On this page
GARP's public FRM exam page, checked on 19 October 2026, displays November 2025 pass rates of 47% for Part I and 50% for Part II. These figures describe those exam cohorts, not an individual's probability of passing a future sitting. The page also says results are pass or fail, with quartile comparisons available to candidates in their portal. The public page's displayed window is the scope of this article; it should not be relabelled as a 2026 result.
What the published figures actually say
On the FRM program and exams page, GARP labels the figures as November 2025 pass rates. Part I's 47% and Part II's 50% are separate shares for separate examinations in that administration. They are not the same candidate pool observed twice, and multiplying them would not produce a reliable probability of eventually completing both parts. GARP's public page does not provide the number of test takers behind these displayed percentages or a breakdown by study time, work background or question topic. The percentages therefore answer a narrow historical question: what share passed each part in the named window according to GARP's displayed result. They do not describe the margin by which any candidate passed or failed.
The page offers an important second piece of context: GARP scores each part on a pass/fail basis and makes a candidate's result accessible through its portal within eight weeks of the window closing. Candidates also receive quartile results that compare their performance with other candidates. GARP's FRM FAQ likewise says candidates receive an email when results are posted and directs them to the portal. A public pass-rate percentage and a private candidate performance report are different products. Someone awaiting a personal result should follow the portal process described by GARP rather than infer an outcome from the public cohort rate. No public percentage can reveal their individual result early.
Why the window label matters
An exam administration is a period, not a permanent property of the credential. The latest window displayed on the GARP page we opened is November 2025. GARP also states that Part I and Part II are offered in May, August and November, but the existence of later exam windows does not establish that later public pass-rate figures are on the page. Equally, an older displayed percentage does not prove that no later candidates have received private portal results. The honest description is that the opened public page displays November 2025. If GARP changes that page before publication or after this article appears, the window and figures should be checked again. Do not retitle an old figure as current-year news.
The two rates are close in the displayed window, but that visual similarity should not be overinterpreted. Part I and Part II test different content, and Part II is taken under different candidate circumstances because candidates ordinarily progress after passing Part I. The public percentages do not adjust for differences in preparation, experience, prior attempts or who chose to sit. Even a consistent series of past rates would not isolate the effect of any particular study method. It would be especially unsound to compare two preparation providers by pointing to GARP's cohort rates; the official numbers contain no provider-level outcome evidence. For a personal plan, use the rates as context about a historical administration, not as a forecast calibrated to your own situation.
A numerical illustration of the interpretation error
Suppose, purely as a teaching illustration, an imaginary exam cohort had 100 Part I candidates and 100 different Part II candidates. If 47 of the first group and 50 of the second group passed, the rates would match the percentages GARP displays for November 2025. Yet nothing in those two counts identifies which people, if any, appear in both groups, and the invented cohort sizes are not GARP data. You could not calculate how many people completed both exams from the two percentages. Nor could you conclude that a particular Part I candidate has a 47% chance of passing next time. Their preparation, the future administration and the scoring decision are not represented by that simple quotient.
Consider a second illustration: a candidate has finished a practice paper and wants to convert their raw score into a pass forecast by multiplying it by the historical pass rate. That operation has no valid interpretation. A pass rate counts people who met an examination standard in a past group; a practice score measures answers in a different set of questions. One is a group outcome and the other is an individual's performance indicator. GARP does not disclose on the public pass-rate panel a raw-score threshold that would let the candidate translate the percentage directly into a target mark. The practical use of a practice result is to identify weak topics and check the quality of reasoning, while the official pass-rate panel remains descriptive background.
Read quartiles without inventing a cut score
GARP says candidates receive quartile results comparing their performance with others. A quartile is a relative placement band, so it can help a candidate identify where performance was stronger or weaker across reported areas. It is not the same thing as a published minimum passing score. A candidate should resist converting a quartile band into an exact number of correct answers or a precise distance from the pass line when GARP has not supplied those values on the opened pages. A stronger band in one topic does not cancel out an unknown deficit elsewhere through simple arithmetic. The personal report is most useful for prioritising review, and its interpretation remains bounded by the information GARP actually provides.
Part I and Part II should be handled separately in that review. GARP's program page lists Part I's areas as foundations of risk management, quantitative analysis, financial markets and products, and valuation and risk models. It lists Part II areas including market risk, credit risk, operational risk and resilience, liquidity and treasury risk, risk management and investment management, and current issues in financial markets. These published area descriptions explain why an undifferentiated statement such as 'FRM was harder this year' is not supported by two pass-rate percentages. The rates alone neither identify which topic drove outcomes nor show how the composition of questions changed.
A decision checklist for a later sitting
When a new exam window closes, there are three separate questions to keep apart. First, has GARP notified individual candidates and placed results in their portal? The FAQ explains that route. Second, has GARP updated the public pass-rate panel, and which window does its label name? The page itself is the evidence for a public cohort figure. Third, what does a particular candidate's own report show? Only that report can support a personal diagnosis. A reader who answers the second question with a figure copied from a previous window, or answers the third with a public percentage, has crossed an evidence boundary. This checklist remains useful even if GARP later replaces the displayed November 2025 figures.
For someone planning a retake, the pass rate is especially tempting as a shortcut: a higher historical rate may look like an easier future exam window. The opened GARP pages do not supply the candidate-level data needed to establish that inference. The better decision is to use the personal quartile report as a starting point, revisit the concepts behind weak responses and confirm the next window's official instructions in the candidate account. For a first-time candidate without a report, a diagnostic set of self-marked work can reveal reasoning gaps, but it is still not a calibrated probability of passing. Keep the labels descriptive: cohort outcome, personal feedback and practice performance.
How a candidate can use the information
Start by recording the source, administration and part next to any pass-rate figure. If the figure came from a screenshot or an undated summary, reopen GARP's page before repeating it. When discussing a personal result, use the candidate portal and its own report rather than social posts about pass rates. When studying for a future sitting, turn any topic-level performance information you actually have into a plan: explain an error in words, reconstruct the relevant concept without looking, and then check another problem of the same type. This process does not require a prediction of the future pass rate. It links action to evidence about the candidate's own work.
The FRM program page is a place to orient your study to the credential, while Omni's methodology page explains its general study method. Neither changes what GARP's public pass-rate panel measures. The useful conclusion is narrow and concrete: GARP publicly displays 47% and 50% for November 2025 Part I and Part II on the page opened for this article; candidates get their own pass/fail and quartile information through GARP; and the historical group figures cannot substitute for a personal result or a measured forecast.
Sources
- FRM Program and Exams | GARP · garp.org · Retrieved
- FRM FAQs | GARP · garp.org · Retrieved
