Warp, I think I may have been a little unclear as to what the two different valuation figures imply, so I'll try to explain it here:
1. The Fed model is based on expected future operating earnings, not trailing. Since we are now in the 3rd quarter, I use the sum of Q3(02) + Q4(02) + Q1(03) + Q2(03). This is the way Yardeni used to measure it on his old website.
2. As the quarter draws to a close, it is relevant to recalculate the model as if we were already in the following quarter, i.e. it's a rolling number. So right now that would be Q4(02) + Q1(03) + Q2(03) + Q3(03).
Since we are in the first week of the new quarter, this is pretty useless right now. However, last week I could see that the degree of undervaluation was going to jump significantly by the beginning of July, purely on the basis of the new estimates.
[Note that for the other inputs, I use the daily numbers, namely, the latest $SPX and $TNX (10 year yield)].
I would prefer to use 'as reported' (GAAP) numbers, rather than 'operating earnings', as I understand the current interpretation of operating earnings includes pro-forma.
The problem with that is 'as reported' forward estimates aren't broken down by quarter and I believe there is enough seasonal variability in earnings patterns to make an arbitrary split meaningless.
I think the S&P will have a better grip on the pro-forma vs. GAAP issue when they start using their new interpretation of operating earnings in the not too distant future. I just don't know what they will do with the historical data to ensure comparability. Ideally they would recalculate these as well, but I'm not sure they will do it.
Lastly, to address the issue of trailing earnings, I'm not convinced these are all meaningful in a highly cyclical environment. The old rule of thumb relating to cyclical companies says to buy when PEs are high and sell when they are low. I think that might work in a repetitively homogenous cyclical industry, but it's too fuzzy for my liking for "the market" as it is now.
That's some of my thoughts on the Fed Model. I find it an interesting academic exercise, which has been conveniently automated in my excel spreadsheet. It's also a piece in the puzzle, but in and of itself should not be used for trading decisions. |
