| Article ID: | iaor2016311 |
| Volume: | 35 |
| Issue: | 1 |
| Start Page Number: | 34 |
| End Page Number: | 42 |
| Publication Date: | Jan 2016 |
| Journal: | Journal of Forecasting |
| Authors: | Nadler Daniel, Schmidt Anatoly B |
| Keywords: | financial, statistics: regression, forecasting: applications |
Returns of several US equity exchange‐traded funds on the days of major macroeconomic announcements are examined for the period of January 2009 to July 2013. The ARMA+GARCH model with external linear regression terms that describe announcement events and their surprises is used. It is found that mean daily returns may be notably higher on the announcement days than those for the buy‐and‐hold strategy, though their difference may be not statistically significant. The ISM Manufacturing Reports, Non‐Farm Payrolls, International Trade Balance, Index of Leading Indicators, Housing Starts, and Jobless Claims turn out to be the most statistically significant factors in the model. Three trading strategies that realize daily returns on the various macroeconomic announcement days are compared with the buy‐and‐hold strategy. The choice of announcements with statistically significant regression coefficients yields higher mean daily returns and better Sharpe ratios but possibly lower compound returns. Transaction costs may significantly affect profitability of these trading strategies.