[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: Fama-MacBeth regressions
> 1. Looks like there is quite a bit of arbitrary tune up: why 24
> months? why 10 portfolios?
Actually, the number of portfolios and the number of estimation months,
though arbitrary, is not crucial...
I and others have done sensitivity analysis and find that these two don't
really matter (in terms of tests of validity of CAPM) as long as:
1) we have a long time-series of data
2) the number of portfolios is not too small (even 5 portfolios suffices).
In fact, my current paper demonstrates that Fama-Macbeth CAPM test results
are extremely "sticky" (non-sensitive) to
1) # of portfolios
2) extimation period length
3) choice of market portfolio
* For searches and help try: