Sovereign Ratings and Risk Pricing, Agency Divergences in the European Union

Using annual EU-27 data for 1995-2024, we examine whether sovereign ratings mainly reflect common macro-fiscal fundamentals or whether agency-specific departures from that benchmark are also priced into sovereign funding conditions. Pooled ordered probits and a machine-learning diagnostic layer for nonlinearities and thresholds for Fitch, Moody’s, and S&P identify a stable set of core rating determinants centred on inflation, debt-to-GDP, current account balance, budget balance rule indicator, output gap, old-age dependency, and revenue capacity, while within-country variation is narrower and concentrated mainly in inflation, debt, and unemployment. The machine-learning analysis confirms that flexible models absorb nearly all systematic variation between fundamentals and ratings, validating the shadow-rating decomposition used in the market-pricing test. ECB-based bond-yield regressions show that both the fundamentals-implied shadow rating and the agency-specific deviation are priced in euro-area Bund spreads: a one-notch more favourable value of either component is associated with about 50 basis points lower spreads. Evidence indicates that this pricing effect strengthens as debt rises and intensifies further once debt exceeds 100% of GDP, while a shorter crisis-period interaction is directionally similar but less precise. Sovereign ratings therefore appear to combine a common fundamentals core with discretionary overlays that markets treat as economically relevant signals.