Saloni Ramakrishna (July 2026)
BTRM Faculty Opinion
“The Green Algorithm – Redifining climate risk in banking with AI”
Climate risk is a material financial risk. Extreme weather events, carbon policies, changing consumer and investor preferences, technological disruptions, supply chain vulnerabilities and growing regulatory scrutiny are impacting counterparty credit worthiness, collateral values, market pricing and an unexpected bouncer in the form of challenges in insurance availability. These effect revenues, capital, liquidity, asset quality, funding costs, and valuation, amongst other aspects. The unprecedented velocity, scale of climate change and its impact on all the known risk classes of financial institutions, demands a different and proactive approach from banks to keep pace with the complexity and inter-connectedness of climate-related financial risks.
The data canvas till recent past -Historical data, financial statements from counter parties, current data from its own books and market + rating data was largely the data canvas of banks – note that most of it was structural data. Modeling time horizons were short – I year horizon to a max of 3 years (Except strategic planning which sometimes went to 5 years).
What changed:
Fundamental change: Moving from depending on past to looking into future and anticipating how it will impact the books of the bank. With the landing of climate risk onto the risk topography, not as a separate category but as one that has an overarching effect on ALL risk classes, that landscape has changed significantly and how!!
Addition to the data classes: Financial statements of either counter parties or banks themselves carried little or no climate related data. Now the requirement sample covers unstructured data like Geospatial imagery, document-based data like the corporate statements, regulatory requirement of using forward looking scenarios for stress testing – to name a few.
Modeling horizon: Broader and deeper – short, medium AND long term. As climate and environmental risks challenge the system both on the time scales and the impact across the entire spectrum, modelling exercises are becoming more complex and multilayered.
Financial connectivity: Ability to connect the complex network of channels through which climate shocks propagate from real economy onto the financial services (FSI) and the intra industry effect of the interplay between the players of the FSI itself.
The Financial cost – Faster and harder crystallization of risks and the resultant impairment of portfolios affect both the balance sheet and the income of the banks. Uncovered, under managed climate exposures directly erode bank’s asset quality, increase stranded assets and impair cashflows that hit the income of banks. Creating and developing the required skill set to proactively identify climate Vulnerabilities is another big cost.
Why AI Augmentation:
AI deployed responsibly and transparently offers a transformational prospect. The real opportunity is in augmenting the system with AI capability. The “why and what” are not the questions – the “how” is. Guardrails are a must – a well-defined “Glass Box” construct is critical, so is the need to embed the same in the DNA of the organization.
Explainable AI frameworks, audit and training of data sets regularly, monitoring alerts, anonymizing sensitive information, clear definition and adherence to access rules, ensuring human in the loop validation, assignment of accountability and responsibility are some pointers towards building responsible AI Maps, architecture designing and operationalizing the same. Below is a sample of how AI augmentation brings value:
Redefining effective climate-related financial risks management at banks requires a shift from reactive compliance to proactive business strategy. Not just for consistency, coherence and context but equally for protecting an organization’s balance sheet and earnings – this requires a solid and grounded AI augmentation. Bottom line “green path and the algorithms” of the bank need to actively collaborate and harmoniously coexist for value creation.
******
Professor Saloni P Ramakrishna
