When doing WACC calculation, does anyone else think CAPM is a pretty shoddy way to value cost of equity. If a company is trading at forward 25p/e, and they issue new common. Than aren’t essentially the new buyers of equity willing to pay $25 for every $1 of earnings. Meaning a 4% rate for the company. So lets so say this 25p/e company , also has a beta of 3 (not unlikely) CAPM = 3% + 3(12%-3%) equals 30% cost of capital. Any thoughts on how this makes any sense…
The literature is full of critiques of capm. What do you recommend people use instead?
1/pe ? I know there’s problems here too, but for goodness sakes at least we aren’t using something like beta that really does have much to do with why i invest in a company or how ‘safe’ their equity is i’m looking at it from the companies perspective, they can raise money cheaper (4%) when their stock is trading at 25x versus it will cost them more (10%) when they are trading at 10x WACC and cost of equity calculations only
If you feel the value beta is shoddy you can calculate beta on your own. The books go into detail about how important key values like beta are in determining CAPM. P/E is ‘relative valuation’ - the value of the company compared with others unlike CAPM.
thanks, i’m going to try the adjusted beta, you’re talking about the b(2/3) + 1(1/3) right, the mean reverting beta idea (referring to CFA Level II book number 4(equity) page 115)
Lots of critiques of CAPM, out there, but empirically, the market factor explains some enormous proportion of variability in stock returns, so it’s not so bad as a first approximation. What you do to get a better approximation is part of the art of modeling. Size and value may have an effect, and presumably other things too.
Art of modeling, sounds like the art of data manipulation. I’m half way kidding, but seriously, I’m building a valuation model right now, and i used CAPM regular first, the results were ‘out of line’ , so usued adjusted beta, got better but not perfect. switched to wacc but some companies arent using debt in my comparison. Im sure theres no perfect way, but it seems to open to ‘choose’ the method that gives you the answer that you want. (which is what I’m doing right now)
> Im sure theres no perfect way, but it seems to open to ‘choose’ the method that gives you the answer that you want. (which is what I’m doing right now) Rather than being dishonest, an alternative is to run the valuation using several approaches and show them all to the client.
CAPM reminds me Black-Scholes model. Both are very useful despite lots of weaknesses.
This is why a lot of sell side valuations are suspect - freedom to choose the method that gives you the answer you want. On the buy side, it’s much more important to consider alternative ways of valuing stuff simultaneously and interpret why they may give you diiferent results, and come to a synthesis that makes investment sense. Remember that CAPM is really about looking at the entire market action to estimate how much return the average investor demands for taking on risk. It is assumed that investor don’t want to take on risk that doesn’t get paid (since this results in volatility drag). You then look at your company and observe how much risk it demonstrates historically, and then figure out what the average investor will require to assume that risk. Now if you have a better model that reduces risk or improves your prediction I earnings or stock returns, you actually have less risk than the average investor (to the extent that your model is really better), and so you can demand less of a return.
bchadwick Wrote: ------------------------------------------------------- > > Remember that CAPM is really about looking at the > entire market action to estimate how much return > the average investor demands for taking on risk. > It is assumed that investor don’t want to take on > risk that doesn’t get paid (since this results in > volatility drag). You then look at your company > and observe how much risk it demonstrates > historically, and then figure out what the average > investor will require to assume that risk. > So why the high correlation between betas and high p/e’s. A high beta would indicate high CAPM (demand more return), and buying at a high p/e (in the sense of paying 25 dollars for 1 dollar of earnings) would indicate to me a willingness to lend at low rates of return. I’m going to take the advice of showing several valuation models, CAPM WACC, AdjCAPM, Thank You!
I don’t know if there is a correlation between high beta and high P/E. Theoretically higher beta means higher (systemic) risk, so you should have a lower price to compensate. However, earnings has both systemic and company-specific risk involved, so they may not be fully comparable. If the data is from the recent bull run, it may simply be that in a bull market, people loaded up on high risk assets as a kind of leverage without borrowing. If you think growth is going to go on forever, a larger portion of the price is going to incorporate future earnings growth, and at the same time, people may want to increase their exposure to risky assets. So there’s presumably some interplay of psychological aspects between earnings expectations and willingness to take risk. I also would challenge your use of 12%-3%=9% market risk premium. That seems like an awful lot for the current environment (if you’re looking one year out), and also very high as a long term average (if you’re just assuming that it will be something in line with historical norms. You won’t get such a high rate of expected return if you adjust your MRP more in line with the current reality.
Just keep in mind that CAPM is one of many tools. You would never want to use just 1 tool when analyzing a security anyway. If you combine CAPM with other valuation factors it has more meaning than just using it by itself.
I would use the BYERP which is what Warren Buffett uses. Very much more practical than CAPM
What would be a more appropriate MRP and why? I used 12% Market return, up until '07 you could have argued that was a plausible 20yr average (just guessing here) And I used 3% RFR because well, the one year is less than that and i was trying to be conservative. Anybody know what the big shops,(CIO, mArket strategists) are using as their Market return number in this current craze…?
What’s BYERP, can’t find anywhere in CFA text?
I’ve never heard of anything higher than 7.5% When I worked in valuation consulting, we used 6%, and therre is plenty of research to justify that. Also: where are you getting 12% mkt return from? I’m pretty sure that’s well above S&P lt returns.
anybody remember FCFE/(k-g) from level 1, because the firms im valuing all have ROE > 15% it takes a pretty decent k to come up with any kinda of usuable number I’m gettting pretty good results with adjusted beta’s, whosever idea that was, thanks so what’s BYERP and why does buffett like it?
BYERP is the current yield on long term company debt plus a equity risk preium of around 3-4%. Buffett does not like CAPM. He has said he does not know what Beta is. BYERP is more based on fundamentals vs. MPT. BYERP = Bond Yield plus Equity Risk Premium
You would have to do a boatload of work on adjusting that long-term debt yield to convince me that cost of equity capital ought to be 3 + cost of long-term debt. In particular, cost of long-term debt is based on downside (you get defined cash flows unless the company goes bust). The cost of equity is supposed to be based on upside, i.e. how much are future free cash flows worth? I’m sure there is a universe in which these are similar, but it’s nothing like as general as CAPM. I would say that the 1/pe model is a DDM without growth explicitly factored in. If you like 1/pe you ought to like DDM-type models better because at least you can make some assumptions about growth. Anyway, everyone knows the CAPM is bunk beyond a rough first approximation. I personally do not feel there is lots of value in presenting three times as many bunky first approximations (the average is unlikely to be better than any of them for example) and it will make you look amateurish. You should do an anlysis with assumptions you think you can justify and then figure out the sensitivity of your analysis to your assumptions. That “analysis works if oil greater than 80” means that the analysis probably needs to include some cost of hedging that assumption or the sensitivity to it.