We examine the performance of real estate mutual funds during January 1991-
December 1997. As a group, the sampled funds outperformed the Wilshire Real
Estate Securities Index on a risk-adjusted basis by more than 5 percentage
points annually. We attempt to explain these surprising findings by examin
ing the fund's asset allocations across stocks, bonds and real estate prope
rty types using Sharpe's (1992) effective-mix test. We find that all of the
superior performance is attributable to fund managers' decisions to overwe
ight outperforming property types (apartments and health care) relative to
the Wilshire Real Estate Securities Index weights. Performance of the funds
matches a multiple-property-type benchmark that takes account of the fund'
s exposure to each property type. Therefore, real estate funds demonstrated
superior allocation across property types, but neither superior nor inferi
or selection within property type, during 1991-1997. Our findings emphasize
the importance of asset allocation for real estate mutual-fund performance
.