NOBODY DOES THIS! NOBODY! Use 5-6% as your ERP and call it a day. If you ever work on Wall Street and try to use CAPM you’d be laughed at…Heck scrap your dumb model all together and just use a discount rate that makes sense.
I’d rather use a higher discount rate to really stress test a potential equity holding to give me a high margin of safety than do any of the nonsense that you’re talking about. What you are doing is “Mental Masturbation”. You will spend a lot of time trying to get the right numbers only to have worse results. Do you think Warren Buffet is sitting here Jerking his brain trying to figure out what the right ERP is?
Nobody does this therefore it’s not appropriate. Wall Street is right more often than not. 5-6% ERP is sustainable in a mature market. Make use of an arbitary discount rate and adjust as nessecary till your new investment idea makes sense. Warren Buffet used discount rates with Coca Cola.
I’m not saying that it’s wrong because nobody does this. I’m saying that nobody does this because it is a complete waste of time. Also, I don’t disagree that discounting cashflows is a valid valuation method. Most equity analyst do use DCF valuation models. What they don’t do is what you are proposing. I am just saying that what you are proposing is akin to mental masturbation. It will not yield any outperformance becuase you determined the ERP is 4.2375% instead of 5%. You also never told me what is the exact time period over which to determine beta?
Using the term structure of interest rates to value bonds makes complete sense, but it does not make sense for equity where there are many outlying factors that can derail your valuation. Why don’t you put your money where your mouth is and let us know how things workout? Sounds about right.
It is not a complete waste of time. I’m already done with it and it took me 5 working days starting from scratch. The output turned out quite simillar in terms of stock valuation prices compared to the common dirty method, but less so with companies with little gearing. Overall, the main advantage of modeling a dynaminc ERP and RFR that matches each cash flow’s duration is it’s adaptibility to sensitive market risk factors through the relative volatility and spread margins of soverign bonds and the credit default swap market, to which I’m using as proxy for the build up approach.
As for the US’s ERP, I’ve merely used 5% as a starting point, since that’s pretty close to the current implied cost of equity premium. And converging over time towards the stable return premium of 3.8-4%. Sounds realistic.
What do you mean the exact time period which to determine beta? I don’t remember mentioning beta anywhere in this thread. My forecast period would typically be comprimised of a 9 year explicit period, followed by perpetuity. The stable period beta would typically fall between 0.8-1.2, but the dynamic total cost of equity is more complicated than that, and most likely needs adjustment beyond the CAPM disseration, such as tax adjustments, liquidity premiums, capital structure changes, and a quantitative sector/idiosyncratic variable where applicable.
I mean, what, according to you, is the appropriate time period to regress the equity returns versus the benchmark to determine beta? Is it 6 month, 1 year, 3 year, 5 year, 10 years of historic returns? And do you use daily, weekly, monthly returns? All of these will give you a drastically different answer which will then give you a drastically different discount rate.
That’s a subjective issue and needs analysis on a case-by-case basis. But generally, you’d use a bottom up industry/sector beta and relever. The statistical choice of an acceptable beta should be examined based on the distribution, and the fundamentals of your data. The regressions used to derive the industry betas should generally match the time period where the fundamentals of the current underlying business has not changed significantly, and where the stock’s liquidity is not a limiting factor.
A stable period beta would fall between 0.8-1.2, since we assume a diversified mix of assets on the long term.
A terminal period is a perpetuity. Best practice would be to discount the stable period inflow by the cost of capital so it assumes either no growth, or growth that creates no value. In some cases, there can be a case for a strong economic moat to earn slightly above the WACC in perpetuity, at a growth rate comparable to the long term outlook of the sector (global sector if a lambda is significant), or roughly the macroeconomy.
Anyway the model output page looks something like this, data is real ltime:
There is a big difference between what financial literature, including the CFAI material, recommends for common practice, and what is the proper application methods in theory. However, theory is much harder to practically apply, so they teach you a one size fits all that is proven to be roughly accurate.
Although I’m not aware what you exactly meant by that statement.
For the ERP in a market like Egypt, I might take a look at the long term volatility of returns for as long as you can get them, compare that to the long term volatility of a market like the US over the same historical period and then scale up the developed market ERP by the ratio of the volatilities. I don’t remember if that’s how the CFA does it or just how years of doing these kinds of things sits in my mind.
Probably it makes sense to repeat this for several developed markets to get a sense of the range of ERPs you get out of it, and then take some kind of weighted average based on the average size of these developed markets over the time period.
You can also try doing this with a simple beta regression which answers the question: how much would I have to lever up or down the developed market to get an average return similar to the the developing market, and how much volatility would that give me if I tried to replicate returns this way, and how much extra volatility would I take on by not replicating and instead investing directly in the emerging market. I generally don’t like the beta approach and prefer the volatiliies approach, but sometimes it generates useful insights, particularly if the developing market is concentrated in a few industries.
Whether you add the U.S. treasury rate or the Egyptian equivalent is a bit of a judgement call and depends on whether USTs are a real investment alternative for your clients or not. My guess is that serious investors do have the option to invest in USTs and Egyptian government bonds have sovereign risk in it so I’d be inclined to use USTs.
You probably want to add a premium for liquidity issues and maybe the sovereign spread makes sense to add on to reflect country risk. (Which would make it ok to use Egyptian govt bonds as a build up complement to the ERP even if they aren’t technically risk free themselves). To some extent, the liquidity issues should reflect themselves in the volatility of the index,but tail event liquidity issues are underrepresented and would require some (probably-subjective) premium on top of it.
