Peer Group Valuation Is (Mostly) Useless

A company’s value is its future free cash flow discounted back to the present. I doubt any readers will argue with me there. And when you break it down, the two things that determine future cash flow are return on invested capital (ROIC) and growth. So if ROIC and growth are the two determinates of future value, then to have a fair peer analysis the included companies should have similar returns on capital and growth prospects. If two companies have different ROICs or growth prospects then it’s not an apples-to-apples comparison.

Having similar growth potential also insinuates the two companies are in the same stage of their life cycle, meaning they’re similar sizes. So including mature businesses in a peer valuation for a small-cap company is meaningless. When a company is selling for less than its peers, it’s almost always because it has lower returns on capital or its growth prospects are not as good as its competitors. There are a few other factors that play into this as well.

How does cash and debt affect this?

McKinsey’s Valuation has an insightful example for this: “Think of Apple as a portfolio with two components: one is an operating business that makes consumer electronics and the other is a pile of cash. The operating business is valued at 7.3x, while if the cash earned 1.5% pretax, it would be valued at 67 times (the inverse of the earnings yield). The company as a whole is valued at the weighted average of the two multiples, 9.8x. Since the 9.8x is a weighted average of two very different numbers, it doesn’t provide any insight into how to think about Apple’s value. It would never make sense to compare the 9.8x with the multiples of other consumer electronics companies.”

Likewise, debt increases the cost of capital and the riskiness of a company. Two companies with similar growth prospects and returns on capital still may not deserve the same multiple if one is riskier than the other. Thus, a peer group valuation that includes companies with far different levels of cash and/or debt may not be very insightful.

So in summary, valuation by peer group is most useful when the companies are close in size, at similar points in their life cycle, have comparable growth prospects and returns on capital, and have similar levels of cash and debt. How many industries really have multiple companies with similar ROICs and growth? Especially since I focus on small and micro-caps, I find being able to do a true peer group valuation is extremely rare. The majority of the companies I look at have zero truly comparable public competitors.

With all that being said, I’m guilty. 

When I wrote about LGI Homes (LGIH) I compared their P/E ratio to the industry average. For the most part, home builders are similar to each other—it’s a commoditized industry where they’re all doing the same thing. Even better, there are a ton of public home builders so using an industry average should be reasonable, right? I’m not so sure. Two of the main reasons I liked (and still like) LGI is that they’re one of the smallest public home builders (meaning they have a longer growth runway) and they focus on the low end of the housing market (lowest average sales price of any public home builder). On top of that, their basic strategy is different than most home builders (how they market, how much inventory they keep on hand, etc). So I don’t think valuing them on a peer analysis is as smart as I was once thought it was. Now I would say it merely adds one more bullet point that supports them being cheap as opposed to being a basis for investment.

Similar situations

Tell me if this investment thesis sounds familiar: “Widget Co. is selling for 8x and has operating margins of 4%. Their peers are selling at 15x because they have better operating margins at 8%. If Widget Co. can simply increase their operating margins by doing x, y, and z, the stock will be a home run.” I suspect doing “x, y, and z” is not nearly as simple as it usually sounds. Different margin profiles for companies insinuates very different corporate structures (efficient vs inefficient, flat vs hierarchical, etc) and that is not an easy thing to change.

Another common investment thesis: “If Widget Co. can just get back to their profitability from three years ago when they were selling for 15x, the stock will be a triple.” The biggest problem here is that even if the company gets back to their former profitability, they are now in a different part of their life cycle than they were before. Because of this, their future growth prospects are probably less than before, which means they deserve a lower multiple than the younger version of themselves.

Comparing companies across industries is even more useless

The logic behind EV/EBITDA is that it gets rid of taxes, capital structure, and how capital intensive a company is. After stripping those out, it’s easier to compare companies across different industries. Why anyone wants to do that is beyond me. A useful comparison to a company within the same industry is hard enough, comparisons across industries is just pointless.

A company selling for 5x EV/EBITDA with a pile of debt is not magically cheap just because we choose to ignore the interest and a bunch of unrelated companies have an EV/EBITDA of 10x. Interest, taxes, and depreciation (or capex rather) are very real cash expenses. I feel like EBITDA is going to be the next generation’s beta.

Comparing companies across industries seems to be most common when valuing SaaS companies. I see all the time people comparing the valuation of a small SaaS company to established SaaS players that are much larger and often in unrelated industries. Just because these ten SaaS companies are valued at 5x revenue doesn’t mean this unprofitable micro-cap should be too. SaaS is not an industry. And while we’re at it, price to sales is not a legitimate way to value a company (but that’s a blog post for another day).

On an unrelated note, I sent out my 2016 shareholder letter yesterday. If interested, it can be viewed here. The letter also has my Where Food Comes From (WFCF) write-up attached to it.