Thoughts On Valuation Multiples
This week we published our semi-annual letter to our investors. If you haven’t read it, you can read it here .
Careful readers will notice we often buy businesses for about ten times pre-tax earnings. And we tend to think of 10x as a long-term valuation floor: i.e. a growing business should be worth at least 10x.
The obvious question is, why?
First I’m going explain the goal that this is a means to. Then, I’ll get into how we derived 10x specifically. Keep in mind that we think of 10x as a rule of thumb, not an inviolable law. It is appropriate in most, but not all, cases.
Bargains in the stock market generally fall into two categories:
A business that is not existentially challenged, but priced as if is.
A business priced as if it will never increase profits, but will.
Both of these scenarios involve a mis-pricing that allows investors to get something for nothing. The difference between what you pay for and what you get is your margin of safety. A margin of safety is always dependent on the price paid and a lower price always makes an investment less risky.
We prefer to focus on the businesses in the second bucket. Dan wrote about why twice in What Should You Pay For Growth? and Falling Markets: Are You Bottom Fishing or Buying Quality?. Though the businesses in the first bucket might re-price quicker, the businesses in the second bucket should earn more over many years and produce higher long-run returns in a tax-efficient manner.
So, our goal is to buy “inevitable” or “virtually certain” growth for free. Now, I’ll explain how we figure out what to pay to achieve this.
A stock’s intrinsic value is the present value of future cash flows. If you had a crystal ball and could see the future, you’d discount each year’s cash flows back to the present and total them. That’d be the stock’s intrinsic value. The stock’s price at the end of the decade is its terminal value.
Most of a stock’s present value is in its terminal value, not its year one or two earnings. If you assume no growth and a 10% discount rate, year one comprises 9% of intrinsic value. Each subsequent year is less valuable. A stock’s terminal value constitutes about 40% of its intrinsic value. This is why we’re not too concerned about a business’s results short-term results so long as we are confident it possess’s long-term earnings power.
Since we don’t have a crystal ball, we can’t predict any given year’s earnings accurately. So, we take a shortcut and focus on the terminal multiple.
A stock’s terminal multiple is given by the formula: 1/(Discount Rate - Growth). If we assume zero growth and a 10% growth rate, we get 1/10% or 10x. Here’s how terminal multiples vary:
Terminal multiples are sensitive to discount rates. Changing the discount rate from 10% to 5% for a no-growth business doubles the terminal multiple. Choosing an appropriate discount rate is important.
How did we decide on 10% then? In theory, the discount rate reflects opportunity costs. You can always buy Treasury bonds, so the 10-year yield (currently 0.6%) is the most academically correct answer.
Fortunately, we have better opportunities than 10-year bonds. Our 10% discount rate is based on our judgements that:
We will have opportunities to lay out capital a 10% rates of return, or higher.
A 10% rate of return on our capital would be satisfactory, though not necessarily exemplary.
Many other investors agree with points #1 and #2 and prices often tend towards this level. The market’s median P/E multiple back to the 1870s is about 15x. If we assume tax rates averaged about 35% then businesses were worth 10x pre-tax earnings.
A 10% discount rate tells us to pay 10x for a no-growth business. If we do and the business never grows (or shrinks), we’ll earn a 10% yield. We hope to do better by getting some growth on top for free.
Worst case, we’e wrong and the business doesn’t grow. At least we didn’t overpay and will enjoy a 10% return. Actually, the worst case is that the business shrinks. In this scenario we’d lose money: our judgements about growth are important. Hopefully we’re not so wrong that “inevitable growth” turns into consistent decline.
Best case, we’re right and earn 10% plus the growth rate. If this happens, our stock’s return will most likely exceed the business’s return. That’s because the market will gain confidence that the business is growing and price it at a higher multiple.
Quick digression — it is actually slightly more complicated to figure out what a business will return. A business’s return will equal its earnings growth plus dividend yield. If you pay 10x for a business that doesn’t grow and it pays out 100% of earnings as dividends, you’ll earn 10%. But if a business retains some earnings, dividends will be less than 10%. The growth rate will equal the fraction of earnings retained times the incremental return on them (reinvestment rate x I-ROIC). A company that retains 50% of earnings and reinvests them at a 30% rate will grow 15% (50% x 30%) and pay a 5% dividend (50% x 10%), for a 20% total return. This is why companies that can reinvest lots of capital at high rates are so valuable. Digression over.
We keep our valuations simple because simple works. Each moving part is a potential source of failure. Every assumption is a potential source of error. Our time is better spent studying businesses and judging their long-term prospects than building complicated valuation spreadsheets.
We didn’t come up with this concept on our own. We came upon it by studying the master, Warren Buffett. We noticed that he paid 9-10x pre-tax earnings for many of his best deals. One of our favorite bloggers, Brooklyn Investor, wrote a great piece about this. Here’s his summary of some of Buffett’s deals:
Berkshire made two large public purchases since:
The Precision Castparts deal is an outlier and shows that 9-10x isn’t an iron rule. However, Apple does conform. He bought below 10x in 2016 and added near the lows in 2017 and 2018. In 2019 Apple traded between 15.2x and 19.7x and Buffett turned from a large buyer into a small net seller.
Paying 10x pre-tax for a growing business is akin to buying an above-average business for a below-average price. Buffett similarly says he prefers to buy a “wonderful business for a fair price.”
In summary, we use 10x pre-tax as a rule of thumb. For most situations, it does a good job. It’s kind of like bowling with bumpers: if you focus on buying inevitable growth for 10x, you can’t get into too much trouble and will probably be quite happy with the result. We focus on getting the “inevitable growth” part right, and don’t waste too much time on complicated valuation models.
Currently our portfolio, in aggregate (as of 4/2), trades at 9.5x pre-tax earnings. We think this bodes well for the future, even if the next few months are rocky.
Addendum (4/5/20)
After I posted this, I realized that Buffett explicitly stated that he likes to buy businesses for 9-10x pre-tax earnings. Here’s a transcript of the question and answer at Berkshire’s 2012 annual meeting.
The question was “What’s a reasonable multiple to apply to the pre-tax earnings of [Berkshire’s] non-insurance businesses?”
Buffett says that he’d like to buy operating businesses with similar competitive positions to what Berkshire already owns for “certainly nine times pre-tax earnings, maybe ten times.” He adds that he might pay a little more if they were similar to what Berkshire already owns because he’d understand them better than dissimilar businesses.
At Berkshire’s 2016 annual meeting Buffett added, “And if you ask me whether I paid a little more for Precision Castparts because interest rates were around zero, than if they’d been six percent, the answer is yes.”
Disclosure: The author, Eagle Point Capital, or their affiliates may own the securities discussed. This blog is for informational purposes only. Nothing should be construed as investment advice. Please read our Terms and Conditions for further details.