
| Artemis - Capital | 0.75 | 15.84 |
| Ecclesiastical - Amity UK | 0.75 | 15.61 |
| Jupiter - UK Growth | 0.75 | 17.18 |
| Premier - Ethical | 0.75 | 15.75 |
| Cavendish - UK Select | 0.76 | 16.09 |
| Jupiter - UK Special Situations | 0.76 | 16.25 |
| Majedie - UK Equity | 0.76 | 16.83 |
| R&M - UK Equity Long Term Recovery | 0.76 | 16.32 |
| Stan Life Inv - UK Equity Unconstrained | 0.76 | 22.68 |
Source: FE Analytics
Unicorn Free Spirit, managed by FE Alpha Manager John McClure (pictured), has the lowest r-squared to its benchmark, of 0.3, and returns of 18.66 per cent, according to data from FE Analytics.
MFM Slater Growth has an r-squared of 0.47 and annualised returns of 16.48 per cent, while Unicorn Outstanding British Companies has an r-squared of 0.55 and returns of 17.91 per cent. However, the problem with using the fund’s stated benchmark is that managers can choose whatever index they want and may end up investing outside of their stated field.
MFM Slater Growth takes the IMA UK All Companies sector as its benchmark, which makes it uninteresting to look at the correlation. Replacing it with the FTSE All Share gives an even lower figure of 0.37.
One way to uncover whether the fund’s benchmark has been well-chosen, and to provide more information about where the manager has been making their money, is to run a regression analysis on the funds.
This calculates which market indices in which proportions best explain a fund’s returns.
For example, a regression analysis on Neptune UK Mid Cap, one of the funds in our list, tells us that the fund has made a lot of its money in the FTSE 250, unsurprising given the name.
Our analysis shows that a model of 24 per cent in the FTSE 100, 54 per cent in the FTSE 250 and 22 per cent in the FTSE Small Cap explains 80 per cent of the fund’s returns over the past three years.
The rest comes from stockpicking and whatever cash weighting the manager has held over the period.
Regression analysis is particularly useful when putting together a portfolio of funds as it allows investors to avoid picking two funds that are doing the same sort of thing.
Schroder Recovery, for example, also has a high dependence on the FTSE 250, although it takes the FTSE All Share as its benchmark.
A model with 4 per cent in the FTSE 100, 78 per cent in the FTSE 250 and 18 per cent in the FTSE Small Cap explains 0.79 per cent of the fund’s returns over the past three years, according to our data.
Although it is not clear from the factsheets, both Neptune UK Mid Cap and Schroder Recovery are both fishing in the same pool, albeit with different styles and strategies. The different strategies might justify holding both together, of course.
Our data shows that Schroder Recovery was actually underperforming its model until recently, but over the past 12 months has surged ahead of it as markets have risen.
An interesting balance for both these funds is provided by Liontrust Special Situations.
Our data shows that a model weighted 62 per cent to the FTSE 100 and 38 per cent to the FTSE Small Cap index explains 74 per cent of the fund’s returns.
For more of a small cap weighting, investors could consider Cavendish Opportunities.
Our data shows that a model weighted 16 per cent to the FTSE 100, 26 per cent to the FTSE 250 and 58 per cent to the FTSE Small Cap explains 77 per cent of the fund’s returns.
Regression analysis also helps to show the funds that make most of their returns from stockpicking.
The best model for a number of funds on our list provides only a relatively low r-squared score. Many of these are funds with a small cap bias, where managers are traditionally able to make better returns through stockpicking.
The best model for Unicorn Free Spirit is 16 per cent in the FTSE 100, 2 per cent in the FTSE 250 and 82 per cent in the FTSE Small Cap index, but that explains only 63 per cent of the fund’s returns. MFM Slater Growth is even less dependent on benchmark returns, according to our data.
A model with 92 per cent in the FTSE Small Cap index and 8 per cent in the FTSE 100 explains only 48 per cent of the fund’s returns.
However, our data also shows that the fund’s outperformance against this model has been steadily diminishing over the past couple of years.
