Global Merger-Arbitrage Forecasting with Language Models
Quick Answer
This paper shows that A novel language-model forecasting system for merger arbitrage predicts M&A outcomes with a 0.151 Brier score, outperforming market-implied probabilities by 24%.
Quick Take
The system utilizes expert-guided and hindsight-based reasoning, achieving superior results on over 400 deals across 42 countries.
Key Points
- The system predicts three outcomes: closing at announced terms, a higher bid, or deal termination.
- Achieved a Brier score of 0.151, outperforming XGBoost by 19% and frontier language models by 25-42%.
- Utilizes long-context reasoning over hundreds of pages of technical documents.
- Combines expert-guided context engineering with hindsight-guided reasoning traces.
- Demonstrates success in specialized financial workflows with -based forecasting.
Paper Resources
📖 Reader Mode
~2 min readAbstract:We present a language-model forecasting system for merger arbitrage, a specialized high-stakes financial setting in which the task is to predict the outcome of announced M\&A deals. Unlike prior work on judgmental forecasting with LLMs, which has focused on broad mixed-topic benchmarks and short context such as news snippets, we study a setting that requires long-context reasoning over hundreds of pages of technical documents. Our system combines expert-guided context engineering with finetuning on hindsight-guided reasoning traces derived from historical deals. Given an announced deal, it outputs a probability distribution over three mutually exclusive outcomes: closing at announced terms, a higher bid, or deal termination. On an out-of-sample set of more than 400 large deals spanning 42 countries, our finetuned system achieves the best performance of any method we evaluate, reducing class-balanced Brier score to 0.151. This is 24\% below calibrated market-implied probabilities, 19\% below XGBoost, and 25-42\% below frontier language models. These results, together with ablation studies, show that LLM-based forecasting can succeed in specialized, long-context financial workflows, with hindsight-based supervision and expert-designed context playing a critical role.
| Comments: | Accepted to ICML 2026 |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2607.09921 [cs.CL] |
| (or arXiv:2607.09921v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2607.09921 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Hinal Jajal [view email]
[v1]
Fri, 10 Jul 2026 19:16:03 UTC (112 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CL
See more →TriAgent: Divergence-Aware Committees for Cost-Efficient Financial Sentiment Analysis
TriAgent introduces a cost-efficient multi-agent system for financial sentiment analysis, combining VADER, FinBERT, and Qwen2.5. It achieves an F1 score of ~0.87 with significant savings of $9.3M/year at a 10M-user scale compared to GPT-4o-mini, while also detecting hallucinations with an AUC of 0.90.