Hi everyone,Here's a cross post of a couple of posts over on the MDP: Activision board, where I lay out some more of my thinking on choosing a discount rate.Cheers,JimWhy do people (myself included) use discount rates of 10 - 15%? I understand the formula using beta, WACC, risk free rate, etc., but I don't believe in beta (a measure of volatility and not risk in my book) and also, I'm trying to think of this as if I were buying a business.Great question and, as always, the answer is "it depends."There are several schools of thought on it. But basically I think they can be broken down into two viewpoints. What is the risk and what is my return?One of the "what is the risk?" views uses CAPM, WACC, beta, risk free rate plus equity risk premium, etc. That adjusts the cost of equity (part of WACC, the other part being cost of debt, as I'm sure you're aware) by some risk multiplier (beta). That beta can be the volatility of the stock versus an index, but the value is highly dependent on time-frame sampled, time of day/week/month sampled, number of samples, and index for comparison. Very flawed, in my book. Betas also exist for country risk and currency risk, among other things. Can get pretty hairy and can lead to a false sense of precision. You could do that with your own risk multiplier (beta), based on a judgment of the riskiness of the investment, say multiplying the equity risk premium by 3 for a biotech, 2 for a risky grower, 1 for a mid-cap blue chip, 0.5 for a mega-cap blue chip, or whatever works for you. (This makes more sense to me than using beta from historical volatility which is what CAPM does.)Another "what is the risk?" view is more along what you outlined. This looks at the stability of the company, the competition, and the likelihood of those future cash flows and adjusts the cost of equity up or down based on a judgment of those risks. A small biotech might use a 15% or 18% (or even higher) discount rate, while a big, stable business like Procter & Gamble might use a 8% or 10% discount rate. If the investment is more like a bond (maybe that cafe you described), then a 7% discount rate might be fine.The viewpoint I prefer is the other one. "What is my return?" This is more like a hurdle rate. I want my investments to return to me some amount (on average) per year, or higher. Personally, I use 15% because I want to double my investments every five years (outside of adding more funds). So, I often use that as the discount rate -- my required rate of return. I might adjust it up or down a bit for individual positions (using thinking similar to that outlined immediately above), but overall, I want the portfolio to bring in 15% or more, on average, per year. Others might use 12%, wanting to do slightly better than the long-term return of the S&P 500. But I tend to use a consistent discount rate no matter what I'm looking at. Of course, that causes me to pass over a bunch of companies.Which approach you choose is more a matter of style and investing philosophy than being "right" or "wrong." However, understanding what is done and why is important so that you can intelligently choose what will work for you.Will you get 15%? When can the investor expect it?IF (and that's a big if) the company performs as modeled, and IF the market is rational (and you can believe that if you want), then the stock price SHOULD go up by the discount rate each year.See all the conditionals I put in there? :-)Trouble is, companies hardly ever perform as modeled, the market is not completely rational -- not even for a big cap like Microsoft -- and stock prices do their own thing over time, but generally rise. But when that rise happens is not determinable.The probability is that the stock will give that return, on average (that's the key), over several years ("several" being kind of open-ended). For instance, if the stock remains flat for four years and then doubles in the fifth, that averages out to 15% per year compounded. Not very smooth at all. But it's only a probability. The stock could end up doing nothing for a very long time or even go down as new information comes to light.What I try to do is review the purchase thesis every now and then and check to make sure I haven't missed anything. Following the company, skimming its news, reading the quarterly earnings releases and listening to (or reading) the conference calls, helps in this regard.
Thanks, Jim. Extremely helpful.
The viewpoint I prefer is the other one. "What is my return?" This is more like a hurdle rate...Of course, that causes me to pass over a bunch of companies.Yes, exactly, you pass over some good investments but what is worse is that this method will select the riskiest stocks. That can be very dangerous. Your screen will pick all the value traps and then you have to manually sift out the ones that are not "traps".I believe you have to account for the riskiness of the investment. You make good critique points about CAPM and beta. I think your "judgment beta" is better than completely ignoring relative risk.tj
Hi tj,Thanks for your comments.Yes, exactly, you pass over some good investmentsOh, certainly. But I'm not trying to find all the good investments. Honestly, nobody can. I'm trying to find enough that the market thinks very poorly of that actually are likely to do better than expected, and then ride the share price up as the market readjusts its expectations. If I manage to pick only winners (defined by rising stock prices after purchasing), nobody will be more surprised than me.but what is worse is that this method will select the riskiest stocks. That can be very dangerous. Your screen will pick all the value traps and then you have to manually sift out the ones that are not "traps".Yes, that is a concern. You always have to ask, "Why is this company disliked now?" And you can find yourself in the position of convincing yourself that it's "better than that, really!" That's where the due diligence comes in, asking questions about the accounting, the management, and the situation the company is operating in (e.g. Transocean). And that's also where my position sizing will play a role. For those that are most questionable, but still deemed good enough, a 2% position is all they get. For those that are pretty good, 4%. For those that I believe are great, 6%. So when I'm wrong, hopefully it won't cost too much.I believe you have to account for the riskiness of the investment. You make good critique points about CAPM and beta. I think your "judgment beta" is better than completely ignoring relative risk.Well, my standard 15% discount rate is already pretty pessimistic. However, if the company lives down to market expectations and only manages to produce free cash flow at the priced-in depressed level that puts it onto the watchlist in the first place (like the screened stocks do -- http://boards.fool.com/screen-results-28942797.aspx), then I'll expect to get an average of 15% return per year.But there's another issue even when using a judgment beta applied to the cost of equity that I hadn't discussed yet. Here's the equation to remind us:cost of equity = risk free rate + beta * equity risk premiumRight now, the risk free rate (commonly, the yield on 10-Year Treasurys) is pretty darn low.Second, what equity risk premium (ERP) should we use? I've seen papers where it's been measured to be anywhere from 4% to 6% or so. Apply a 2.0 multiplier to that and that's a 4-point swing between the top and bottom of the range, which really affects the calculated intrinsic value. Why use 4% or 5% or 6%?Not only that, but I seem to remember reading recently that the ERP has been declining for the past 20 years or more, based on what investors have been willing to accept. Why? And how does that historical view play out when we pick one that looks forward to use in our discount rate?For instance, suppose we assign a 1.5 multiplier to an ERP of 4% and use a risk free rate of 3.3% (the current 10-Year Treasury rate). We get a cost of equity of 9.3%. If the company has any debt at all, the resulting WACC (which is supposed to be the discount rate) is almost certainly less than that because debt is usually cheaper than equity. And that's for a company that we've judged to be more risky than the alternative (that 1.5 multiplier).OK, so then we "correct" it and say, no we should use a 6% ERP (resulting in a cost of equity of 12.3%). How is that decision to arbitrarily use a 6% ERP instead of a 4% ERP any different from just sticking a discount rate on the thing, as I do, and going on from there?Here's another way to look at the same issue: If I want my portfolio to return 15% per year or more, which I do, what benefit do I get from using a discount rate less than 15% in determining what expectations are priced in for a given stock? Remember, discount rate and expected return are mathematically identical, just looked at in two different ways.Another aspect of this issue is, if I use a 15% discount rate, does that mean I will automatically invest in the riskier companies, as your first comment implies? I don't think so. One, I'm not estimating growth (a source of error when generating the cash flows to be discounted) and then applying a discount rate to it (another possible source of error, compounding any errors from the first bit). I'm letting the market tell me what is expected for a given discount rate at a given price.The question I need to answer, then, is just, "Is that growth rate reasonable, given what I know about the company?" In today's article on Microsoft, the answer I come to is, "Yes." http://www.fool.com/investing/value/2010/12/13/im-not-buying... And so, I decided not to invest.Two, high discount rates have been traditionally applied to riskier companies, but there's nothing that says we can't use high discount rates to help determine an acceptable entry price for less risky stocks. Unusual maybe, but not against the rules (what rules?).Three, I'm using behavioral investing to inform my decision. Is there disgust or ennui or despair surrounding this company? Look at the articles being written about it. Emotions like that tend to be shorter lived (though not always -- look at Microsoft over the past few years) and can often lead to mispricing.Regardless, I'd love to read your or others' further thoughts on this issue. Maybe I'm doing something completely wrong.Cheers,Jim
Jim,I think your comments about ERP are spot on. And I also agree that DCF analysis is a minefield of potential error compounding.I don't think there is a chance that you are doing anything completely wrong. Reason is that you are accounting for risk outside the DCF with "behavioral investing" and broader analysis. For example, you dumped financials from your screen results. Personally, I really like that angle about disgust/ennui, because we know from behavioral studies that the human emotional pendulum has a wide arc. I see some of that in health insurance, solar (esp Chinese), military (MANT, SAI, OSK), drillers after BP, semiconductors, and others. The challenge is to figure out how much of the disgust is excessive...and MUE is one place I visit for that!Best,tj
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |