In this Paper we evaluate (return based) style analysis. The portfolio and positivity constraints imposed by style analysis are useful in constructing mimicking factor portfolios without short positions. We use a simple simulation experiment to show that imposing these constraints in estimating the factor portfolios leads to significant efficiency gains, if the factor loadings are indeed positively weighted portfolios. If this is not the case though, imposing the constraints can substantially bias the exposure estimates. We also show that the actual portfolio holdings will in general not reveal the actual investment style of a fund because of cross exposures between the asset classes, and because fund managers may hold securities that on average do not have a beta of one relative to their own asset class. Style analysis may be used to determine a benchmark portfolio for performance measurement. If the actual exposures are a positively weighted portfolio and if the risk free rate is one of the benchmarks, then the intercept coincides with the Jensen measure. In general, the intercept in the style regression can only be interpreted as a special case of the familiar Jensen measure.
