Criticism of funds of funds tends to focus on how their layers of fees have a negative compounding effect on returns. A new study, though, suggests a greater problem may be their poor selection of funds – particularly when they are restricted to internal products.
Edwin Elton, Martin Gruber and Andre de Souza of the Stern School of Business at New York University investigated the 1,612 funds of funds available in the US between 2002 and 2015. These funds represented more than $1.7 trillion in 2015, equivalent to 11% of the total assets in all mutual funds, and thanks to an increasing preference for asset-allocation solutions the academics recorded their five-year growth rate in assets under management at 13.4%, well ahead of the 5.7% for all mutual funds.
Many funds in the full sample were duplicate products, however: target-date funds with the same management and essentially identical holdings, funds in the same range distinguished only by slightly different risk mandates, and so on.
The authors excluded duplicates but retained one representative fund in each suite to avoid overweighting decisions by single management teams, and also removed purely passive strategies from their database.
This left a final haul of 219 funds of funds, offered by 115 fund groups.
A first screen revealed the extent to which this market is dominated by managers who invest only in their own house’s funds. Through the paper’s timeframe, between 51% and 77% of funds of funds allocated exclusively to products within their own family. Only 14% to 27% invested solely outside their parent firm, generally because it did not run other fund ranges, with the balance of 9% to 29% a mix of both. Of the funds that did invest externally, more than 85% did so from necessity because there was no internal option.
Taking the fun out of fund selection
So how did managers of these funds of funds perform? Elton, Gruber and de Souza compared their holdings in a given sector with the average fund in that category. So rather than questioning the managers’ asset allocation, they asked whether managers were buying above-average funds.
They were not. For all funds of funds, the average holding-weighted alpha sacrificed by not owning the average fund in a chosen sector was 19.9 basis points per year.
‘This implies that the average fund of funds forfeited almost 20 basis points in alpha per year by not selecting at random funds’ in a given sector, the authors noted.
This could not be attributed to picking expensive funds either: the typical fund chosen by funds of funds had an expense ratio 5.4 basis points per year less than the average alternative, so they underperformed despite using cheaper products.
Funds of funds that did not invest in internal products did at least fare slightly better than the full peer group, outperforming them by 40 basis points a year. Again, this was not simply because they bought funds with lower costs.
On the other hand, internally focused funds of funds were disappointing. Here, the professors considered whether managers could have selected superior products not from the whole market but from their within their own group’s line. For example, if a Fidelity fund of funds owned a Fidelity large-cap growth strategy and there were other Fidelity funds available in that sector, did it own the best such manager?
Again, no. The average alpha from the funds they held was 34 basis points per year less than from the average fund offered by the family in the same category. ‘These results suggest that either managers of funds of funds are using some family or manager goals in selecting the particular funds to include, or that their selection criteria are perverse since they are doing worse than random selection,’ commented Elton, Gruber and de Souza.
One hint towards an answer was that the average fund selected in these internal funds of funds charged five basis points per year more than the average fund in that family.
‘All of the evidence presented above on comparative performance suggests that pressure to meet fund family goals supersedes the benefit from private information in determining the choice of funds by funds of funds,’ they concluded.