Looking for an old(er) report

[LC]

Hey folks,

 

Does anyone remember a report/article/whitepaper from a few years ago that tracked blue-chips from the 50s/60s? Essentially it looked at PG as one example, and the P/E multiple it was trading at in say 1950. Then it took the next 50 years of earnings/dividends/splits etc. and discounted it back to that same point in 1950, in order to figure out what P/E it *should* have been trading at. Essentially making the point that quality companies are undervalued over the very long term.

 

I thought I saved it on my computer or google drive but I can't find it :( Hopefully one of u guys remembers or has it saved. Cheers

LC

[Travis Wiedower]

I don't have what you're looking for, but I did something similar myself years ago (can't find the spreadsheet right now unfortunately). I picked 20 or 30 high quality companies (P&G, Coke, J&J, etc) and tracked their multiples and earnings/dividends back several decades. The results were quite fascinating and lined up with what you found. The companies I looked at were almost always undervalued and there were only a few exceptions.

 

It's one reason I don't mind investing in more expensive businesses like IBKR that I think are very high quality. Mr. Market is focused on the short-term quarterly results and probably isn't as good at valuing companies that are going to compound value for decades.

 

Disclaimer: lots of hindsight bias

[wachtwoord]

You already mention it as a drawback but the hindsight bias in this is off the charts and probably disqualifies the entire result. Maybe to reduce it pick up a large number of stocks at high relative multiples at the time you start your analysis and see whether the results hold for the entire group (probably not).

 

The key here is really to identify quality businesses that will survive for decades to come. Can you reliably do this without the advantage of hindsight?

[KCLarkin]

You already mention it as a drawback but the hindsight bias in this is off the charts and probably disqualifies the entire result. Maybe to reduce it pick up a large number of stocks at high relative multiples at the time you start your analysis and see whether the results hold for the entire group (probably not).

 

We know that as a group high multiple stocks underperform, so I don't think this analysis would be useful. Probably the best "simulation" is the Nifty Fifty review by Jeremey Siegel:

https://www.aaii.com/journal/article/valuing-growth-stocks-revisiting-the-nifty-fifty

 

This largely avoids hindsight bias since the "nifty fifty" lists were widely available at the time.

 

But note that even though Coca-Cola's "warranted P/E" was 82x, this doesn't factor in the extreme volatility you would suffer if you actually paid 82x. You would suffer many 50%+ drawdowns.

 

[Cigarbutt]

Can't find the specific article but here are potential helpful links:

 

http://economics-files.pomona.edu/GarySmith/Nifty50/Nifty50.html

 

http://www.burgundyasset.com/data/newsletter/2001_11_Wise_Passivity_Cautious_Opportunism.pdf

 

Useful variables to dissect: what to include/exclude (especially important looking retrospectively) and effect of market cap weight within the "index" chosen

 

The more long term you look the less important the price you pay becomes. Not many people have time horizons of 30 years or more as most stocks turnover is about 3 months.

 

Dry powder versus interruption of compounding.

Price is what you pay and value is what you get but cash return is 0%.

The game is on.

[Travis Wiedower]

You already mention it as a drawback but the hindsight bias in this is off the charts and probably disqualifies the entire result. Maybe to reduce it pick up a large number of stocks at high relative multiples at the time you start your analysis and see whether the results hold for the entire group (probably not).

 

The key here is really to identify quality businesses that will survive for decades to come. Can you reliably do this without the advantage of hindsight?

 

High multiple stocks underperform, that's already been proven many times. That's not what I was testing though (and in reality I wasn't scientifically testing anything, just followed my curiosity and discovered some interesting results, I'm sure my method had errors). I wanted to see how accurately the market values long-term compounders and it was pretty overwhelming evidence that the market doesn't value them as well as I expected before I went through the exercise.

 

Ask me in ten or twenty years if I was correct on the quality of IBKR's business  ;D

[LC]

The key here is really to identify quality businesses that will survive for decades to come. Can you reliably do this without the advantage of hindsight?

 

Relatively, I think you can skew the odds in your favor...the key is to remove the current price (or current multiple) from the equation entirely. The only way to do that is to think on such a long-term horizon (decades) that price becomes more irrelevant. Which reminds me of that Ben Graham documentary quote (spinoza): "You must look at things in the aspect of eternity"

 

And, maybe even more importantly, by doing so you can generalize a lot more and make arguments based on industry and economic dynamics rather than daily minutia.

 

Take the SP500 constituents. Imagine they are all trading at the market multiple. Can someone make relative claims on the value/"moat" of one company vs. another? Which company has a better chance of being around in 20 years vs. another? I think so, maybe not conclusively but you can make a strong argument of one vs. the other.