Question

I'm trying to calculate a regression beta based on daily change of market index and a stock price, the beta i get is really low 0,2 for and IT stock. So i tried YTY changes for regression beta and i got 1,6, which is meaningfull. Is it a correct method ? Thx

Answer 1

In most cases the first problem ie. low co-relation is due to having taken a very large range of dates. I would do the same day to day regression with a more limited range, say one year and see if it is similar to the YTY. Sometimes the day to day co-relation also goes wrong due to the mismatch of dates. Double check that. To avoid this problem I use a v lookup table with dates as reference rather than simple copy paste. Although a beta of 1.6 is more meaningful and variations are expected when dealing with 2 different models the difference between .2 and 1.6 is too large to be accounted for as a systematic error. Also during the period in question check for major reasons for rise and fall of the stock market. In a abnormal market ie. recession, crisis etc... we can not expect meaningful betas.