GPT and Claude failed Bridgewater's finance tests because the right answers were never public

They defined six tasks drawn from an investor's daily routine. One example: deciding whether a financial article is relevant to an executive. Another: whether a central bank document signals the direction of future rate changes. For investors, these calls are trivial, but they can barely put their reasoning into words. The report gives a telling example. A headline about Trump's claim to Greenland gets flagged as irrelevant, while Trump's threat of new China tariffs is highly relevant. Both touch on geopolitics and finance.

Frontier models failed in the authors' tests. Variants of Gemini, Claude, and GPT hit only about 50 percent accuracy with a basic prompt. Expert-written instructions and a three-tier rating system ("relevant and interesting," "relevant but uninteresting," "irrelevant") pushed accuracy into the mid-70s. That still fell short of the 80 percent threshold the authors set for trustworthy deployment.

Newer models barely improve per dollar, the report says. GPT 5.4 costs 43 percent more than 5.2 but is only marginally more accurate.

The real value lives inside investors' heads

The solution was fine-tuning, retraining an open-weight model on proprietary examples. The key ingredient was the Bridgewater investors' judgment: At first, cheap outside contractors labeled the documents, but many of those labels were wrong. To avoid having expensive professionals review everything, the researchers used a workaround. A first model learned from the flawed labels and re-evaluated the same documents. Wherever the model and the original label disagreed, there was likely an error. Only those disputed cases went to investors for correction.

Training ran on the Tinker platform from Thinking Machines Lab, built on top of the open model Qwen3-235B. In the team's own evaluation, the fine-tuned model hit 84.7 percent accuracy versus 78.2 percent for the best frontier model tested. It also cost nearly 14 times less to run. This isn't a truly independent comparison, of course. Both companies have a clear interest in selling their product.

Still, the finding beyond the numbers is worth noting. It shows once again that the big labs like OpenAI haven't absorbed all the data out there. Huge pools of proprietary corporate data and untrained human expertise still exist, and they hold real room for improvement. That's especially true where companies deliberately keep their most valuable data private. Anyone who hands that data to a frontier lab risks competing against a product built on top of it.

Fine-tuning open models through tools like Tinker gives companies an alternative. They keep the weights, the data, and, depending on the setup, the GPUs themselves.