Portfolios in disguise ? Window dressing in bond fund holdings

This paper enlarges the current research on window dressing practice on strategic allocations. The existence of window dressing is of special interest for investors in order to know whether portfolio holdings vary across months with public portfolio disclosures. The monthly bias free sample of bond funds contributes to the scarce works on this phenomenon in the literature. The monthly frequency provides a more comprehensive analysis than the quarterly perspective traditionally applied in previous research. A comparison between disclosed and undisclosed portfolios is followed by the test of the window dressing practice on public debt allocations. We carry out complementary analyses such as the seasonality of this practice or the significance of some fund characteristics (size, fees, age and past returns). Empirical evidence also shows that some funds significantly modify portfolios prior to disclosure decreasing the public debt allocation, but it is not a general behaviour in the sector. EFM classification Codes: 340; 370