Application: emerging market risk premium calc

To all: Any one have any good ideas for a consistent way of calculating a company specific risk premium in an emerging market? CAPM is out the window because of skew/kurtosis… I am thinking g-sec yields+avg CDS spreads on industry+ X… Have no idea what to do about X… Thanks

Are you calculating a premium for equity, debt, sovereign debt? WACC? IIRC, there are slightly different methods for each. If equity (possibly debt too, but I’m not so sure), you might try a model that has risk premia for co-skew and co-kurtosis, but that’s pretty data intensive. I’d have to think that one through a bit to figure out how to specify it in a way that can be estimated.

Looking for a discount rate for time value of money (largely for DCFs)… I am finding info on using modified CAPM and solving for the company specific risk premium, but I don’t like it. I am also thinking about using the WACC as of the latest financials (usually some what stale), just to incorporate the company specific, zie, and marketability risks… and then adding maybe a country risk premium all to a g sec yield… country risk premium will be default spread* (stddev of equity)/(stddev of bonds)… I am unsure to what extent I am being redundant or making this too complicated…

> country risk premium will be default spread* (stddev of equity)/(stddev of bonds)… This was one of the ways to calculate a country risk premium given in L2. It basically assumes that the country risk for equities is proportional to the country risk for debt, and then scales it up using the relative volatilities of debt and equity as the scaling factor. It’s not bad as a first cut, but the assumption that equity has proportionately the same statistical properties as debt seems suspect to me. For one, bankruptcy law will affect the SD of debt, but probably won’t affect the SD of equity all that much, since equity will likely be wiped out in any case. You might use a CAPM using a local market factor and then a CAPM using a global market factor and then use a weighted average of those two results. The value of the correlation coefficient between local and global market factors might be the weight you put on the global CAPM result and use (1-corr) as the weight of the local CAPM. It’s not fully mathematically defensible, but it’s not necessarily a terrible alternative.