{"id":2260,"date":"2025-10-20T11:00:30","date_gmt":"2025-10-20T11:00:30","guid":{"rendered":"http:\/\/thetoptenwebhosts.com\/?p=2260"},"modified":"2025-10-20T13:58:09","modified_gmt":"2025-10-20T13:58:09","slug":"ai-systems-pass-toughest-finance-exam-in-stern-study","status":"publish","type":"post","link":"http:\/\/thetoptenwebhosts.com\/index.php\/2025\/10\/20\/ai-systems-pass-toughest-finance-exam-in-stern-study\/","title":{"rendered":"AI systems pass \u2018toughest finance exam\u2019 in Stern study"},"content":{"rendered":"

\"\"<\/a><\/p>\n

Stern researchers found 23 artificial intelligence models that successfully\u00a0passed the highest level of the Chartered Financial Analyst exam, which is globally used to evaluate candidates\u2019 readiness. They said that last month\u2019s findings could make financial advising more accessible to companies with lower budgets.<\/span><\/p>\n

The researchers collaborated with the AI-driven wealth platform GoodFin to examine the \u201chigh-stakes analytical reasoning\u201d abilities of large language models such as ChatGPT and Gemini. The tested models successfully passed a mock version of the exam\u2019s third level, which contains multiple choice and essay questions that require synthesizing complex financial concepts.\u00a0<\/span><\/p>\n

Stern professor Srikanth Jagabathula told WSN that he began the research unsure of whether AI could play a role in financial advising because it is such a high-stakes industry.\u00a0<\/span><\/p>\n

\u201cYou need to be careful about what you deploy in practice,\u201d Jagabathula said. \u201cThese studies clearly demarcate areas where machines might be good and humans might be good, and points to a future where we could have both work together to offer the best advice.\u201d<\/span><\/p>\n

Previous studies have found that LLMs pass the first two levels of the CFA exam, which are composed of simple multiple choice questions. In the new study, researchers saw that the performance of \u201cfrontier\u201d LLMs were able to surpass their predicted passing threshold of 63% in the overall exam \u2014\u00a0with ChatGPT\u2019s o4-mini model passing at 79.1% and Gemini passing at 75.9%.<\/span><\/p>\n

The new findings also identified that reasoning versions of the LLMs \u2014 which are designed to spend more time on analyzing prompts \u2014 outperform non-reasoning models by an average of 19% when handling essay questions. LLMs also respond best when they receive precise description of the steps to take in an analysis, which researchers said is a first step to relying on AI in higher-stakes situations.<\/span><\/p>\n

\u201cBecause the space that we\u2019re in is heavily regulated and we\u2019re dealing with people\u2019s money, and usually professionals in this space will have credentials like the CFA Level III, it\u2019s really important that we have this understanding of how performant these models are,\u201d GoodFin founder and CEO Anna Joo Fee said in an interview with WSN.<\/span><\/p>\n

The study also found that the number of details LLMs require for accurate analysis will demand more computing power, which translates to a heightened cost for users \u2014 a 7.8% increase in multiple choice accuracy would see a jump in three to 11 times in cost.\u00a0<\/span><\/p>\n

To balance cost and efficiency, the researchers suggest that financial advising firms could assign smaller, faster models for simpler prompts, and only utilize more advanced models for equally complex prompts. However, Fee said that companies will still need to ease into AI usage.<\/span><\/p>\n

\u201cSome might argue that financial advising is more relationship-based than finance-based,\u201d Fee said. \u201cIt\u2019s no longer going to be enough to do the more basic things, but really understanding your client, being able to empathize and building those relationships \u2014\u00a0that\u2019s going to become paramount.\u201d<\/span><\/p>\n

Contact Mia Shou at mshou@nyunews.com.<\/em><\/p>\n

This story AI systems pass \u2018toughest finance exam\u2019 in Stern study<\/a> appeared first on Washington Square News<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"

Stern researchers found 23 artificial intelligence models that successfully\u00a0passed the highest level of the Chartered…<\/p>\n","protected":false},"author":1,"featured_media":2262,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[1],"tags":[],"_links":{"self":[{"href":"http:\/\/thetoptenwebhosts.com\/index.php\/wp-json\/wp\/v2\/posts\/2260"}],"collection":[{"href":"http:\/\/thetoptenwebhosts.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/thetoptenwebhosts.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/thetoptenwebhosts.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/thetoptenwebhosts.com\/index.php\/wp-json\/wp\/v2\/comments?post=2260"}],"version-history":[{"count":2,"href":"http:\/\/thetoptenwebhosts.com\/index.php\/wp-json\/wp\/v2\/posts\/2260\/revisions"}],"predecessor-version":[{"id":2263,"href":"http:\/\/thetoptenwebhosts.com\/index.php\/wp-json\/wp\/v2\/posts\/2260\/revisions\/2263"}],"wp:featuredmedia":[{"embeddable":true,"href":"http:\/\/thetoptenwebhosts.com\/index.php\/wp-json\/wp\/v2\/media\/2262"}],"wp:attachment":[{"href":"http:\/\/thetoptenwebhosts.com\/index.php\/wp-json\/wp\/v2\/media?parent=2260"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/thetoptenwebhosts.com\/index.php\/wp-json\/wp\/v2\/categories?post=2260"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/thetoptenwebhosts.com\/index.php\/wp-json\/wp\/v2\/tags?post=2260"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}