a study<\/a> found that leading models could pass the first two levels of the CFA, but Level III\u2019s essay questions remained a sticking point because of their open-ended nature.<\/p>\n Not anymore. Two developments helped the latest generation of LLMs clear the essays. First, AI labs have released \u201creasoning” models like o4-mini and Gemini 2.5 Pro that can break down questions into smaller steps, improving their critical thinking. Second, researchers used \u201cchain-of-thought prompting,” where they asked models to explain their reasoning while approaching problems from multiple angles.<\/p>\n
The LLMs\u2019 performance is particularly striking because they passed without receiving any specific training, according to GoodFin CEO Anna Joo Fee. \u201cWhat\u2019s so exciting is thinking how much better you can make them with human experts, with training, with context,” she says.<\/p>\n
The tools still have room to improve, says Fee. LLMs aren\u2019t yet good at grasping user intent, especially when expressed through nonverbal cues like body language. Accuracy scores could go up as well. \u201cIt will dramatically improve over time,” she says.<\/p>\n
What CFAs think.<\/strong> Kate Feeney, a wealth advisor at Summit Place Financial, studied three hours each workday for a year before sitting for the CFA Level III, which she passed in 2016. She says she isn\u2019t surprised that AI can now pull off the same feat. \u201cI personally had the hardest time with time management while taking the Level III,” she says. \u201cAnd AI is incredibly fast.”<\/p>\n Derek Williams, an advisor at Veratis Advisors, earned his CFA in 2023 after five solid months and hundreds of hours of studying. \u201cYou can\u2019t really wing it,” he says.<\/p>\n
Just because AI is getting smarter doesn\u2019t mean analysts will be out of work anytime soon, CFA charterholders say.<\/p>\n
\u201cMy experience as a financial planner is that people don\u2019t seek me out for technical knowledge,” says Williams. \u201cOur industry is already moving towards not really relying on a person to do base-level analysis.”<\/p>\n
Summit\u2019s Feeney adds: \u201cI think there\u2019s definitely elements of advice that can be automated.” However, \u201cYou have to be able to effectively communicate with your clients. I don\u2019t think empathy and that human component can be replicated right now.”<\/p>\n
Mike de Massa, CIO of Forza Wealth Management, isn\u2019t worried about AI coming for analysts\u2019 jobs. But he does worry that AI tools could automate the low-level tasks that typically teach young CFAs important skills.<\/p>\n
\u201cIf all the elementary schools gave kids calculators to learn their times tables, then when they get to algebra class, they\u2019re not going to know what the heck they\u2019re doing,” says de Massa. \u201cAI is like a calculator for CFAs.”<\/p>\n
The three-part test to become a chartered financial analyst is widely viewed as one of the hardest exams in finance. But new research shows that many of today\u2019s AI models can pass it in a matter of minutes\u2014and they continue to improve at a rapid pace. Just because AI is getting smarter doesn\u2019t mean analysts […]<\/p>\n","protected":false},"author":2,"featured_media":406188,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[7],"tags":[6467,263,262,40961,40960,260,259,40962,258,265,202,261,264,40963],"class_list":["post-406185","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-metaverse","tag-ai-models","tag-axie-infinity","tag-axs","tag-cfa-level-iii-exam","tag-chartered-financial-analyst","tag-decentraland","tag-facebook","tag-financial-analysts","tag-game","tag-mark-zuckerberg","tag-nft","tag-sandbox","tag-vr","tag-wealth-advisory-firm"],"_links":{"self":[{"href":"https:\/\/dripp.zone\/news\/wp-json\/wp\/v2\/posts\/406185","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/dripp.zone\/news\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/dripp.zone\/news\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/dripp.zone\/news\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/dripp.zone\/news\/wp-json\/wp\/v2\/comments?post=406185"}],"version-history":[{"count":1,"href":"https:\/\/dripp.zone\/news\/wp-json\/wp\/v2\/posts\/406185\/revisions"}],"predecessor-version":[{"id":406189,"href":"https:\/\/dripp.zone\/news\/wp-json\/wp\/v2\/posts\/406185\/revisions\/406189"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/dripp.zone\/news\/wp-json\/wp\/v2\/media\/406188"}],"wp:attachment":[{"href":"https:\/\/dripp.zone\/news\/wp-json\/wp\/v2\/media?parent=406185"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dripp.zone\/news\/wp-json\/wp\/v2\/categories?post=406185"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dripp.zone\/news\/wp-json\/wp\/v2\/tags?post=406185"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}