“To the extent that it matters for investors, climate risk is already priced in.”

I joined Grace Osborne on Bloomberg’s ESG Currents podcast for an Oxford-style debate with Jakob Thomae of Theia Finance Labs to debate this motion. Jakob argued that climate risk is priced in. I argued against.

My view is that while climate risk is increasingly being priced in, there is a long way to go. Supervisors are more sophisticated, disclosure has improved, and investors are paying closer attention than they were. But the gap between what is priced and what should be priced remains very large.

It is an important question to debate and was great fun to do!

Listen here:

🎧 Spotify - open.spotify.com/episod…

🎧 Apple -

Transcript of my remarks:

THE BURDEN

I have a great deal of respect for Jakob and for his work at Theia Finance Labs. But respectfully, the Proposition has set itself an impossible task today.

Let me explain why.

The motion is: “To the extent that it matters for investors, climate risk is already priced in.” To win this debate, the Proposition must convince you that all climate-related risks that matter for investors are already reflected in prices. Across all markets. For all investors. All the time.

My task is easier. I need only show that some material climate risks are not currently priced in, by some investors, some of the time. If markets are anything short of perfectly efficient on climate, the motion falls at the first hurdle.

That is the asymmetry at the heart of this debate. And I believe the Proposition has already fallen at that first hurdle. But to keep this interesting, let me unpack the case.

I am going to draw on the findings of the ECB, the Bank of England, and the Bank for International Settlements to show that the institutions that supervise the global financial system have reached the opposite conclusion to the motion. And I am going to use the Proposition’s own organisation’s research to make the case.

WHY MARKETS STRUCTURALLY UNDERPRICE CLIMATE RISK

There are structural reasons why capital markets are poorly equipped to price climate risk. These are not temporary inefficiencies that will self-correct. They are features of how markets work.

First, climate risk involves deep uncertainty. Not the kind of uncertainty markets handle well — not coin-flip uncertainty with known distributions — but ambiguous probabilities, contested payoff distributions, and non-linear dynamics. Where uncertainty is this deep, markets either attach the wrong prices or apply discount rates that bury long-tailed losses. That is not “priced in.”

Second, investment mandates, benchmarks, and career risk shorten investor horizons. A pension fund CIO who knows that coastal real estate will be impaired over twenty years still faces a benchmark measured over twelve months. Long-dated, foreseeable risks are systematically discounted away. And when the repricing does come, it comes all at once, because physical climate shocks and policy shifts hit many portfolios simultaneously. You cannot diversify the climate.

But these are theoretical arguments. Let me move to the evidence.

THE EVIDENCE: WHAT SUPERVISORS HAVE FOUND

Exhibit one. The ECB and the path to enforcement. The ECB’s 2022 climate risk stress test found that almost 80% of significant eurozone banks had either basic or no climate-related risk management practices in place. What followed was not reassurance that markets would self-correct. It was a multi-year enforcement escalation.

In March 2023, the ECB issued binding supervisory decisions to 28 banks that had failed to manage climate-related risks, with the threat of periodic penalty payments. In November 2025, the ECB issued its first climate-risk fine, against Abanca. In February of this year, it fined Crédit Agricole €7.6 million for the same category of failure. And in January 2026, the ECB formally embedded climate- and nature-related risks into its core supervisory and monetary policy functions.

If climate risk were already priced in, the supervisor of the eurozone banking system would not be issuing binding decisions, imposing fines, and restructuring its own operations to address it. The progression from the 2022 finding of widespread inadequacy, through binding decisions, to fines tells a story. Banks were not pricing this, supervisors told them to, and some still did not.

Exhibit two. The Bank of England and the measurement problem. The Bank of England’s Climate Biennial Exploratory Scenario found that climate risks would create a persistent drag on profitability, with nearly half of projected mortgage losses concentrated in just 10% of postcode districts. But the most important findings were about capability. Different firms estimated loss rates for the same counterparty that differed by a factor of ten. Let me repeat that. A factor of ten. For the same borrower.

The BoE found widespread data gaps in information about the location of corporate assets and in standardised data on value chain emissions. Its 2023 follow-up concluded that firms’ capabilities are not sufficiently well developed to support the internal calibration of capital. If the UK’s largest banks cannot agree on the magnitude of climate risk for the same borrower within an order of magnitude, the claim that the risk is “already priced in” is not credible. You cannot accurately price what you cannot accurately measure.

Exhibit three. Sovereign bonds and physical risk. BIS Working Paper No. 1275, published last year, examined climate risk in sovereign yields across 52 economies over two decades. It found that transition risk is associated with higher sovereign yields, particularly for high-emitting countries after the Paris Agreement. But high-temperature anomalies, a key physical risk indicator, do not appear to be priced into sovereign borrowing costs at all.

Sovereign bonds are the bedrock of the global financial system. They are the benchmark for interest rates. They influence the pricing of every other asset class. They are the most liquid, most scrutinised securities on the planet. If physical climate risk is not in sovereign yields, the absence propagates through the entire pricing chain. If it is not priced in the most analysed asset class in the world, the motion cannot hold.

THE PROPOSITION’S OWN RESEARCH

And here is the tension at the heart of the Proposition’s case.

Jakob’s own organisation, Theia Finance Labs, published research — as part of the 1in1000 initiative — showing that once you incorporate climate tipping points, ecosystem decline, and social risks into financial stress scenarios, the losses in equity markets from climate change could be amplified by a factor of 2.5 to 3.5 times compared to standard estimates.

Let me say that again. Jakob’s own shop says that standard models may be underestimating climate-related financial losses by a factor of 2.5 to 3.5.

If that is right, and the Proposition’s own organisation published it, then current market prices, which are informed by those standard models, are also underestimating the risk by that factor. That is not a marginal discrepancy. That is a fundamental challenge to any claim that climate risk is already priced in.

CLOSE

So where does that leave us?

The Proposition must defend a world where every material climate risk is already in the price, for every investor, everywhere, all the time. That is not the world we live in.

We live in a world where the ECB has moved from guidance to fines because banks were not managing climate risk. Where the Bank of England found that its largest banks’ risk estimates differ by a factor of ten for the same borrower. Where the BIS found that physical climate risk is not reflected in sovereign yields at all. And where the Proposition’s own organisation says standard models underestimate climate equity losses by a factor of 2.5 to 3.5.

There are also climate risks that not even the best research groups in the scientific community fully understand or properly model. Tipping points, cascading hazards, compound events. You cannot accurately price a climate tipping point. There are known unknowns and unknown unknowns.

It only takes some risks, for some investors, some of the time, to be mispriced for the motion to fall. That is the asymmetry of the burden in this debate, and it is why you must reject this motion.