Comparison of actual funds to their pre-closure (“ghost”) portfolios after closure.

I form (“ghost”) portfolios, each consisting of the same stocks that a fund held immediately before closing to new investors. The initial portfolio weights are the same as the weights of the actual portfolio of the fund at the time. This method resembles the approach of Lapatto and Puttonen [10]. Each portfolio is formed with the corresponding fund holdings as of the last available report date prior to closure, given that it is not more than a quarter before the event. A fund must have been open to investors for at least 24 consecutive months prior to the closing event in order for its “ghost” portfolio to be formed. The portfolio is held passively for 24 months after the event. Portfolio alphas are calculated at the individual portfolio level using time series regressions of the monthly portfolio excess returns against the relevant factors. Fund alphas are calculated at the individual fund level by regressing the fund excess returns before fees against the corresponding factors. The post-closure gross returns of a fund must be available for no less than 12 months. The table contains the average monthly performance measures (%) of the funds and the portfolios as well as the average monthly pairwise differences in performance (%). Panel A shows the results when the “ghost” portfolios are not rebalanced, whereas Panel B portrays the results when the portfolios are rebalanced quarterly. The t-stats are reported in parentheses. Panel C contains the average percentage of the actual fund portfolio that is invested in the stocks of the corresponding “ghost” portfolio as of a given number of months relative to “ghost” portfolio formation.