The human body replaces around 1% of its cells every day, so a few times a year you are, setting aside ship-of-Theseus-style concerns, a statistically new person. Companies turn over their assets in the same way: much of what makes them valuable is either depreciating (capital equipment), being quickly converted into cash (working capital), or growing in value but liable to walk out the door at any time (employees). So, over time, this means that the vast majority of the assets a company owns consists of things bought with their retained earnings.

Which brings us to an important question: how do they choose what to buy? For the moment, let’s set aside strategic concerns, and just consider the 1-to-N kind of scaling: what are the commonalities in the decision process that leads Netflix to buy a new show, Exxon to drill a new well, Chipotle to open a new store, Equinix to buy a new datacenter, etc.?

Internally, at least in theory, companies are measuring every potential investment against a “hurdle rate”, or required rate of return, which is ideally higher than their cost of capital. So if a business can borrow at 6%, and its cost of equity capital is 10%, and it finances growth with half equity and half debt, its blended cost of capital is the simple average of those numbers, i.e. 8%, which means that, in theory, investors should be indifferent between a) the company using some capital to buy an asset with an expected return of 8% and b) saving the capital and returning it to shareholders.¹ But that's making some hefty assumptions. Sure, if that 8% return were guaranteed, investors would be truly indifferent to which choice the company made, but a truly guaranteed 8% return doesn't exist in a world where the 10-year treasury yields around 4.3%. And, aside from index funds, shareholders don't really want companies to convert $1 of shareholder cash into operating assets worth exactly $1; the level of CEO pay, not to mention the fact that it's mostly structured as incentive-based variable compensation, implies that CEOs are not supposed to be running a generic index fund of miscellaneous average assets, instead they’re there to generate some upside.²

A straightforward way to handle this problem is to set a hurdle rate high enough relative to the cost of capital that companies have to be judicious, but low enough that they don't refuse to make any investments at all. Common numbers here are 15%, or maybe 12%. Interestingly, this paper from a decade ago shows that hurdle rates used to be a bit higher than bond yields, with firms demanding ~16% returns on equity when treasury bonds yielded 12% or so. But when rates dropped ten points, hurdle rates dropped roughly one, with hurdle rate surveys still showing 15%. Some companies and industries choose a higher rate than that: EOG Resources, the fracker, says that they aim for a 30% well-level return (i.e. ignoring corporate costs, so not a true 30% return). But there's a reason they think this way: they were talking about this target in November, and their model assumed that Henry Hub natural gas would be at $2.50. Natural gas has since dropped over 35%. So that high hurdle rate is partly a way to account for the fundamental uncertainty in some businesses. For an energy company, getting to a decent return over the full cycle means being extremely conservative any time there's enough cash flow to invest to expand production and an appetite to do so, because other industry participants feel the same way and will quickly flip the supply/demand imbalance.

For other companies, IRR targets aren't round numbers and aren't calculated based on approximations. Prologis, the warehouse real estate company, talked about this on a recent investor call: "Let's take cost of capital, for example. This is a metric that we evaluate internally every 4 to 6 weeks [in] partnership with our treasury, valuation and research team and that will guide most of our investment decisions." That's a fast clip, but it also means that they aren't investing today based on what made sense in last year's economic environment.

There's a broad question here: why have hurdle rates been stuck at such a high level? It's hard to have a definitive answer. In theory, if there's a "wedge" between a hurdle rate acceptable to investors and the one companies choose. Investing within the wedge means accessing a set of opportunities that are good enough for investors but that don't have much competition. In other words, if you need to hit 10%, your competitors aim for 15%, and you target 12%, you're still adding value relative to paying out earnings as a dividend or buying back stock, and you can deploy a lot of capital this way.

But in practice, that doesn't seem to happen much. The economic world isn't being taken over by companies that invest a lot at lower-but-still-decent rates of return and continuously issue stock to do so. Instead, the dominant companies are the ones that get very high returns, and can comfortably reinvest a lot while also buying back stock. One possibility is that there's a wide swathe of fairly mediocre businesses that are simply not that interesting to public markets, but that can produce a decent rate of return with enough leverage, so those businesses tend to get taken private. Meanwhile, the public markets select for companies that are unique enough that they have high returns but are constrained in how much more they can invest at those returns, and then by a long tail of companies that are too mediocre to qualify based on this criterion but haven't gone under just yet. Which is actually basically fine; it's not a tragedy if the 100th-best oil company, retailer, or real estate business isn't available to public investors, because there are still plenty of companies to choose from. And the companies that are truly distinctive and truly great tend to end up being publicly-traded after a while, simply because it's the best way for founders to monetize what they've built and for the company to pay talent with equity.

So the best answer to the hurdle rate mystery might be that companies demand higher returns than they used to because companies are better than they used to be, and high hurdle rates are both a cause and effect of this. The world might be better-off if these companies, with their economies of scale and their abundance of human capital, grabbed every opportunity for a 10% return they could find, but in the meantime the current system works perfectly well.

Read More in The Diff

Internal hurdle rates are implicitly part of the story for any growth company, but here are some Diff posts that dig deeper and give more examples:

• You can think of growth stocks as being valued as an option on future reinvestment. What you’re paying for is not just the stream of 25% return-on-equity returns from their existing assets, but for the returns on the capital they can deploy at 20%, 18%, etc.

• Sometimes great companies invest in less exciting adjacent businesses, but this makes sense if the main business’s core asset makes adjacent ones more valuable. (This is a good time to remember that the two most valuable browser businesses in the world make essentially all their money from Google search: Safari, by charging Google for placement, and Chrome, by saving Google from paying other search engines more for the same.)

• In lower-rate times, we looked at how you should value equities when the risk-free rate is zero ($).

• Hurdle rates probably follow a very long cycle ($).

• If the Prologis quote grabbed your attention, there’s much more on Prologis here ($).

• Some businesses are partly a hurdle rate arbitrage, like aircraft leasing ($). A leasing company with assets that can be moved around the globe will tend to have a lower cost of capital than a geographically-bound business with direct exposure to fuel and labor cost fluctuations. So expensive assets like planes find their way onto a different balance sheet than that of the airlines that use them.

1. A quick terminological note: we say "fund with equity," but this usually doesn't mean that the company issues stock to make every new investment. Instead, the way to think about it is to imagine that shareholders are entitled to 100% of the cash a business generates above its cost of servicing debt (plus whatever prudent buffer is necessary to make sure they have decent odds of paying back that debt), and any time dividends plus buyouts sum to less than free cash flow, the retained earnings are the funds from which "funded by equity" projects are backed. So perhaps “funded by profits” is a better way of saying what companies do here. Conceptually, it's a lot easier to imagine a new company coming into existence to buy whatever the return-producing asset in question is, and choosing its capital structure just that one time. This footnote is a bridge between what's easy to talk about in theory and what actually happens in practice.

2. We can get even more lost in the thickets of financial theory here. What if the cost of capital is a good reflection of not just the market's assessment of the required rate of return for certain industries, but for certain companies? If two companies have a similar financial structure, and one has a lower cost of capital, it implies that the market has higher confidence in the low cost of capital company's ability to make judicious investments. If the market tells you that a 5% return on equity is good enough for them, why not take the market at its word? The strongest explanation here is that economic profits tend to drift towards zero over time, and, specifically, many of the high-ROE, extremely stable companies out there have ran out of room to expand. Someone like Verisign just can't do much to expand their core market. That’s ultimately why businesses like Hilton or Marriott ditched their capital-intensive pieces (owning hotels) to focus on the high-ROI ones (licensing brands, managing properties, operating a loyalty program). If they used their cheap equity capital to start growing their owned and operated segment, that cheap capital would get expensive in proportion to how aggressively they deployed it. Finally, markets have mood swings, and it's best for managers to at least consider the possibility that the marginal price-setter has wildly but temporarily overestimated the profitability of cannabis, or "AI" businesses that require suspiciously few researchers, or, say, movie theaters and mall-based video game retail.

Join the discussion!