I've come across a research paper where for a specific period of time, the portfolio has negative returns (or roughly flat returns). During this same period of time, the portfolio's Fama-French 3-factor alpha is fairly positive.
Here is a photo of the returns from the paper, I'm looking at the 2000-2007ish time frame. My understanding is as such, but I'm not completely sure if this is right:
$$ R-R_f = B_1 (R_m-R_f) +B_2 (SMB) + B_3 HML + alpha $$
If in a given year $R_m - R_f$ was -5%, and SMB and HML were zero, but my particular stock (or portfolio) was -2%, I would have +3% alpha. In this way, I would have positive alpha while having negative returns.
Is this understanding correct, or something else?