Independent Researcher · AI-Driven Climate Financial Risk
Measuring how climate risk is repricing American mortgage credit.
I apply causal machine learning and explainable AI to loan-level mortgage data and federal hazard records to quantify a risk that credit models don't yet price, and what that means for lenders, borrowers, and policy.
22M+U.S. mortgage records analyzed
2,827U.S. counties
8Research papers
2Program committees
About
From the credit desk to the climate question.
I'm an independent researcher based in Columbus, Ohio, working at the intersection of climate risk and mortgage credit. My research uses causal machine learning, Double Machine Learning and Causal Forests, together with explainable AI methods (CatBoost, SHAP, LIME, DiCE) on Fannie Mae and Freddie Mac loan-level data matched to FEMA hazard records, to measure how flood and disaster exposure translates into default risk across U.S. counties.
Before research, I spent five years inside bank credit operations and operational risk at Emirates Islamic Bank in Dubai, where I was twice recognized with the bank's GEM Award for operational excellence. That practitioner grounding shapes the questions I ask: not whether climate risk is real, but whether it is priced, and who bears the cost when it isn't.
Alongside my own research, I serve the field as a program committee member and peer reviewer for IEEE and international venues.
- Focus
- Climate-adjusted mortgage credit risk, causal ML, explainable AI
- Data
- Fannie Mae, Freddie Mac, FEMA National Risk Index
- Memberships
- IEEE (Columbus Section), INFORMS
- Service
- PC Member, IEEE CIFEr 2026 and ATLC25, reviewer across international venues
- Background
- 5 years, credit operations and operational risk, Emirates Islamic Bank, Dubai
- ORCID
- 0009-0007-3378-180X
Research
Publications
Featured · Under review
Causal Machine Learning for Climate-Adjusted Mortgage Default Risk
Published
Springer
Exchange Rate Regimes, Financial Constraints, and Export Pricing: Evidence from Chinese Firms
IEEE Xplore
DevOps for Web-Native Bioactivity Graphs: A Scalable RDF Framework for ChEMBL Data Integration
Applied ML methods
Foundational techniques that carry directly into my climate-finance modeling: detecting regime shifts in non-stationary data, and building representation pipelines for high-dimensional signals.
IEEE Xplore
Regime Detection in Non-Stationary Time Series Using Hidden Markov Models and Support Vector Machines
IEEE Xplore
VoxTransmute: Transformation-First Waveform-to-Codebook Pipeline for Speaker Recognition
Under review
Springer
Empirical Analysis of Monetary Policy Optimization in Ghana under Uncertainty using a Soft Actor-Critic Reinforcement Learning Framework
Peer Review & Service
Serving the field's quality bar.
Manuscript peer review across international venues, with program committee appointments at international conferences.
Program Committee Member
IEEE CIFEr 2026
IEEE Symposium on Computational Intelligence for Financial Engineering & Economics, Tokyo, Japan. Reviewed computational-finance submissions.
Program Committee Member
ATLC25
Advances in Teaching & Learning Conference, Georgia Tech, Atlanta. Listed in the printed conference program.
| Venue | Role |
|---|---|
| IEEE CIFEr 2026, Computational Intelligence for Financial Engineering & Economics | Program Committee · Reviewer |
| IEEE ECCE 2026, Energy Conversion Congress & Exposition | Reviewer |
| ATLC25, Georgia Tech | Program Committee · Reviewer |
| Computology: Journal of Applied Computer Science & Intelligent Technologies | Reviewer |
| REST Publisher journals (JEMM · JITL · JDAAI) | Reviewer |
| IEEE CAI 2026, Conference on Artificial Intelligence | Reviewer |
Reviewer training: Nature Masterclasses, Focus on Peer Review, certificate, April 2026.
For Editors & Conference Chairs
Inviting review, program committee, and editorial roles.
I welcome invitations to review, to serve on program committees, and to join editorial boards in my areas. I respond on deadline and deliver careful, technically grounded reviews. I do not use AI to write reviews, so manuscripts get real scrutiny and no fabricated citations.
Review expertise
How to cite my work
M. Merchant, "Adapting Credit and Asset-Backed Financing to Climate Threats: AI-Driven Modeling and LLPA Policy Reform," in Proc. 4th Int. Conf. Big Data and Artificial Intelligence Applications (ICBDAIA'25), Lecture Notes in Networks and Systems, vol. 1724. Cham, Switzerland: Springer, 2026, doi: 10.1007/978-3-032-10895-1_7.
M. Merchant, "Exchange Rate Regimes, Financial Constraints, and Export Pricing: Evidence from Chinese Firms," in Proc. 4th Int. Conf. Big Data and Artificial Intelligence Applications (ICBDAIA'25), Lecture Notes in Networks and Systems, vol. 1724. Cham, Switzerland: Springer, 2026, doi: 10.1007/978-3-032-10895-1_22.
M. Merchant, "Blockchain-Based Finance for Advancing Astronomy and Astrophysics," in Recent Developments in Computational Finance and Business Analytics (CFBA-2025), Learning and Analytics in Intelligent Systems, vol. 53. Cham, Switzerland: Springer Nature, 2025, doi: 10.1007/978-3-031-99477-7_16.
M. Merchant, "DevOps for Web-Native Bioactivity Graphs: A Scalable RDF Framework for ChEMBL Data Integration," in Proc. Int. Conf. Innovations in Intelligent Computing and Cybersecurity (ICEENG/IICC 2026), Cairo, Egypt. IEEE, 2026. [Online]. Available: https://ieeexplore.ieee.org/document/11582075
M. Merchant, "Regime Detection in Non-Stationary Time Series Using Hidden Markov Models and Support Vector Machines," in Proc. 2026 Int. Conf. Artificial Intelligence and Materials (ICAIM). IEEE, 2026. [Online]. Available: https://ieeexplore.ieee.org/document/11601227
M. Merchant, "VoxTransmute: Transformation-First Waveform-to-Codebook Pipeline for Speaker Recognition," in Proc. 2026 Int. Conf. Artificial Intelligence and Materials (ICAIM). IEEE, 2026. [Online]. Available: https://ieeexplore.ieee.org/document/11601370
Under review
M. Merchant, "Causal Machine Learning for Climate-Adjusted Mortgage Default Risk," under review, International Review of Financial Analysis (Elsevier). Working paper: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6759885
Recognition & Memberships
Recognition, memberships, and talks.
Award
GEM Award, Emirates Islamic Bank
Operational Excellence, April 2016. Certificate signed by the bank's CEO.
Award
GEM Award, Emirates Islamic Bank
Operational Excellence, Q3 2018.
Membership
IEEE Member
Columbus Section, Region 2.
Membership
INFORMS Member
Institute for Operations Research and the Management Sciences.
Certificate
Nature Masterclasses, Focus on Peer Review
Springer Nature, April 2026. Signed by the Editor in Chief of Nature.
- IEEE ICAIM 2026, 2026 International Conference on Artificial Intelligence and Materials · two papers presented
- ICBDAIA'25, 4th International Conference on Big Data and AI Applications, Kenitra, Morocco · two papers presented (July 2025)
- CFBA-2025, 3rd International Conference on Computational Finance and Business Analytics, IIMT Bhubaneswar · April 2025
- ICSBP 2025, International Conference on Sustainable Business Practices, IMI Kolkata · January 2025
Media
For journalists and editors.
I'm available for expert comment, data points, and background on climate risk in housing and credit markets. Three findings from my current research, in plain English:
- Flood risk is already in the data, and mispriced. In 200,000 Fannie Mae loans, the highest flood-risk quartile carries nearly five times the predicted default risk of the lowest, yet both pay identical loan-level price adjustments, a 30 basis-point gap worth roughly $488M a year in misallocated premium on a $500B book.
- The risk is concentrated, not uniform. Effects vary sharply across the 2,827 counties studied. A small share of geographies carries a disproportionate share of the exposure.
- Pricing hasn't caught up. Current loan-level price adjustments don't reflect measured climate risk, leaving mispriced exposure with lenders, guarantors, and ultimately taxpayers.
"The question is no longer whether climate risk is real in mortgage markets. It's whether anyone is pricing it, and who pays when they don't."
Usable with attribution: Mujahid Merchant, independent researcher, AI-driven climate financial risk. Areas I can speak to: climate risk and housing markets, mortgage credit and default modeling, FEMA hazard data, explainable AI in finance, GSE pricing and LLPA policy.
Short bio · 50 words
Mujahid Merchant is an independent researcher in AI-driven climate financial risk, based in Columbus, Ohio. His research applies causal machine learning to 22M+ U.S. mortgage records matched to FEMA hazard data to measure how climate risk affects default. He is a Springer and IEEE published author and IEEE program committee member.
Bio · 150 words
Mujahid Merchant is an independent researcher working on AI-driven climate financial risk, how flood and disaster exposure translates into mortgage credit risk, and whether markets price it. His research applies causal machine learning (Double Machine Learning, Causal Forests) and explainable AI to more than 22 million U.S. mortgage records, Freddie Mac and Fannie Mae loan-level data and a 622,000-observation national panel across 2,827 counties, matched to FEMA disaster and flood-risk data. The core study is under review at the International Review of Financial Analysis (Elsevier). His work is published by Springer and IEEE, and he serves as a program committee member at IEEE CIFEr 2026 (Tokyo) and ATLC25 (Georgia Tech). Before research, he spent five years in credit operations and operational risk at Emirates Islamic Bank in Dubai, earning two GEM Awards for operational excellence. He is a member of IEEE and INFORMS.
Writing
Medium · Policy
The Disaster That Doesn't Get Declared: What FEMA Reform Means for Your Mortgage
How the FEMA reform bill quietly moves undeclared disaster losses onto escrow accounts and fixed-rate mortgages.
Medium · Analysis
43 Michigan Counties Just Got a Disaster Declaration. Here's What the Data Says Happens Next to Mortgages.
A federal declaration starts a two-year mortgage story that standard risk models miss.
Behind the Paper · Springer Nature Research Communities
Climate threats, mortgage pricing, and AI for resilient asset-backed finance
The story behind the Springer chapter: why loan-level price adjustments should carry a climate signal, and what disciplined modeling looks like when hazard inputs are uncertain.
Medium · Analysis
Florida's Insurance Market Recovered. Its Mortgage Market Never Priced the Risk at All.
One market looked at climate risk and repriced. The other still isn't looking.
Medium · Methods
Why Causal Inference Is the Missing Piece in Climate-Mortgage Risk Modeling
Prediction tells you who defaults. Causal inference tells you why, and what policy can change.
Medium · Data
FEMA Data Meets Fannie Mae: Building a Climate-Mortgage Research Pipeline
How federal hazard records and GSE loan-level data become one research-grade dataset.
Medium · Findings
The Average Is the Lie: Why Climate Risk Demands Heterogeneous Models
National averages hide the counties where climate-credit risk actually concentrates.
Contact
Get in touch.
For research collaboration, peer review and program committee invitations, or media inquiries.