I’ve made some changes in how I figured and use the 1YPEG. (This is not necessarily related to the discussions we’ve been having about SKX, etc). I had previously capped very high rates of TTM earnings growth at 200% (even if the trailing rates were greater) but I decided that even that was unrealistic, as companies probably can’t maintain those high rates for another year, and the high growth rates excuse unrealistically high PE’s. I therefore decided to cap the earnings growth I used in figuring the 1YPEG at 150% even if they were over 200%. I then thought that if actual rates were between 150% and 200% I’d cap them at 125%, and if actual rates were 100 to 150% I’d cap them at 100%. I haven’t entirely decided on this though, and would welcome input.
What it would mostly do is reduce some of the effect of little companies just starting out with just a few cents in earnings and showing huge percentage gains which are less meaningful. It wouldn’t have affected some of our stocks that have recently come off their highs, such as SKX, SWKS, AMBA, BOFI, or INFN, who would all still have had low 1YPEG’s. I’ll discuss my thoughts about them in the next post.
Saul
For Knowledgebase for this board please go to Post #9939.
To the extent that you’re expecting that your “forecasts” (which is implicitly what you’re doing when you apply past 4 quarters’ results) to be valid, I totally agree with the idea of capping them somewhat tightly. The only exception would be if you were looking at microcaps, which I’m pretty sure you almost never do. IMHO, those super-high results are obviously nonrepeatable for very long … the longer term you’re thinking about, the more tightly you need to think about capping them.
I’m going to make a guess that not only are super-high results difficult to repeat for very long, but they’re probably also going to occur with great volatility and/or uncertainty/risk. Capping them is an implicit way to risk-adjust your calculations, IMO.
From an aesthetic perspective, I dislike the idea of discontinuous adjustments, particularly with arbitrary boundaries. So, if I were going to favor such an adjustment, my inclination would be to pick some function that would gradually temper any amount over 100%, the farther above, the stronger the temper.
Which said, I’m not sure that any adjustment is required since this is, after all, a screening tool, not the basis for a decision. If one sees enormous growth rates, one should be looking for the source of that rate and seeing that it is the change from 1 penny to $2 tells one a great deal more than arbitrarily reducing the 200% to 150%. Likewise the discussion we have had about GILD and the sudden balloon of revenue from Solvaldi which is unlikely to continue with the same rate of growth. If one doesn’t look beyond the single 1YPEG figure to the numbers that created that figure and to the company and market to understand how likely it is that the pattern will continue, then one is not exercising due diligence as an investor.
Given that you’re talking about using 1YPEG as a screening tool, and that any “adjustment” can be taken into account by examining all of the numbers in context later on, I guess that I’m now tempted to reverse my previous opinion.
I agree with what tamhas wrote. The 1YRPEG is backward looking. We also want to be forward thinking. I use a 2 step process where the 1YRPEG is the first step. There is no judgment call because it just uses the historical numbers so I would not make any adjustment.
For the second step, one should use the available information about the company, the market, the competition, the situation, and anything else to predict what will happen in the future. How will the company do and what will the likely future earnings be. From there you can figure what the likely future P/Es and 1YRPEGs will be. This second step is subjective and everyone will come up with a different future assessment and base their investment decisions accordingly. But changing the 1YRPEG will get messy. If you want to make an adjustment, then just use the 1YRPEG and then make a second value: “adjusted 1YRPEG” but keep in mind that different people will have different adjusted 1YRPEG based on their future view of each particular company.
If that helps you Saul, fine. But numbers are just numbers, and while we may take high rates with different understanding, I prefer uncapped data.
If we arbitrarily cap high rates of TTM growth, we lose the ability to see trends in that data until they break out below the cap, and we lose the ability to discern between capped equities.
Just as I might choose to avoid stocks with P/E >50, I still want to know what distinguishes a 50 PE from a 200 PE stock, and how those values compare with similar firms in a sector. Capping stocks whose P/E > 50 with “50” impairs my selection, and confuses my “watch lists”.
I agree with Chris, making arbitrary changes to how the 1YPEG is calculated seems to be added effort without an obvious benefit. The 1YPEG alone regardless of how it is calculated or adjusted is a historical number, and without any other context can only tell you a stock “might” be cheap. Thus as many others have stated here, it is best used as a screening tool.
A better metric would be something that looks forward at least to some degree. I had offered a 1 yr forward PEG previously. I understand the difficulty with predicting the future, but using an average of analysts estimates or the company’s own estimates would give a basic clue.
The 1YRPEG is backward looking. We also want to be forward thinking. I use a 2 step process where the 1YRPEG is the first step. There is no judgment call because it just uses the historical numbers so I would not make any adjustment.
Chris, I guess that is what I was trying to do in a way, trying to adjust for what the future will be in a subjective way, and recognizing that the 1YPEG is just a guidepost of a sort.
Saul
I think capping it makes a lot of sense when you think in the context of a backward looking metric to guide forward projections. It actually makes your method a bit more conservative and prevents the story of newly increasing earnings to looks better than they are.
Of course it is only one piece of data so making that more conservative probably strengthens your overall approach. I highly doubt that it would disqualify many investments that you would have otherwise held on to. One simple way could be to backtest some prior quarter data and then look at your notes from ERs and see if the differences would cause you to rethink your opinion (positively or negatively).
Another suggestion would be to look at what average earnings growth rates for all your holdings over however many years you have data for. Taking a standard deviation, you may make your caps or weighting at a 1, 2, and/or 3 sigma deviation from the earnings growth you typically like to see in your portfolio. This may make the calculation more tailored to the types of stocks and the growth you want to hold.
Not sure if this helps but probably some ideas I’d play with depending on how much data I had at my fingertips.
Given that my results are not nearly as outstanding as Saul’s, I feel a little funny offering these thoughts.
I use 1yrPEG followed by calculating ftmPEG (future twelve month PEG) using consensus estimate of adjusted EPS from Schwab’s earnings tab. Since analyst estimate consensus is often much more conservative than 1yrPEG, it tends to act as a counterweight, so I do not modify 1yrPEG beyond truncating above 200%.
Following up with my own view of the company’s future prospects is the final step, but this final step is qualitative rather than quantitative, encompassing many elements discussed on this and other boards as well as reading company reports.