Arun Gupta

Research Papers:

• Agentic AI for Investment Data. Work-in-progress

  • I am developing a multi-agent system to automate return and risk analysis using a dataset containing millions of position-level holdings (equities and fixed income) of the largest global banks. In particular, I leverage python libraries (CrewAI and LangChain) to orchestrate specialized agents for cash flow volatility, tail risk probability, and macroeconomic stress assessments. These agents are then used to generate interpretable, professional-grade reports tailored for investment, structuring, and regulatory insight. This project bridges explainable AI, agentic automation, and scalable investment risk reporting for institutional platforms.

• Forecasting Macroeconomic Risk in Private Markets. Work-in-progress

  • This study compares the forecasting ability of ten machine learning methods in predicting future default probabilities for loans to private firms. I then use the best model (gradient boosting) to calculate the sensitivity of loan portfolio losses to macroeconomic shocks and calculate the optimal risk premium to cover these risks.

• Machine Learning for Financial Tail Risk Forecasting. Working Paper. SSRN 

  • Many alternative investment funds rely on heuristic stress-testing or traditional statistical methods for forecasting tail risk. Using electricity spot prices (one of the most volatile and fat-tailed assets in the world), I show that a well-calibrated machine learning model (LightGBM) provides superior out-of-sample tail risk and volatility forecasts (67% to 85% more accurate) in comparison to a suite of traditional statistical methods.

• The Performance of Private Equity: Evidence from Confidential Filings. Working Paper. SSRN 

  • Most insightful for helping institutional investors better optimize their capital allocation decisions, I show new evidence that (1) larger PE funds earn significantly higher returns, (2) more diversified PE funds earn lower returns, but present higher reward-to-risk ratios (due to industry diversification), and (3) leveraged buyouts yield no return premium for the risk incurred.

• A Natural, yet Imperfect Hedge for Interest Rate Risk: Estimating the Value of Deposit Franchise. Working Paper. SSRN

  • I show direct empirical evidence that markets price in deposit franchise (worth on the order of $1 trillion USD) when valuing bank equities, and quantify how imperfect this deposit franchise is as an interest rate hedging tool.

• The Collateral Channel and Bank Credit (with Vladimir Yankov and Horacio Sapriza). Under revision at the Journal of Financial Economics SSRN, Federal Reserve

• Utilizing confidential loan-level data from global lenders, our paper provides causal evidence that fluctuations in collateral values have a strong effect on the future credit capacity to private firms.

• The Internal Capital Markets of Global Dealer Banks. Journal of Financial Crises, 2022. Link

• Provides evidence that loans between sibling entities in global dealer banks experienced substantial runs during the 2008 financial crisis. In short, global dealer banks experienced internal runs.

• Un-used Bank Capital Buffers and Credit Supply Shocks at SMEs during the Pandemic (with J. Berrospide and M. Seay). International Journal of Central Banking, 2024. SSRN Federal Reserve

• Shows evidence that one of the Basel III capital regulations (regulatory capital buffers) was more costly on banks than intended by regulators. During the pandemic, the excessive costliness of the regulation led banks to inefficiently curtail lending to small and medium-size private businesses more than otherwise warranted by the covid recession.

• Do Costly Internal Equity Injections Reveal Bank Expectations about Post-Crisis Real Outcomes? (with Horacio Sapriza). SSRN, Fed 

• Show direct evidence that reveals the sharp and accurate forecasts lenders have regarding future economic outcomes (including future growth in local business revenues, employment, and establishments). 

• Measuring Global Bank Complexity (with Linda Goldberg and Nicola Cetorelli), FRBNY Liberty Street Economics, 2014. Link

  • Addressing the macro-prudential concerns of the growing “complexity” of global financial institutions, we look present comprehensive data analyses on various metrics for measuring the “business line”, “geographical”, and "organizational” complexity of the largest financial intermediaries.

• Ring-Fencing and “Financial Protectionism” in International Banking (with Linda Goldberg), FRBNY Liberty Street Economics, 2013. Link

  • We discuss recent policy and regulatory efforts to induce banks to shift away from global activity, including ring-fencing of domestic banking operations, other forms of financial protectionism, and enhanced oversight and prudential measures.