Norwegian green bond growth showed no clear link to local power prices
An arXiv preprint finds that green bonds became a larger share of Oslo Børs listings, but detects no statistically significant link between local electricity prices and issuers’ bond choices.
AI-generated conceptual illustration; not documentary evidence.
Did higher electricity prices during the Nordic energy crisis push borrowers toward green bonds? Phan and colleagues examined Norwegian bond listings and found no statistically significant evidence of such a relationship, although their study could not rule out smaller effects.
What the study found
In an arXiv preprint, Phan and colleagues studied bonds listed on Oslo Børs from 2019 to 2025. They combined European securities reference data, including matured bonds, with Euronext green bond classifications and issuer registry information. They then matched issuers to electricity bidding zones and analyzed bond choices using issuer and quarter fixed effects.
The authors report that green bonds rose from 15% of new listings in 2019 to 48% in 2024. But they found no statistically significant evidence that higher local electricity prices increased the likelihood of choosing a green bond over a conventional bond. That finding held across price-exposure windows of three to twelve months and for utilities and other issuers.
Why it matters
The results suggest that the growth in green bond listings was more closely associated with broader market trends than with local differences in electricity prices, according to the authors. The study also found no clear association between carbon prices and green bond choice.
Limits and unknowns
The main estimates rely on 21 identifying issuers. At 80% statistical power, the minimum detectable effect was 3.1 percentage points for a €10/MWh increase in electricity prices; smaller effects cannot be ruled out. The estimate for issuers with an earlier green bond was larger but imprecise, and the carbon-price analysis covered only eight utilities. The abstract does not provide the full effect estimates, confidence intervals or further details of the dataset and analysis.