Issue: September 2026 · Market data: as of 31.08.2026 (UST – close of 28.08.2026) · Inflation month: July 2026 (latest published monthly data) · Updated monthly · CIRA analytical data
| Currency | Risk-free rate | How it is derived |
|---|---|---|
| US Dollar | 4.4% | 10-year UST 4.73% − US CDS 0.33% |
| Euro | 3.3% | 10-year Bund 3.32% − 0 (Aaa/AAA) |
| Hryvnia (derived from the dollar)* | 8.7% | Fisher: (1 + 4.40%) × (1 + 7.7%) / (1 + 3.4%) − 1 |
* A 10-year dollar base adjusted by the current inflation differential of the two currencies; updated with every new inflation print. In the headline tables values are rounded to 0.1 p.p.; calculations are performed on unrounded values.
| Tenor | Government bonds, weighted average yield (auction 18.08.2026) | Risk-free rate (UAH) |
|---|---|---|
| ~1 year (UA4000239016, 329 days) | 15.17% | 9.2% |
| ~1.8 years (UA4000239008, 658 days) | 15.63% | 9.7% |
| ~3.5 years (UA4000239081, 1274 days) | 16.47% | 10.5% |
The internal curve ends at ~3.5 years: no hryvnia market exists beyond that tenor, and no point of this curve can be presented as a 10-year rate. We publish it nonetheless – for capital inside the hryvnia circuit it is the working benchmark at the tenors that exist. At the auction of 25.08.2026 no issues with administratively distorted yields were placed – every point on the curve is a market point.
US Dollar.
| Step | Component | Value |
|---|---|---|
| 1 | 10-year UST (US Treasury par curve, close of 28.08.2026) | 4.73% |
| 2 | − US CDS (5-year, 31.08.2026) | 0.33% |
| 3 | = USD risk-free rate | 4.40% |
Euro.
Euro. The 10-year German Bund yield (Aaa/AAA, zero credit spread; Germany's CDS of 7 basis points (bp) is negligible): r_f EUR = 3.32%.
Hryvnia – the international derivative.
| Step | Component | Value | Source |
|---|---|---|---|
| 1 | USD risk-free rate (calculation above) | 4.40% | |
| 2 | Actual USD inflation (US consumer price index (CPI), July 2026, y/y) | 3.4% | BLS |
| 3 | Actual UAH inflation (CPI, July 2026, y/y) | 7.7% | State Statistics Service |
| 4 | Fisher formula: (1 + 4.40%) × (1 + 7.7%) / (1 + 3.4%) − 1 | 8.74% |
The simple-difference approximation: 4.40% + (7.7% − 3.4%) = 8.70%.
Hryvnia – the internal circuit.
| Step | Component | Value |
|---|---|---|
| 1 | OVDP UA4000239016, ~1 year (auction 25.08.2026, weighted-average yield) | 15.17% |
| 2 | − Sovereign credit spread (local-currency rating CCC+ ≈ Caa1; source 7, update of 01.07.2026) | 5.97% |
| 3 | = UAH risk-free rate (internal, ~1 year) | 9.20% |
Cross-check of the internal rate.
| Component | Value | Source |
|---|---|---|
| Inflation (CPI, July 2026, y/y) | 7.7% | State Statistics Service |
| Real GDP growth (Q2 2026, y/y) | +0.6% | State Statistics Service |
| Macroeconomic estimate of r_f (UAH) | 8.3% | |
| Market calculation (step 3 above) | 9.20% | |
| Difference | 0.9 p.p. |
Both routes here are backward-looking: a bond yield looks 12 months forward, while year-over-year CPI looks 12 months back. With forward-looking inputs the estimate is higher – household inflation expectations (9.95% for the next 12 months, NBU survey, July 2026) would give ~10.5%; this gap measures how much future disinflation the OVDP market has already priced in – and the surveys have not.
Issue commentary. This month the international derivative (8.7%) and the internal one-year point (9.2%) differ by 0.5 p.p. The inflation block has not been updated: July 2026 is the same month as in the previous issue, because fresher CPI data for both currencies are released after the cut-off date, and the methodology requires the same month for the hryvnia and the dollar. The internal rate rose from 8.82% to 9.20%, and almost all of that move is non-market: the switch to the current edition of the default-spread table (6.37% to 5.97%) accounts for 0.40 p.p., while the OVDP yield itself took off 0.02 p.p. (15.19% to 15.17%). The long end of the hryvnia curve has lengthened from ~2.5 to ~3.5 years (a market point at 16.47%): the horizon is wider, but still nowhere near ten years. Going forward, the gap between them is the informative object: the derivative moves with the inflation differential, the internal point with the OVDP market, and the distance between them will show how much disinflation the domestic market has already priced in. For long-dated hryvnia cash flows (the terminal value in a discounted-cash-flow (DCF) model), the international derivative serves as the anchor; the choice of the long anchor is a modelling decision to be disclosed in each specific valuation.
The risk-free rate is the return on an investment whose outcome the investor knows in advance: the issuer cannot fail to pay (no default risk), and the instrument's term matches the investor's horizon (no reinvestment risk). It has only two components: the real reward for postponing consumption – the price of giving up money today for tomorrow – and compensation for the currency's inflation. The first is roughly the same everywhere by its nature; the second is each currency's own. So in essence there is one risk-free rate, and what separates currencies is inflation: the difference between the hryvnia and dollar rates is first of all the difference in their inflations. Two practical rules follow: cash flows are discounted at the rate of the currency they are denominated in, and the rate of one currency can be restated into another through the inflation differential – which is exactly how the hryvnia derivative in Table 1 is built.
The entire cost of capital stands on the risk-free rate: the cost of equity is the risk-free rate plus the equity risk premium (ERP) and the country risk premium (CRP); the cost of debt is the risk-free rate plus the borrower's credit spread. An error of one percentage point (p.p.) at this foundation shifts a company valuation by 5–20%, depending on the level of rates and the growth rate (for a growing perpetuity the relative value change per 1 p.p. of the rate ≈ 1/(k − g)) – which is why we publish the benchmark regularly and show the calculation in full.
The NBU's currency restrictions have made Ukraine a separate financial ecosystem. NBU Board Resolution No. 18 of 24.02.2022 (paragraph 14) is built on the principle of "everything is prohibited except what is permitted": cross-border transfers are allowed only for a defined list of operations, and until August 2026 the purchase of foreign securities was not on it. Resolution No. 90 of 10.08.2026 (in force since 11 August) opened the purchase of foreign securities to individuals for the first time since 24 February 2022 – within a monthly limit of UAH 200,000, roughly $4,000–5,000: a household channel, not an institutional one. Building a position in US Treasuries (UST) or German government bonds (Bunds) from inside Ukraine is therefore practically impossible – external risk-free instruments are unavailable to hryvnia capital, and Ukraine's normalised Chinn–Ito capital-account openness index equals zero (latest available observation; the database is published with a lag of several years).
The hryvnia risk-free rate therefore has two distinct definitions, and we publish both:
These are not two conflicting estimates of one indicator but different objects: the first answers "what a long hryvnia rate would be in an open world", the second – "what it actually is inside the closed circuit at the tenors that exist".
US Dollar. The 10-year US Treasury yield minus the US CDS spread (credit default swap – the market price of insurance against the issuer's default; it is what measures the residual credit risk). The United States no longer holds an Aaa/AAA rating from any of the three global agencies, so USTs are quasi-risk-free and this small credit risk is removed.
Euro. The 10-year German Bund yield: Aaa/AAA rating, zero credit spread.
Hryvnia — international derivative. No direct 10-year hryvnia instrument exists, so the hryvnia rate is derived from the dollar rate via the Fisher relation – through the difference of the two currencies' inflations (both readings from the same month). We use actual year-over-year inflation, not forecasts or surveys: actual values are observable, leave no room for subjective assumptions, and are precisely what lets the page update with every new inflation print. The cost of this choice is disclosed in "Limits of the method".
Hryvnia – the internal circuit. The closest domestic instrument to risk-free is hryvnia domestic government bonds (OVDP): government securities that the Ministry of Finance places at primary auctions. But even in its own currency the sovereign is not a risk-free borrower, so the sovereign's own credit (default) spread is removed from the OVDP yield:
r_f (UAH, internal) = yield on hryvnia OVDP (~1 year, primary auction) − sovereign credit spread by local-currency rating
With no 10-year point available, the benchmark is the ~1-year issue – the shortest clean market point (excluding issues with administratively distorted yields), comparable in horizon to the cross-check. The credit spread is taken at the sovereign's local-currency rating from the default-spread table (source 7): S&P CCC+ and Fitch CCC+ correspond to Caa1 on Moody’s scale. We rely on the S&P and Fitch ratings rather than Moody’s Ca for one reason – recency: Moody’s has not reviewed its assessment since May 2025, whereas S&P and Fitch upgraded after the debt restructuring was completed (December 2025 – January 2026); moreover, the spread at the stale Ca would produce a "risk-free" rate below inflation.
The subtracted spread is the same rating-based quantity that returns to the cost of capital as a separate term (the sovereign credit spread in the cost of debt, the country premium in the cost of equity): subtracting it from r_f and adding it back in the premia offset each other, reproducing the observed yield – there is no double-counting of country risk. The resulting number is a computed anchor for building the cost of capital, not a yield any investor can "buy" in the market.
Cross-check. Each month the internal rate is checked against an independent bottom-up estimate: inflation (CPI, year over year) plus real GDP growth of the latest quarter – the same "inflation + reward for postponed consumption" logic, with the reward approximated by the economy's growth. It is a check, not a standalone estimate; a material divergence between the routes is explained in the issue commentary, never ignored.
.png)