The problem with compounders

Given the choice between a new item and a slightly used item there aren't many people who would willingly choose the slightly used item for the same price.  Why purchase something older with a shorter lifespan when the alternative is a brand new unused object with a full lifespan?

This is similar to the argument the market has been making about companies affectionately called "compounders".  These are companies that have reliably turned $1.00 into $1.20 year after year, effortlessly, like money printing machines.  The argument is why purchase a company that turns $1.00 into $1.09 when you can purchase something that turns $1.00 into $1.20?  Or even worse, why purchase something that turns $1.00 into $.95 or less!

I agree, all things being equal I would prefer my dollars are turned into a dollar twenty, verses something less.  The argument follows that if these compounders can consistently turn $1.00 into $1.20 then whatever price you pay for this perpetual money machine is consistently too cheap.  I know this sounds absurd, but I want to walk through the math.

Let's say you are offered interest in one of these money machines at $75.  That means for each $1 invested they earn $1.20 in earnings.  At $75 you're paying 62.5x earnings.  This seems really high, it will take you 62.5 years of earnings to earn back your original amount.  Except the company compounds.  Which means the $1.20 invested this year turns into $1.44, and $1.72 the next and so on.  In 20 years the company will be earning $46 a year, and in 30 years $284!  In 30 years earnings will have grown 284 times from that original dollar.

The theory is that you're paying 62.5x earnings the first year and even if multiples compress to 1x earnings you'll make money.  At 1x earnings in 30 years you'll have an investment that returned 4.5% compounded.  That seems like an absurdly low multiple, so let's say they compress to a disastrous 10x.  Now the investment compounded at 12.8% a year over those thirty years.

The math above is simple, this appears to be a can't fail investing strategy.  You can buy something that compounds at 20% a year at almost any price and as long as the company continues to compound at that rate you will have a market beating return.  The break-even where the compounded return matches the market's historical return is about 200x initial earnings.  That means as long as you pay under 200x earnings you should beat the market.

You can see why this investment strategy is popular.  Pay anything less than 200x earnings for a company growing at 20% a year, sit back and count your stash.

The problem is that there is a small flaw with this strategy.  Finding those companies that can sustain high growth for 30 years.  The issue is we really don't know the future and predicting something thirty years from now is really hard.  In the market we have a solution.  We look at the past and reason that if a company has the type of culture that can grow at 20% for the last thirty years then that culture will likely let them grow at 20% for the next thirty years.  Maybe we don't need a history of thirty years, maybe only five or ten.

But here's where things break down for me.  A company that has grown at such a high rate will struggle to continue to grow at such a high rate for a long period of time.  Why you ask? Simple math.  A company earning $1b today will be earning $237b in 30 years and have a few trillion market cap.  Maybe that's reasonable, I don't know.  But what about a company earning $5b or $10b?  It will be earning $1t, and a company at $10b earning $2t.  Those are large numbers, especially considering that if the US grows at 2.5% our GDP will only be $38t by then.  Some simple reasoning would show that there isn't enough room in the future market for the number of compounders the current market currently has.

And that's a problem.  Hopefully the US economy won't consist of a dozen companies doing everything from selling washing machines, to clothes, to cars, to search advertising.  And that's the problem with this strategy, there isn't much room for more than one or two of the established large companies compounding at 20% a year to exist in thirty years.  Or for compounding at such a high rate to continue.  A company earning $1b a year compounding at 20% will be 100% of the US economy is less than 60 years.

But that doesn't mean this investment strategy is dead!  The math is still fantastic and it seems like we should be able to work something out.  The only thing is buying larger established companies with sizable profits doesn't work.  Time isn't on their side.  So we'll need to do something different.  We'd need to find small companies growing at 20% a year that can compound.  The reason for this is because a small company with a small profit base has a lot more room to grow before their earnings dwarf the US economy.

And here is where the market diverges from where the strategy works.  The market, and investors in general are looking at large, or larger companies that have a history of compounding.  To make this strategy truly work you need to look at the potential to compound, not a history of it.  And discovering this is a completely different skillset.

Let's dive in.  Two factors are needed to compound capital, the first is growing revenue.  The second is a business model that has operational leverage.  This means for each dollar invested above a certain threshold generates incrementally more.  There are a lot of businesses that operate like this, it doesn't really matter what you decide to invest in either.

What is most important is you find a business with the correct business model that can grow sales.  The sales engine of the company is the most important aspect, and also the one most overlooked by investors and analysts.  Sure, cost structure matters, and business model matters as does "capital allocation", which is what they do with the tiny bit of leftover money, but what matters most is sales.
Herein lies a problem.  How do you determine that a small company with the correct business model will grow sales at a high rate?  The only way to do that is to visit the company and talk to management.  But talking to management isn't enough.  You need to sit down and discuss their sales strategy, understand who their employees are and evaluate the ability to execute on their plan.

This is clearly a dark spot for most analysts and investors.  How do you determine if the sales manager is selling you, or knows what they're talking about?  Especially if there isn't much in the way of results to look at?  I believe it's possible, but instead of having a solid background in financial analysis you need to have sales experience and understand the sales process.  Instead of reading the newest book on investing strategies your bookshelf should be full of books on pricing, call strategies, how to approach demos, and prospecting.  It's also worth remembering that enterprise sales is a different beast from consumer sales, or small business sales.

When you start to put all the pieces of this puzzle together it starts to become more apparent why everyone didn't invest in Starbucks, or Microsoft, or Oracle when they were tiny companies.  To truly catch a compounder when they're in infancy you need a set of skills that few investors possess.  It's not impossible to build out that skill set.  Understanding this paradox also helps to expose the myth that buying high growth companies is a surefire way to success.  Buying high growth companies IS a surefire way to success if you can buy them when they're small enough and their market is large enough.


  1. The current obsession with compounders at any price is a bit odd. It is great for institutional investors as it provides a great narrative and illustrates their ability to do something that Joe Public can't i.e. find compounders therefore they can raise money.

    As you have pointed out, any additional due diligence refutes the notion that compounders are good investments. Michael Mauboussin's base rate book illustrated that there is mean reversion in earnings growth.

    From 1950 to 2015, only 7.5% of companies grew earning greater than 20% over a ten year period. Extrapolating the ten year statistics over a thirty year period means only 0.4% of companies grow earnings over 20% for a thirty year period. This does not take into account the survivorship bias in the study.

    Great post, it seems like you are spot on and the probability of making a good investment from buying a compounder at elevated prices is low. I'll stick to out of favor companies that offer a higher probability of success.

  2. "The math above is simple, this appears to be a can't fail investing strategy. You can buy something that compounds at 20% a year at almost any price and as long as the company continues to compound at that rate you will have a market beating return. The break-even where the compounded return matches the market's historical return is about 200x initial earnings. That means as long as you pay under 200x earnings you should beat the market."

    How did you calculated the breakeven P/E multiple?