A recent study explored the application of large language models (LLMs) in counterfactual analysis, specifically using the GPT-3.5 model. Counterfactual analysis aims to predict potential outcomes under hypothetical scenarios, offering valuable insights for decision-making. The researchers applied this method to the context of online lending, evaluating the counterfactual return on investment (ROI) under different interest rate schemes. The results showed that prompt engineering significantly enhanced GPT's predictions, with the R-squared increasing from 1.97% to 2.84%, closely approaching the 3.48% achieved by gradient-boosted regression. Furthermore, the researchers used GPT to generate counterfactual ROIs under a set of alternative interest rates, and the results demonstrated GPT's logical coherence and causal reasoning abilities. These findings underscore the potential of LLMs as effective tools for counterfactual analysis in online lending, suggesting broader applications for LLMs in various predictive and decision-making contexts. Blogger's Review: This paper reveals the great potential of large language models in counterfactual analysis, especially in the financial sector, such as online lending. By optimizing the prompt, the GPT model can provide predictions that are close to those of traditional machine learning algorithms, providing new tools and ideas for complex decision-making.