Beyond Window Dressing
Coverage of the 2017 Northern Finance Association Conference
BY Caroline Cakebread | September 14, 2017
Scant or non-existent data is a challenge when it comes to selling new mutual funds, which is why a lot of young funds rely on “window dressing” to attract potential clients. Window dressing involves creating a hypothetical set of returns that show how the fund, based on its backfilled holdings, might have performed in the past. In the absence of real data, such window dressing is the only way to demonstrate that the fund managers are worth investing in.
But how valuable is window dressing as a decision-making factor for investors – does it really show whether or not a fund with no track record is going to be a good investment?
A new paper by Oleg Chuprinin and Thomas Ruf aims to answer that question. The paper is to be presented this weekend at the 2017 Northern Finance Association Conference – Canadian Investment Review is a media sponsor for the event.
Turns out that the need to market through past performance often drives the portfolio compositions of new funds. The authors compare what they call “ghost returns” (the implied fund returns based on past performance and on which the fund is marketed) to the actual returns in the 12 months after the fund is launched:
“…we find that portfolio compositions of young funds are structured to convey the impression of strong hypothetical past performance, significantly more so than those of mature funds. The average difference between the holdings-implied fund return computed in the 12 months before the initiation (the “ghost return”) and that computed in the 12 months after the initiation is over 4% for broad U.S. equity funds and over 19% for sector funds.”
At the same time, the authors attribute ghost return performance to three choices:
- timing of fund initiation relative to market performance (48% of the effect)
- choice of the fund investment focus as reflected in the benchmark index (21%)
- overweighting of well-performing stocks in excess of the market and the benchmark performance (31%).
The authors also find that that good ghost returns result in strong fund flows, but those flows fade over time as real returns become available.
Bottom line – buyer beware: “our results suggest that fund managers make extensive use of marketing devices rooted in investors’ backward-looking bias, significantly more so at times when realized fund returns are too scarce to provide a reliable alternative signal of managerial quality.”
Read the full paper here.