Having spent a fair few hours looking at fund performance figures myself, I am always interested in the ways that professional fund pickers go about analysing the data. It is true that in the past some IFAs have not bothered much with doing their own research, preferring in at least one notorious case simply to auction places on their best buy list to the provider offering the highest commission rates.
These days, thank goodness, such chicanery is harder to get away with and virtually every intermediary who recommends funds has at least to go through the motions of screening the available universe. Some remarkably sophisticated packages that analyse fund behaviour in extraordinary depth are now available to make the life of the professional analyst easier.
The best firms of course carry out a thorough exercise themselves, believing that, as Richard Timberlake of Investment Manager Selection puts it, picking good funds is one of the few areas in investment where you can still find significant market inefficiency. (As the man who recruited Anthony Bolton to Fidelity, and later founded Fund Research, the first fund ratings service in the UK, Timberlake can claim to know of what he speaks.)
Here is how one financial intermediary goes about the job. Tim Cockerill, head of research at Rowan & Co Capital Management, looks for funds that display the greatest consistency in producing superior risk- adjusted returns.
While raw cumulative performance figures dominate most fund advertisements, such figures are notoriously unreliable. A high-risk fund may be temporarily top of the performance tables if it has four lousy years in a row followed by one spectacular success. Such a fund makes for a more dubious investment than a fund that grinds out small but positive risk-adjusted returns more often than not.
How Cockerill sets about avoiding these traps is by ranking funds in each of the main sectors, as defined by the Investment Management Association (IMA), on six different measures.
The fund's absolute return counts as one of the six, but other measures include the fund's volatility, its Sharpe Ratio (a standard measure of risk-adjusted returns) and its performance in the three most recent 12-month periods, with the most recent period given a slightly heavier weighting.
These scores are then combined to give an overall ranking for the fund's performance over three years. A similar score is computed for more recent performance. The range of scores runs from six (the theoretical best) to 600 (the worst). The numbers are recalculated every month, with each fund being assigned a grade from A to F. The top five grades are given to funds that have a combined score between 0 and 50, and every other fund receives an F.
Cockerill emphasises that this methodology is only a tool to help in fund selection. It is not the be-all and end-all of the process. Its great merit is that it allows him not only to compare different funds in the same sectors, but to track trends within sectors. You can be certain, at the very least, that the methodology will weed out most of the dismal funds that clutter up the lower half of the fund tables.
The results are certainly interesting. Very few funds score an A. In the May 2006 half-yearly review, for example, not one fund had an A rating. There were 13 Bs out of 256 funds in the All Companies sector, one (out of 62) in Equity Income and one (out of 52) in Smaller Companies. Unlike some fund-rating systems, which award top ratings to a fixed proportion of funds, a system that does not require an outright winner gives an intuitively appealing result.
A second interesting feature is that, while there is clearly a strong correlation between the absolute performance numbers of a fund and its rating, there is nothing automatic about it. So, for example, 10 of the 50 funds with the best cumulative three-year performance figures in the All Companies sector score the lowest rating, an F.
Typically this will reflect that the fund's performance has been patchy and inconsistent. To take a specific instance, Cavendish Opportunities, ranked 17 out of 256 funds on absolute performance in the May 2006 review, received an F while the fund above it, Anthony Bolton's Fidelity Special Situations, achieved a B.
Cockerill says that tracking the direction of a fund's rating over time provides useful information about which funds are prospering or struggling in the current market environment.
At the moment, as I have noticed myself, a number of traditionally strong funds are finding the short-term going tough. This is often an indication that some kind of inflection point in the market is approaching.
More than half the funds in each sector score an F, as you would hope to find in any system that reflects the reality of the industry's overall aggregate performance. Cockerill insists that just because a fund has an F rating does not mean that it cannot be chosen if further investigation provides a plausible explanation for its lowly ranking. For example, a fund that has been struggling and subsequently appoints a highly regarded new manager might well offer an appropriate investment opportunity.
But such occasions will usually be rare. In the end, as he is the first to admit, statistics can only take you so far in analysing funds. The best funds will always have good ratings, but funds with good ratings are not necessarily ones you want to own. As I have found myself, nothing ultimately compensates for knowing the style, ability and character of an individual fund manager, and the integrity, motivation and efficiency of the firm that he works for.
One final notable feature of Rowan & Co's most recent screening exercise (data to the end of September) is that two funds have achieved an A+ rating, the only time in Cockerill's experience that such a thing has occurred. Both are run by the same individual, Neil Woodford of Invesco Perpetual. While his funds are now larger than conventional wisdom says is consistent with the ability to outperform, his long-run track record underlines the truth of the saying that the best fund managers break, not make, the rules.Reuse content