these are startling performances gaps between different structures, including funds of funds, investment trusts and life funds
This is the third in a series intended to quantify the great conundrum of the investment business - the greater the marketing success; the worse is subsequent investment performance.
The first installment demonstrated how badly the free advice industry delivers investment performance. It took ONS statistics for the balance sheet and cash flow statements of the UK"s unit trust sector as a good example because investment advice is given away for free when bundled with front-end loads for those products.
This data was used to show performance for the average unit trust investor, as opposed to the average unit price. It found that bad market timing cost investors 4.4% per annum in terms of under-performance against the FTSE index since 1986.
The second installment took the proposition that US mutual sales are an effective reverse indicator in predicting the stock market"s performance and applied that to UK unit trust launches. Using S&P Micropal"s funds database, this showed that changes in popularity for different types of equity mandate also proved to be an effective reverse indicator for asset allocation by British retail investors.
This third installment concerns itself with performance by different types of investment packaging. A quarter of a century ago they were sold by different fund management groups. However, nowadays most groups deliver most kinds of investment package. That makes it possible to use the TrustNet database for overall comparisons of unit trusts, life funds, pension funds, multi-manager funds, fund-of-funds and conventional investment trusts.
investment trust winners
Starting first with the dartboard exercise, investment trusts provide the best odds of picking winners. Over five years the average trust has risen by nearly half, while the next best group, unit trusts, have risen by less than a quarter. However in the past three years, they were among the worst performers. Nevertheless they out-performed substantially again in the past twelve months.
This indicates that trusts are more volatile, which is not surprising, given the very specialised investment mandates for many, typically in emerging markets or smaller companies. The former has performed especially well in the last three years. However, apart from that, these mandates did not confer any favourable bias.
One suspects that the superior investment performance of investment trusts is related to their lack of success in marketing. The TrustNet database contains around 2,000 each of unit trust, life and pension funds, but barely a tenth of that for investment trusts, despite their century-long tradition. One suspects that the size of the front-end load is inversely correlated with investment performance.
If so, life funds should be the worst performers. Unsurprisingly, they are, as they returned only 16% over the full period. This could be excused by the more conservative nature of life policies. If so, their performance during the three years of the worst bear market in living memory should be much better. While the FTSE All Share fell 21%, the FTA British Government All Stocks index rose 16%, so life funds should have performed much better during the bear market.
Next, the selection was reduced to managed funds to eliminate possible biases for or against particular types of fund, as described above for investment trusts. This reduces the difference in performance between the different types but life funds still perform worst.
Traditionally it was widely thought in the City that unit trusts had to have the brightest managers because their performance was so visible, while life companies could afford the sleepiest because their performance was not transparent. One suspects that there is still an element of truth to this belief, despite the blurring of demarcation lines in recent years.
This suspicion is especially obvious at the opposite extreme, where there is the least opportunity for closet-index linking - smaller companies funds. Here there is an enormous gap between the long-term performance of unit and investment trusts on the one hand as against life and pension funds on the other hand. In the last year pension funds have closed the gap, but this may be too short a period to be meaningful.
While their numbers are still small, the results suggest that multi-managers do earn their keep, especially when comparisons are made on a like-for-like basis among managed funds only. Only 22 have been operating for five years, but this field is rapidly becoming popular and now there are 45 operators. Their results continue to be the best for all types of fund in the most recent period, thus validating the concept.
As the biggest issue over the full period has been asset allocation, this suggests that multi-managers have been more aggressive than the balanced managers in timing switches from stock to bond markets and back again. Failure of the latter to do much more than tinker at the margins, while stock markets were crashing, has exposed their vaunted flexibility as hypocrisy.
Further analysis reveals that the advantage of multi-managers is even greater when deciding which equity markets to back. While multi-managers beat unit trusts by 8% over five years in managed funds, their out-performance increased to 24% in global equities. However as the numbers are very small for multi-manager funds that have been operating so long, comparisons over shorter terms may be more meaningful. Over three years the multi-manager advantage is still 10%, of which 6% was in the last year alone, so the concept pays off here as well.
When it comes to stock picking, multi-managers appear to be no better than average in picking the right individual fund managers. This is shown by the figures for the UK All Companies sector, but as the number of multi-managers involved are still few, the statistics are only indicative.
This suggests that the value-added by multi-managers is primarily in deciding which investment game to play rather than in picking individual stock-pickers. They do well when given broad mandates such as global equities, which provide wide scope for top-down decisions, but less well when given narrow mandates.
Asset allocation committees
Managed funds under-perform a random selection of underlying funds. It is confirmed by the statistics, whether performance is measured over one, three or five years. It is confirmed by the statistics whether performance is measured for unit trusts, life or pension funds. It is confirmed by the statistics in large numbers, as the data includes more than 600 managed funds.
Typically asset allocation committees are comprised of the chief investment officer, representatives of all the major asset classes, an economist and/or quantitative analyst plus possibly a marketing manager. One suspects that, as in committees elsewhere, decisions are taken not by vote but by volume. One suspects that the quant may be politely ignored in an environment dominated by star fund managers. One suspects that the fund manager, whose asset class has the best recent performance, has the greatest confidence to argue his case, the most willing audience and the most marketing clout.
Distribution of returns
These suspicions are strengthened by looking at the distribution of returns by type of managed fund. The chart below adds up the number of funds in each 10% performance band over five years. Performance is shown on the vertical axis, while the horizontal axis ranks the bands in ascending order. Investment trusts are excluded because they compete for specialised rather than generalist mandates.
There is comparatively little difference in performance between different life and pension funds. Almost half lie between 0% and 10%. That suggests prevalence of core-and-satellite or closet index-linking strategies.
Returns for managed unit trusts are more widely dispersed, which indicates greater willingness to take bets. Superior performance confirms either that fortune favours the brave or that what fund managers lack is not brains but guts. In as far as different types of investment packaging are increasingly provided within the same group, managers of these packages share the same market intelligence, so the difference must be guts.
Returns of multi-manager funds are even more widely dispersed. This suggests that they make even bigger bets relative to their benchmark. This supports the idea that it is guts that make the difference, because the multi-manager does not have to risk offending colleagues on the asset allocation committee, as his relationship with external fund managers is more impersonal and objective. That could explain the superior performance of multi-managers in the managed fund field discussed above.
The value of packaging
However there is a surprising anomaly, when fund of funds are extracted from the unit trust statistics. These are essentially internal multi-managers and the wide dispersion of results is very similar, yet their performance is even better than that of external multi-managers, so the explanation must lie elsewhere. There any many of these fund of fund unit trusts, so this is unlikely to be a statistical fluke.
It seems that it is the extra layer of packaging that makes the difference. This forces managers to concentrate on the big picture, and it is the big picture that makes the big difference. It looks like managed funds are a dying concept. That could be because the manager may be over-enthused by the hype of star colleagues seeking to differentiate themselves by their stock picking ability and bored by the statistics of economists and quantitative analysts.
If you want management, concentrate on funds of funds. At least they should perform averagely. If you want performance, go for the investment trust menu. That way you can build a portfolio of attractive niches. If you want to earn commissions, sell life assurance, but be prepared for poor performance.
Acquisition completed earlier this month.
Changes to take place by next year
Launched 18 November
Investment Association to create new labels