The very last trading day of
each quarter is often the busiest and a lot of that activity is related to
window dressing portfolio managers.
Recent academic research confirms investor's worst
fears: Window dressing erodes mutual fund shareholder value.
Just one more reason among many
to reconsider the benefits of traditional index funds and exchange-traded
funds (ETFs), which largely avoid this problem.
Ernst Schaumburg
is a PhD in Economics from Princeton University and has studied the window
dressing phenomenon extensively. Schaumburg currently works at Kellogg
School of Management at Northwestern University and he talks with us.
What is
window dressing and why does it matter?
ES: “Window dressing” is the
practice by some fund managers of adjusting their holdings immediately before
reporting their portfolios
to the SEC at quarter end. The ultimate goal of window dressing is to make the
fund manager look better to investors than he really is.
The way is works is as follows:
While investors can observe daily fund returns, they have no way of directly
observing the composition of the portfolio which generated those returns.
Instead they must in general rely on the quarterly portfolio snapshots reported
to the SEC. A manager can for instance hide the fact that he invested outside
his universe by eliminating these positions before filing with the SEC.
This all matters to investors
because, in order to construct their desired overall portfolio of funds, they
need to know what kinds of risks each manager takes on to generate his returns.
Based upon your research, does
window dressing add shareholder value?
ES:No, on the contrary,
window dressing actually destroys value by incurring transaction costs. It
results in excess turnover not associated with any legitimate (alpha generating)
trading strategy.
Can window dressing be
easily detected by the average investor?
ES: No. Almost by
definition it must be hard to detect since rational investors would want to
abandon such a fund. The managers have to stay under the radar to avoid being
detected by techniques such as style analysis which are commonly used by firms
such as Morningstar and Lipper.
What distinguishes a momentum trader from a window dresser?
ES:The two can in some cases
appear similar. For instance, a window dresser might remove “embarrassing”
quarter losers from his book and replace them by stocks that did well during the
quarter prior to reporting. The momentum trader on the other hand engages in a
legitimate strategy which involves investing in recent winners and shorting
(relative to benchmark) recent losers. The idea being that the recent
performance will continue over the next quarter.
There are however two key differences. While the
momentum trader in general will trade continuously throughout the quarter, the
window dresser will have abnormally high turnover towards the end of the
quarter.
The second difference is that a skillful momentum trader knows which
winners/loses to trade and hence will tend to outperform his benchmark although
this can be hard to detect for funds with a short track record.
What are
securities regulators doing to protect fund shareholders?
ES:We
have already seen an increase in the required filing frequency from semiannually
(the old N-30D filings prior to May 2004) to quarterly (the new N-Q filings). It
is not likely that this change will eliminate window dressing although it does
make it a more costly proposition. To make life harder for window dressers will
require more mandatory disclosure of the timing of trades, such as weekly
turnover numbers and/or the timing of the top trades.
What about
ETFs? Are they guilty of window dressing too?
ES:I
do not know of any evidence of window dressing for ETFs or closed-end funds.
Moreover, given that the shares are exchange traded the incentives for window
dressing would be much less clear than for open-end funds trying to maximize
fund inflows.
