A few weeks ago, we published an article discussing our approach to discount rates. Since then, we’ve been distributing the piece to people in the hedge fund and academic world, and they have encouraged us to publish the piece more widely.
We sent a copy to ValueWalk, and the editor there, Jacob Wolinsky, was kind enough to have run the piece as one if his top articles. A few people have commented, and one comment was especially interesting as it gets to the heart of the difference between orthodox methods for choosing a discount rates and Framework’s. Here is the complete exchange.
Reader Comment
I find your approach problematic for two reasons:
- Let’s hypothesize two firms capitalized identically. One of them is in a highly cyclical economically sensitive industry like machine tools (presumably a high beta) the other is in a more stable industry with low economic sensitivity like branded food products. Are you really saying both should be discounted at the same rate despite their disparate risk profiles?
- Alternatively let’s take two companies in exactly the same business. One has debt / market cap of 4 to 1 (very highly levered) while the other has a .25 to 1 ratio. The two have dramatically different risk profiles but you would have us discount them both at 10%?
I must be missing something but this doesn’t seem to make sense.
Erik’s Reply
Thanks much for your thoughtful comment. In your response, you’ve provided two reasons for doubting my argument, but in fact, I would counter by saying that these two reasons turn out to represent different aspects of the same premise: Namely, that a discount rate can be altered to reflect operational risk.
Let me say that, with regards to equity securities, I disagree with this premise; I’ll give you an example of why in a moment. While I disagree with the premise, I can also understand its origin. A bond investor is offered a contractually-established cash flow stream. Basing the discount rate on the likelihood of receiving those cash flows makes sense under those conditions – a short-term, contractually-determined cash flow stream with legal recourse for non-performance.
As I mention in the article, in the case of a bond, one is not betting on whether a firm will succeed, but rather, on the relative likelihood that it will fail to deliver one its promised cash flow stream.
However, an investor in an equity is an owner of a perpetual, residual claim to excess profits. No payment is promised, let alone guaranteed; in fact, many corporations pay out little or no profits to owners. My base argument is that requiring an imaginary “risk premium” due to uncertainty related to operational outcomes makes no sense if one does not have legal recourse to capture a flow of future cash payments.
Now, for an example to highlight why I reject the idea that a discount rate can be used to reflect the riskiness of an equity. Imagine a regulated electric power utility. The regulatory environment is friendly and the region the utility serves is stable in terms of power demand. Of course, there are issues of demographics, and temporary drops in demand related to recessions and lowered business activity, but in general, the demand environment and the regulatory environment are well-understood. The sovereign long-bond is trading at a very low yield – let’s say 1%. What discount rate would you use to value that regulated utility? Think of a number. Okay, now take your number and double it. What’s that number?
The company of which I’m thinking is TEPCO – the Japanese utility whose reactor was damaged in the Fukushima earthquake and tsunami. Was your discount rate assumption high enough to offset the risk of a nuclear disaster? Of course not.
When an investor invests in an equity, they need to understand the operational risk associated with that enterprise, risk which can be exacerbated by financial leverage. But an investor need not invest in that security; an investor always has an alternative. The simplest alternative is to buy the index. Doing so, and holding the index for long enough, you’ll return around 10% on a nominal basis.
The important risk from an investment perspective (as opposed to an operational one) is whether holding a single stock will outperform a good alternative investment. In other words, the future cash flows to which an equity investor has a residual claim must be considered in light of the opportunity cost for investing in that individual company’s residual claim rather than in an asset that will generate a greater return.
Your second example, I find particularly interesting because of how an orthodox thinker might approach this problem of operational uncertainty. Let’s say that the highly-levered company can issue debt cheaply and let’s further say that from its covariance with the market and the going rate for risk-free debt, unlevered, its discount rate would be 10%. By increasing the amount of debt issued at a low rate, you are actually making the firm less risky from a WACC standpoint even though, as you point out, the operational outcomes of the firm become much less certain. This obviously does not make sense.
A firm’s value is the sum total of its future cash flows. Its future cash flows are uncertain and are dependent upon only a few drivers: revenue growth, profitability, investment level, and (related to investment level) investment efficacy in the medium-term. Financial leverage works to increase profitability when revenue growth is high, but works to decrease profitability when revenue growth is low or negative. My argument is that the leverage level of a firm creates uncertainty regarding the future cash flows of the firm, and that uncertainty is independent from the rate at which you choose to discount those future flows.
Operational uncertainty must be considered when assessing the capacity for the firm to generate cash flows. Opportunity cost must be considered when assessing the attractiveness of one asset against another. And assessing the attractiveness of one asset against another is exactly the role played by the discount rate.
If the market price of a stock is so low that – even in the worst operational case – an investment in the company would allow the owner to generate a greater return than if one had invested in an asset whose returns represent a common yardstick for equity returns, one should invest in the company.
I realize my view on this is unconventional, so I appreciate your taking the time to read my article and to make your counterpoints.
All the best,
Erik