ESG, Environmental, Social, and Governance, has become finance’s favorite shorthand for “future-proofing.” In theory, firms that manage environmental exposure, social relations, and governance quality should be safer borrowers: fewer regulatory shocks, fewer reputational blow-ups, cleaner compliance, better management discipline. That intuition has helped ESG migrate from corporate PR to investor decks, bank risk frameworks, and supervisory guidance.

But here’s the uncomfortable question: does ESG actually show up in credit risk metrics in a measurable way, especially in smaller, bank-centric economies? Croatia offers a useful testbed because the dataset pairs HGK’s ESG rating with HGK creditworthiness measures for Croatian firms in 2024–2025, giving a single-provider, internally consistent view (avoiding the classic “ESG ratings disagree” problem).

The headline from the report is not that ESG is meaningless. It’s subtler and more realistic: in this short window, ESG is not a dominant, standalone driver of credit ratings once you account for the basics, firm size and sector. ESG may matter, but it doesn’t behave like a simple lever that moves credit risk up or down in a clean, monotonic way.

A story of signals, and louder signals

Think of credit assessment as a noisy room. In Croatia’s data, the loudest voices are still the traditional ones:

• Firm size: Bigger firms tend to look safer in credit terms for familiar reasons, diversification, stronger buffers, more stable cash generation, better access to finance, and (often) stronger reporting capacity.
• Sector: Industry structure matters because credit risk is cyclical, regulatory, and exposure-driven. Utilities, energy-related activities, manufacturing, trade, each carries different cost structures, demand stability, and shock sensitivity.

Against those “macro” drivers, ESG behaves more like a second-order signal. ESG scores frequently co-move with size and sector: larger firms may have better governance frameworks, more formal policies, and greater capacity to document them. Some industries face tighter regulatory pressure and therefore invest earlier in ESG-relevant practices. In that sense, ESG can look like a proxy for institutional capacity as much as a proxy for risk reduction.

So the core empirical message becomes intuitive: once you control for the big structural factors, the incremental contribution of ESG is harder to detect, especially in a two-year snapshot.

Why “weak link” doesn’t mean “no link”

The report’s results are consistent with a key idea in modern finance: ESG can matter, but its impact is often conditional. In deep capital markets, ESG shows up more clearly through pricing (spreads, yields) and longer horizons where climate, regulatory, or governance risks have time to materialize. In bank-based systems, credit decisions can be relationship-driven and shaped by collateral, cash-flow history, and sectoral risk conventions. ESG might be considered, but not necessarily decisive.

Also, ESG is not one thing. Environmental improvements can look expensive in the short run (capex, compliance costs), even if they reduce long-run risk. Governance reforms may “pay” sooner (controls, transparency, risk oversight). Social factors can be real but slow-moving. In a one-year horizon, it’s entirely plausible that ESG changes don’t translate into immediate rating changes, because ratings are designed to be relatively stable and anchored in financial fundamentals.

The time-lag problem: sustainability isn’t quarterly earnings

A crucial economic point that often gets lost in ESG debates: many ESG mechanisms operate with lags. Decarbonization, supply-chain redesign, safety culture, compliance systems, board reforms, these take time to implement and longer to convert into measurable risk outcomes.

Credit metrics also have their own inertia. Ratings (even internal scoring systems) typically avoid overreacting to short-term noise. So the question “Does better ESG improve credit rating within a year?” may be the wrong benchmark. A fairer question is: Does ESG improve resilience across cycles and reduce tail risk over several years? Croatia’s data window is simply too short to let that story fully emerge.

What about directionality—could credit strength enable ESG instead?

Another advantage of the report is that it treats the relationship symmetrically. It does not assume “ESG drives credit.” It also asks: Do better-rated firms become better at ESG because they can afford it?

This “capacity-to-comply” channel is extremely plausible in emerging and mid-sized European economies. Better credit can mean cheaper financing, which can fund energy efficiency, compliance upgrades, reporting infrastructure, and professionalization. Yet even here, the Croatian evidence in this period remains muted once controls are included, again consistent with the idea that size and industry dominate the observable variation, and that two years is not much time for structural adjustments to show up.

A reality check from machine learning

The report adds a useful “non-econometric” complement: predictive modeling (random forests) to see whether ESG variables help predict HGK numeric ratings out-of-sample. The finding is pragmatic: prediction is modest, and breaking ESG into E/S/G does not substantially improve predictive power over total ESG.

This matters because it separates two questions:

• Inference: Is ESG statistically linked to credit outcomes once controls are included?
• Prediction: Even if not “significant,” does ESG improve real-world forecasting?

In Croatia’s dataset, ESG doesn’t emerge as a strong predictor in either sense, at least not in a way that would let banks or analysts treat ESG as a reliable shortcut for credit risk classification.

What should we conclude, without overclaiming? A sensible, finance-minded conclusion is:

1. ESG is not a credit magic wand. In the Croatian firm sample (2024–2025), ESG does not act like a dominant, standalone driver of HGK credit ratings once size and sector are accounted for.
2. Classic fundamentals still rule. Sectoral structure and firm scale remain the main pillars of observed credit differentiation.
3. ESG may still matter through longer horizons and specific channels. The absence of a strong short-run association does not invalidate ESG as a risk framework, it suggests measurement limits, time lags, and context dependence.
4. Don’t confuse reporting capacity with risk reduction. In smaller economies, ESG scores may reflect how well firms document policies as much as what they do. That can weaken the link with credit outcomes if credit scoring already captures fundamentals.
5. Better data would likely change the game. A longer panel, richer financial controls, and direct credit pricing variables (spreads, defaults, restructuring events) would provide a tougher test of ESG materiality.

Serbia box: why this matters even more next door

Serbia currently lacks comparable, systematic public research linking ESG-type scores to credit risk metrics at the firm level. That gap matters for three reasons:

• Regulatory convergence: As EU-linked supply chains and financing standards tighten, Serbian firms face growing ESG-related compliance demands regardless of domestic regulation.
• Bank-centered finance: If ESG risks are priced anywhere in Serbia, it will likely be through bank credit policies, making local evidence crucial.
• Policy credibility and capital access: Without transparent empirical work, ESG can drift into ideology (pro/anti) rather than becoming a measurable risk-management tool that improves access to capital.

A Croatian-style matched dataset (single provider ESG + credit metrics over multiple years) would let Serbia answer the question investors actually care about: does ESG add information beyond size, sector, and balance-sheet fundamentals, and over what horizon?

The bigger takeaway

The most useful message here isn’t “ESG doesn’t work.” It’s that ESG is not automatically a credit variable, especially in small, bank-centric economies and short time windows. Croatia’s evidence points to a world where ESG is real, but its footprint in credit outcomes is modest, conditional, and easily overshadowed by structural determinants.

If ESG is to become more than a buzzword in credit markets, the next step is clear: longer horizons, better financial controls, and outcome variables that reflect credit pricing and distress, not just categories. Until then, ESG is best read as a complement to fundamentals, not a replacement.