Resumen
We explore monthly seasonality in long term corporate bond monthly total returns from January 1926 to December 2013. We test three types of mont returns from January 1926 to December 2013. We test three types of month effects. We further slice the data into three sub-periods. In addition, we analyze the data based on Republican and Democratic presidencies, periods of contraction and expansion, as well as periods of crisis.The mean of monthly total returns for the entire data set (0.51%) is significantly greater than zero. The mean of monthly returns of December was the highest (0.85%), but it was not significantly higher than the mean of monthly returns of the other eleven months stacked. However, the mean of March (-0.01%) was significantly lower than the mean of the returns of the other eleven months stacked. We find the volatilities of the returns of some months are significantly higher or lower compared to the volatilities of the returns of the other eleven months stacked. January?s mean return was the second highest followed by a dip in February and March, and then an upward trend peaking in December. When the data is sliced into four sub-periods, we do not find a consistent pattern in seasonality. The mean of monthly returns during the Republican presidencies (0.64%) is significantly higher than duringthe Democratic presidencies (0.38%). The mean of monthly returns during contraction periods (0.79%) is significantly higher than during expansion periods (0.44%). Given that no consistent pattern emerges with respect to month effects in the full data set and for various sub-periods goes to strengthen the argument that though not fully efficient the U.S. corporate bond market exhibits a high degree of efficiency.Keywords: Month effect, seasonality, volatility.