Stop reporting. Start steering.
Climate risk is now priced into the financial system. A technologist’s case for treating sustainability as a data and decision problem.
By Dr ir Rouzbeh Amini
I began my career designing spacecraft. Telemetry, in that world, has one job: to fly the machine. Nobody streams ten thousand signals from orbit to store them. Every reading shapes the next decision: correct the attitude, protect the payload, extend the mission.
A note on where I stand. I have spent the past decade building the data and technology layer under corporate sustainability, so this comes from the build side, not the podium. I am not a climate scientist, and this is not a sustainability lesson. I still look at it the way an engineer looks at any complex system: what are the signals? Can they be trusted? Do they reach the decision loop in time to change the outcome?
Judged that way, we spent a decade building one of the largest data-collection exercises in corporate history and pointed nearly all of it at the archive. I include myself in that. For years I built the machinery of disclosure and called it progress, and it took me too long to see we had never disclosed more and steered less.
That decade is ending, and not because of anything inside sustainability teams. It is ending because the financial system has begun to hard-code the risk.
The signal is now priced
Two things happened this summer, and the real news made neither headline. The European Central Bank began applying a climate factor in its collateral framework: the corporate bonds a bank pledges are now valued partly on the issuer’s exposure to transition risk. The more exposed the issuer, the less the bank can borrow against the same paper. Central banks move last, after an argument is over. This one ended in a collateral schedule.
One detail should stop any data professional cold. Alongside sector stress, emissions and transition plans, the ECB weighs the quality of a company’s climate disclosures. Weak data now carries a haircut, in the literal sense. The factor runs on forward-looking scenarios, not rear-view statistics; history holds no precedent for what is coming. Your reporting has a new reader: the models that price your paper.
Weeks later, on 2 July, the EU’s ESG Ratings Regulation put rating providers under ESMA supervision. Read the definition: a rating is an opinion built on a methodology. Europe licensed the opinionmakers and left the opinions as opinions. A board whose climate view is a letter grade bought from a vendor now knows what the law thinks it is worth.
So the pricing has begun: model by model, contract by contract, parameters rather than politics. You can win the ESG argument and still lose the haircut. It reaches revenue when a tender asks for a footprint you cannot produce; margin, as carbon pricing and border adjustments turn embedded emissions into landed cost; capital, as the ECB just showed; and operations, where volatility and supply concentration turn externalities into disruption.
These signals, climate or otherwise, surface among suppliers, products and customers before the accounts register them. Financial statements are lagging telemetry; risk arrives first. The board question is no longer belief in sustainability. It is whether someone else’s model already reads your exposure better than you do. The answer, in euros and probabilities by product line, customer portfolio and supplier tier, takes what no rating agency sells: your own sustainability and financial data, merged, at a quality worth betting on.
You can win the ESG argument and still lose the haircut.
Data doesn’t decide. Leaders do.
Here is the uncomfortable part, and I say it from the technology side of the table: the constraint is almost never the data or the tools. It is whether leaders let the signal into the room where money moves.
Insight creates value at one point only, the moment of decision. Pricing. Sourcing. Capital allocation. M&A. Everywhere else it decorates the strategy deck, and you can tell which you are in within ten minutes of a leadership meeting. In the companies pulling ahead, the CFO treats carbon as a cost curve to predict and manage rather than a number to assure, procurement weighs supplier emissions beside price and lead time, and commercial teams sell with footprint data because customers buy with it.
The bridge between data and value is not a platform. It is leadership: designing decisions so the signal arrives, is trusted, and gets acted on.
No resilience on a spreadsheet
None of it works while sustainability runs as a standalone layer, a parallel universe of spreadsheets bolted onto the enterprise. The biggest unlock is also the dullest, which is probably why it keeps being deferred. Sustainability data must be engineered to the standard of financial data: governed, integrated, traceable from source to statement. Call it the company’s second financial-grade substrate. That means wiring it through the systems where the business lives, ERP and procurement and logistics and the commercial stack, not rebuilding reality once a year for a report. Data architecture is business architecture, and when a central bank prices collateral partly on your disclosure quality, data governance is a treasury matter. The payoff is double: the foundation that makes disclosure cheap makes risk visible. You cannot manage a supply-chain exposure you never modelled, and cannot model it on data you never collected or do not trust.
From the rear-view mirror to the windshield
Build it, and something better than reporting becomes available. Carbon signals inside commercial conversations. Supplier risk flags inside sourcing decisions. Transition metrics inside investor dialogue, continuously rather than once a year.
That is where sustainability stops being an obligation and becomes a differentiator. It wins tenders competitors cannot bid for. Within a few years product-level sustainability data will move between companies as routinely as price and lead time, and the firms already treating it as infrastructure will set the terms. A report tells you where you have been. Intelligence tells you what to do next, and what portfolio you sell in five years.
AI is the accelerant, not the autopilot
This is where my own field changes the economics, and where most boards ask the wrong question. The models stopped being the bottleneck some time ago. The substrate is. And the prize is not automation but anticipation.
Running on trusted data, AI can simulate transition scenarios across thousands of products: the same forward-looking logic the ECB applies to collateral, turned on your portfolio before the market gets there. It can read supplier networks no human team could, absorb compliance drudgery consuming your most expensive experts, and put decision support in the hands of buyers and planners.
Feed those same systems fragmented, unverified data and AI does not fix the problem. It industrialises it, at speed, with total confidence and nothing underneath. The sequencing is unforgiving: foundationsfirst, intelligence second, autonomy last. Get it right and the advantage compounds. Get it backwards and you have automated your blind spots.
The agenda
I invite you to bring three questions to your next board agenda (see box). If the honest answer to the first is “nowhere”, the other two are already answered. In orbit, you never get to stop the satellite and recalculate. You steer in real time, with instruments you trusted enough to build properly. Business now runs the same way: continuous exposure, continuous signals, no pause button. The slow-reporting decade gave us the instruments. The fast one starting now belongs to whoever learns to fly.
Three questions for your next board agenda
In the last quarter, where did a sustainability signal change a commercial or capital decision?
Do we know our sustainability value at risk in financial terms, by product line, customer portfolio and supplier tier, or only our enterprise ratings?
Is that data governed to the standard we apply to financial data, and does a named leader own the convergence?
Key Takeaways
Reframe the board question: not “are we compliant?” but “what is our value at risk, in euros, by product line, customer portfolio and supplier tier?”
Watch the plumbing, not the politics: the ECB now discounts collateral for transition risk and weighs disclosure quality. Weak data has a price.
Treat ratings as what the law says they are: opinions. Decision-grade insight comes from merging your own sustainability and financial data.
Move the signal to the moment of decision: pricing, sourcing, capital allocation, M&A. Give one named leader ownership of the data convergence.
Sequence AI strictly: foundations first, intelligence second, autonomy last. The prize is anticipation, not automation.
Dr ir Rouzbeh Amini
is Global VP Sustainability Performance at dsm-firmenich, leading sustainability digital & AI strategy. He holds a PhD in Aerospace from TU Delft. He was EY EMEIA Sustainability Data & Technology Lead Partner and founded Cognizant’s Global Sustainability practice.