“…A number of processes exist for this, including the prominent adoption of different proxies for one or more variables in the model or alternative data frequencies, for example. We embrace the second option, banking on previous clues that the predictability models of financial series such as those assessed here could be sensitive to the choice of data frequency (e.g., Narayan and Liu, 2015;Narayan and Sharma, 2015;Salisu and Adeleke, 2016;Narayan et al, 2018;Salisu, Ndako, and Oloko, 2019). Hence, having employed monthly observations in the main analysis, we thought it revealing to determine the sensitivity of the results using a quarterly data frequency.…”