# Fama French sorting

I am trying to calculate my own monthly fama french factors SMB and HML and I have a general question about a final screening act.

I understand that to sort companies in june of year $t$ they need to have market value, book equity data etc, but what about the period from july of year $t$ until june $t+1$ when I calculate the returns for the 6 portfolios? In these 6 portfolios do I only include companies which have from july of year $t$ to june of year $t+1$ companies which have for the entire (!) period from july of year $t$ until june of year $t+1$ a viable return and a viable market value? So for example if one of my companies has a missing return value for, say, december 2000, then I can not include it in the june 2000 sort for the portfolio creation?

Let's start replicating the Fama-French portfolio construction:*

1. Portfolios are created at the end of June each year $t$, based on size and book-to-market ratio (BE/ME).

2. The size breakpoint for year $t$ is the median NYSE market equity at the end of June of year $t$.

3. BE/ME for June of year $t$ is the book equity for the last fiscal year end in $t-1$ divided by ME for December of $t-1$.

4. Calculate the value-weighted return for the portfolios for July $t$ to June $t+1$.

As stated on Kenneth French website:

The portfolios for July of year $t$ to June of $t+1$ include all NYSE, AMEX, and NASDAQ stocks for which we have market equity data for December of $t-1$ and June of $t$, and (positive) book equity data for $t-1$.

So there are three cases:

1. If there is no data for market value in December $t-1$, June $t-1$ or book-value for $t-1$, this stock can not be considered in the portfolio sort. As there is no data for including this stock in the breakpoint calculation, it is also excluded in calculating the portfolio return (as it is not clear in which portfolio it should be added).

2. If there is valid data for a stock in the quoted points of time, it is considered in the breakpoint calculation and in the portfolio return calculation. If a stock is delisted, it's return is calculated until delisting (and therefore until data is available).**

3. If a stock is newly listed e.g. in April of year $t$, it is first added in the breakpoint (and portfolio return) calculation the next year (since there is no data for $t-1$).

* These steps are simplified for clarification. Read their paper common risk factors in the returns on stocks and bonds for all the details, e.g. excluding stocks with neg. book-to-market ratio etc.

** Be aware of the delisting-return in the CRSP universe, which is described in detail in Bali/Engle/Murray (2016), Empirical asset pricing: the cross section of stock returns, John Wiley & Sons., chapter 7.

• Just reading this now, thank you very much for this. So guess my approach to this was right. but I dont understand why my results are so weird. Sep 18 '18 at 21:50