Mohnish Pabrai Use of the PE Ratio

[nickenumbers]

All,

 

I would like to get the groups opinion on PE, PEG, growth rates, etc as tools to understand stock pricing and future price appreciation.

 

I am a hugh Monish Pabrai fan and I love his work.  I have heard him talk about his work day, and I believe that he has said that he reviews investment company ideas that come to him daily, but he will not look at anything that is not under a 4 PE, or a 2 PE, etc.  I think he talks about the PE as a filter, or a low bar..

 

I am confident that he is far more complicated than this simple filter, for sure.  But, I am growing my realization on the best way to filter for investment prospects.  I think it must involve, Price, Price to Book, PE, PEG, and it must have some larger understanding of the growth rates.  As well as all the business intangibles and moat evaluation.  This is shockingly obvious, I get it..  But why does Mohnish hand out PE as the primary filter?

 

I guess I am recognizing how little value there is in a stock market Screen as a way to select companies to evaluate.  Filter on PE, is wrong.  Filter on PEG, is wrong.  Filter on LTM EPS growth rate, is wrong  Etc.

 

Thoughts?  Thanks all.

[LR1400]

I believe most of these pros are simple with valuation but broad and “complicated” with analysis and research. Which to me means looking at industries, the future picture, etc. And some like Tepper throw in a large macro view to make a determination. All of this being based on fundamentals of some sort.

 

I listen to Tepper who seems really complicated at times because he’s so well rounded, but when he talks about past decisions he talks about how much of s BJ brained and how simple the decision really was. He probably uses some kind of model but it’s probably similar to the analyses he did when he was in credit at Goldman.

 

I read an interview with Gabelki and another semi-famous investor. They were discussing a natural gas shipper and the other guy went on and on about how the company didn’t hit this number or that number. Gabelli keeps saying, it’s cheap, buy it. It’s cheap, just buy it. He was right, thankfully I listened to him.

 

I believe Buffett is similar. He’s going to look at the earnings yield versus other investments and the future prospects of that yield and its growth.

 

Pickens even thought similarly, lots of fundamental research and analysis, thinking, but he’s making his decision based on what he thinks will happen in the future based on that fundamental research.

[racemize]

I think he (and most other sophisticated investors) are referring to a normalized earnings or owner earnings.  It's quite difficult to come up with such a number that is accurate, but once you do, you get a normalized P/E which is likely the most important thing (ultimately, most investments come down to cash flow). 

 

Most of the other metrics are there to tell you about normalized earnings, e.g., P/B is usually a short-hand way of figuring out what price to pay based on a normalized RoE for an industry, but if you had the normalized earnings, you could skip it.

 

On the other hand, RoE, RoIC, etc., tell you about growth, which you also need to know to assign an appropriate multiple to the earnings.

[Graham Osborn]

I did a little calculation yesterday I think provides a reasonably practical solution to this problem.  I assumed that the price one is willing to pay for a business is the next 10 years' worth of unrestricted earnings.  This only works for simple-capitalization businesses (in other cases EV/ FCF would be more accurate).

 

You can see how the PEG agrees pretty well with this result for businesses growing around 20%, but becomes inaccurate for lower or higher growth rates.

 

There was a debate on an earlier thread about the obvious limitations of the PEG, and I think this helps resolve that debate.

 

For a business where you can't establish some boundaries on probable FCF for the next 5-10 years, PE's are largely worthless unless they are ridiculously low.

PEG_vs_10-year_earnings_calculator.xlsx

[KCLarkin]

You can see how the PEG agrees pretty well with this result for businesses growing around 20%, but becomes inaccurate for lower or higher growth rates.

 

Thanks for this. I think your use of only the next 10 years undervalues the high growth stocks. Your terminal value would be very high for a company growing 20%. I suspect the fair multiple for a 15% grower is well over 30x.

 

Mathematically, PEG is clearly wrong. It does have the advantage of implicitly assuming some reversion to the mean. A company that you know will grow 20% per year for the next 10 years, is maybe worth >40x. But a company that is currently growing at 20% is worth significantly less (since you don't know how long growth will last).

[Graham Osborn]

I think your use of only the next 10 years undervalues the high growth stocks. Your terminal value would be very high for a company growing 20%. I suspect the fair multiple for a 15% grower is well over 30x.

 

Yes, you are right that this approach chops off the tail of the DCF.  I would argue this is the best approach for the investor concerned with establishing a margin of safety in his valuation.  As a practical matter, all too often the projected growth rate proves over-favorable, in which case even the calculations presented here may be high.  Just look at Microsoft circa 2000.  Prudent investing is not about determining a single correct value for a business, but rather about establishing a "price ceiling" below which a profit is high probable.  I have seen more overvaluation atrocities perpetrated by the "terminal value crowd" than any other.  For me, 10 years is the limit - and only in very special cases.  I will leave the theoreticians to debate 20- and 30-year projections.  As your broker will tell you, even a theoretician can receive a margin call - and when he does I'll be happy to buy his shares as he debates projections vs practicalities :)

 

Peter Lynch didn't use the PEG because it was theoretically accurate - he used it because it worked.  I think this method illustrates that the PEG is actually pretty darn useful between PE's of 15-25.  I think the modified curve extends this utility.

[KCLarkin]

Peter Lynch didn't use the PEG because it was theoretically accurate - he used it because it worked.  I think this method illustrates that the PEG is actually pretty darn useful between PE's of 15-25.  I think the modified curve extends this utility.

 

10 year treasuries were almost 10% when he wrote this book, so the ratio might have "worked" in many cases back then. I suspect you are undervaluing growth. But then again, there are exceedingly few companies that can grow 15% or 20% per year for 10 years.

 

[LR1400]

I just can’t see the value a DCF with terminal value for many, if any situations.

 

Stock as a bond concept where you are looking for returns far exceeding the risk free rate and the returns from other opportunities. Simple calculation and how many greats do

It apparently.