A cross-border investment committee paper that quotes one country-risk premium and moves on has hidden three choices the committee was entitled to see. The first is what the number is trying to price: sovereign default exposure, equity-investor compensation, or operating cash-flow uncertainty. The second is which translation from credit to equity has been adopted and whether the modeller treats it as a theorem or a working convention. The third is whether country risk attaches to the country at all or to each operating company by its actual exposure. A country-risk premium is not a measurement; it is three assumptions stapled together, and the IC's only useful question is which of the three you have hidden from them.
Why a single CRP number hides the assumption the IC needs to see
Three institutional methods dominate the field. The Damodaran build-up takes a mature-market equity-risk premium, adds a sovereign default spread for the target country, and scales the spread by a volatility ratio intended to translate credit risk into equity risk [1]. The market-implied route substitutes a five-year sovereign credit-default-swap spread for the rating-based spread, then applies the same scaling. The McKinsey approach refuses to load country risk into the discount rate at all and instead expresses it as probability-weighted cash-flow scenarios discounted at a single global cost of equity [6]. The three methods do not produce small differences around a common answer; they answer different questions.
The IC's first job is therefore to fix the question. If the committee is pricing the chance the sovereign defaults on its obligations, the natural anchor is a credit instrument. If the committee is pricing the additional return an equity investor demands for the residual uncertainty after sovereign default risk is accounted for, the natural object is an equity-risk premium and a credit instrument is at best an input to it. If the committee is pricing operating uncertainty in a single multinational with revenues in several jurisdictions, the natural object is a company-level exposure measure rather than a country label. A paper that does not state which of the three is being priced has answered the wrong question precisely.
The build-up method: what the volatility ratio is actually doing
The Damodaran build-up is the most widely used construction in institutional practice and the most openly documented [1][2]. The mechanic is direct. A mature-market equity-risk premium is set, typically from the implied premium on the S&P 500. A country default spread is taken from the local-currency sovereign rating by mapping the Moody's rating to a CDS-implied spread by rating bucket. The default spread is then multiplied by the ratio of equity-market standard deviation to sovereign-bond-market standard deviation, and the result is added to the mature-market premium.
The volatility ratio is the load-bearing step. It claims that the extra equity-risk premium an investor demands for a country is the country's default spread scaled up by the relative volatility of its equity market against its sovereign bond market. In current practice the ratio is computed once for the asset class as a whole (the S&P emerging-market equity index against an iShares emerging-market sovereign-bond ETF) and applied universally across countries that lack a deep local sovereign bond curve [1]. This is a known simplification. The volatility ratio is not estimated country by country; one ratio is applied to many countries; and the substitution is openly acknowledged on the source methodology page.
Damodaran's own caveats belong in front of the IC, not buried in an appendix. He flags that the resulting equity-risk premium can fall below the underlying default spread when the volatility ratio is low, which is theoretically uncomfortable, and that the equity-bond volatility ratio is unstable across cycles [2]. The Kruschwitz, Loeffler and Mandl critique goes further: the bridge from credit spread to equity premium is not derivable from CAPM and rests on an analyst's working assumption rather than a theoretical result [5]. The honest framing is that the build-up is a convention with broad practitioner uptake, not a derivation. GIVE Analytics treats it accordingly: disclose the convention, run sensitivity around the ratio, and decline to defend the resulting figure as a derived quantity.
The practical consequence at the IC table is that the build-up premium should travel with at least two stress points. A high-volatility-ratio case, reflecting periods in which emerging-market equities sell off harder than emerging-market sovereign debt, widens the premium and tightens the equity hurdle. A low-volatility-ratio case, more common in benign global conditions, can produce an equity premium too close to the underlying default spread to look credible. The Fernandez survey of country-risk premiums used in practice [7] shows how widely practitioners disagree on the same country at the same date; that dispersion is itself the most direct empirical argument against a point-estimate framing.
The CDS route: continuous pricing with imported global noise
The market-implied variant replaces the rating-based default spread with a directly observed sovereign CDS spread, usually the five-year USD-denominated contract. The appeal is operational. CDS spreads update continuously and reflect the current market view rather than a periodically refreshed rating. For a committee that wants the cost of capital to move with the market the CDS route is the more honest instrument.
The cost is well documented. Bank for International Settlements work has shown that global risk factors, rather than country-specific fundamentals, dominate the cross-section of sovereign CDS pricing, particularly since 2008 [4]. The implication for the modeller is that the CDS spread carries a global risk-aversion component that has nothing to do with the country in question. Using it as a country-default anchor imports liquidity beta and global investor sentiment into the discount rate whether the modeller wants them or not. More sophisticated practitioners decompose the spread into expected-loss and risk-premium components before scaling; most do not, and an IC paper should say which version is in use.
The same volatility-ratio bridge then has to be applied to translate the CDS spread into an equity-risk premium. Substituting a continuously priced default anchor does not avoid the bridge question; it inherits it. The CDS route also imports a tenor choice. Five-year contracts are the most liquid and the conventional reference point; longer-dated contracts carry materially less depth and produce noisier spreads. An IC paper should state the tenor used and whether the contract is USD-denominated, since local-currency CDS where it exists encodes a different mix of default and currency risk.
The relative-volatility hybrid: cleaner in theory, unstable in practice
A third construction takes the mature-market equity-risk premium and scales it directly by the ratio of country equity-index volatility to mature-market equity-index volatility, with no sovereign anchor at all. The construction is theoretically cleaner because it removes the credit-to-equity translation question; it replaces it with a volatility-only question. Damodaran flags the method as empirically unstable [2], and the practical objection is that equity-index volatility is dominated by index composition and float, which differ across countries in ways that have little to do with country risk.
The relative-volatility method is useful as a cross-check rather than a primary construction. Where the build-up and CDS routes converge but the volatility method diverges, the divergence usually reflects market structure rather than country risk. Naming it as such is a more disciplined disclosure than averaging the three.
Lambda: the question most CRP users skip
A single country-risk premium applied uniformly to every firm in a country misstates exposure, because firms differ in how much of their revenue, cost base and operating capital sit inside the country whose risk is being priced. Damodaran's lambda framework names the choice [3]. Three variants are documented. A revenue-share lambda weights the country premium by the share of company revenue earned in the country. An accounting-earnings regression estimates lambda by regressing company earnings against country economic conditions. A price-return regression estimates lambda by regressing company stock returns against the country sovereign-bond return.
The point of naming lambda in front of the IC is not to insist on a particular variant. It is to demonstrate that "country" risk in a portfolio of multinationals is not a country label but an operating-exposure measure, and that an unlambda'd build-up applied across a diversified business is a precision claim the model cannot honour. A defensible cross-border build separates the country-level premium from each company's lambda, exposes the lambda choice on the assumptions register, and stress-tests the result against a uniform-exposure alternative. Lambda dispersion across the portfolio is often the larger source of misstatement, particularly for SWF and family-office allocations that combine listed multinationals with concentrated operating exposures in single markets.
The choice between the three lambda variants is itself a disclosure. Revenue-share lambda is the easiest to evidence and the easiest for the IC to read; it understates exposure where local costs, financing or regulation are the binding constraint rather than where revenue is booked. Earnings-regression lambda picks up that operating sensitivity but requires a longer earnings series than many portfolio holdings can supply. Return-regression lambda is the most market-consistent but is noisy at the single-company level. Naming which variant was used, and why, is more useful to the committee than the lambda figure itself.
The McKinsey objection: should country risk be in the discount rate at all
The valuation textbook position from Koller, Goedhart and Wessels argues that country risk is more honestly modelled in the cash flows than in the discount rate. The mechanics are scenario weighted. A base, upside and downside cash-flow path is specified for the target investment, with explicit probabilities reflecting the country's macroeconomic, regulatory and currency risks; the resulting expected cash flow is then discounted at a single global cost of equity [6]. The argument is methodological. A discount rate compounds; a country-risk premium loaded into the discount rate therefore penalises distant cash flows disproportionately, whether or not the underlying risk grows over time. A scenario specification lets the modeller place the risk where it actually sits.
The objection is not a knock-out. Scenario-DCF asks more of the analyst, depends on probability judgements that are themselves contestable, and produces a number that committees often find harder to challenge than a discount-rate adjustment. The serious institutional practice is to run both: build the discount rate with one of the CRP methods, build the cash flow with explicit country scenarios, and present the two answers side by side. Where they converge the conviction is defensible; where they diverge the divergence is the conversation. The discipline behind this is set out in the cross-border WACC build, which sits one layer up the stack from the country-risk premium itself, and in layered scenario architecture versus single-variable sensitivity, which addresses how the scenarios themselves should be specified.
What an institutional CRP disclosure should contain
The argument across the four constructions reduces to a short disclosure list. An IC paper that quotes a country-risk premium should state, on the same page as the number:
- which question is being priced (sovereign default, equity premium, or operating cash-flow uncertainty);
- which bridge convention has been used to translate credit risk into equity risk, and that the bridge is a convention rather than a theorem;
- what volatility ratio applies and at what frequency it is re-estimated;
- what lambda is applied to each portfolio company and how it was derived;
- whether the scenario-DCF cross-check has been run.
Where the construction uses a CDS spread the disclosure should add that the spread carries a global risk-aversion component. Where the construction uses a rating-based spread the disclosure should add the date the rating was last refreshed.
This is the standard GIVE Analytics applies on cross-border financial modelling and forecasting work, and the standard the firm sees adopted by the more deliberate SWFs and family offices in GCC and emerging markets where cross-border allocation is a continuous activity rather than an occasional question. The construction itself matters less than the disclosure. A CRP table without the disclosure list is not a number; it is a hidden assumption. The IC's job is to ask for the list.
Notes
1. Damodaran, A. (2026). Country Default Spreads and Risk Premiums. NYU Stern. https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/ctryprem.html
2. Damodaran, A. (2026). Equity Risk Premiums (ERP): Determinants, Estimates and Implications, 2026 Edition. SSRN working paper. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6361419
3. Damodaran, A. Measuring Company Exposure to Country Risk: Theory and Practice. NYU Stern methodology note. https://pages.stern.nyu.edu/~adamodar/New_Home_Page/valquestions/CountryRisk.htm
4. Amstad, M., Remolona, E. M., Shek, J. (2016). How do global investors differentiate between sovereign risks? The new normal versus the old. BIS Working Paper No. 541. https://www.bis.org/publ/work541.pdf
5. Kruschwitz, L., Loeffler, A., Mandl, G. (2010). Damodaran's Country Risk Premium: A Serious Critique. SSRN working paper, later published in Business Valuation Review (2012). https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1651466
6. Koller, T., Goedhart, M., Wessels, D. (McKinsey). Don't overthink your approach to valuation in emerging markets; chapter 22, Valuation: Measuring and Managing the Value of Companies, seventh edition. https://www.mckinsey.com/capabilities/strategy-and-corporate-finance/our-insights/dont-overthink-your-approach-to-valuation-in-emerging-markets
7. Fernandez, P., Garcia de la Garza, D., Fernandez Acin, L. (2025). Survey: Market Risk Premium and Risk-Free Rate used for 54 countries in 2025. IESE / SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5260463