The concept underneath the concept
Most diagnostics tell you which topic you scored worst in. That is rarely where the problem is. Risk Taxonomy is a prerequisite for 3 other concepts in the FRM Part 2 syllabus, spread across different topic areas — so if it is shaky, you lose marks in several places and it looks like several unrelated weaknesses.
These are the real dependencies this diagnostic reasons over. They are read straight from the syllabus map the tutor uses — not a marketing simplification of it:
| If this is weak… | …these suffer, in other topic areas |
|---|---|
| Risk Taxonomy | Digital Assets & Crypto Risk, Geopolitical Risk, Liquidity Risk Types |
| Duration & Convexity | Arbitrage Pricing and the Shape of the Term Structure, Fixed Income Market Risk, Mortgage-Backed Securities |
| Bond Pricing & Yields | Arbitrage Pricing and the Shape of the Term Structure |
| Credit Risk Fundamentals | Private Credit, Securitisation Mechanics |
| Credit Risk Transfer Mechanisms | Credit Derivatives, Securitisation Mechanics |
| Option Valuation | Arbitrage Pricing and the Shape of the Term Structure, Structural Default Models |
| Risk Management Framework | Model Risk & Inventory, Operational Risk Framework |
| Time Series & Volatility | Term Structure Models: Ho-Lee, Vasicek, CIR and Gauss+, Volatility Modelling |
| VaR Estimation Methods | Risk Budgeting & Monitoring, Stress Testing |
| CAPM & Performance Measurement | Factor Theory & Models |
| Central Clearing | Counterparty Risk & CVA |
| Commodity & FX Markets | Cross-Currency Funding |
| Correlations & Copulas | Portfolio Credit Risk |
| Country & Sovereign Risk | Risks of Rising Government Debt |
| Credit Ratings & Corporate Bonds | Credit Risk Fundamentals |
| Cyber & Third-Party Risk | Digital Resilience |
| Fund Management | Hedge Fund Risk |
| Linear Regression | Factor Theory & Models |
| Machine Learning Foundations | AI & Machine Learning Risk |
| Measuring Credit Risk | Portfolio Credit Risk |
| Model Validation | AI & Machine Learning Risk |
| Mortgages and Mortgage-Backed Securities | Mortgage-Backed Securities |
| Portfolio Credit Risk | CDOs & Correlation Products |
| Regulatory Capital & Basel | Fundamental Review of the Trading Book |
| Statistical Inference | Operational Risk Measurement |
| Stress Testing | Stress Testing |
| Value at Risk Foundations | VaR Estimation Methods |
That is what the 25 questions are for. Spread across every topic area, they are enough to separate "you do not know this topic" from "you do not know the one thing this topic is built on".
What you get at the end
A finding, not a percentage. Something like "your Bond Pricing is costing you marks in two later topics, not just its own" — with the topics named, so you can check it against your own experience of the exam rather than taking our word for it.
You also get the score and a per-topic breakdown, because you will want them. They are just not the point. A score tells you where you are; it does not tell you what to do on Monday morning.
When the 25 questions are not enough to support a claim, it says so rather than inventing one. A single wrong answer on a concept is not evidence that you do not know it, and being told something false about your own preparation is worse than being told nothing.
Questions
Is the diagnostic really free?
Yes — 25 questions, no account and no card. You can take it without signing up, and you get the full result including which concept is underneath your wrong answers.
How long does it take?
About ten minutes. There are 25 questions, mixed across every FRM Part 2 topic area and across difficulty, weighted harder than a random draw because easy questions do not tell you much.
Is this a mock exam?
No. A mock measures where you are against the pass mark; this looks for the cause of the marks you are dropping. It is deliberately shorter and deliberately harder than a representative sample.
What happens to my answers?
They are stored without a name, an email or an IP address, and they help us learn how hard each question actually is. They are kept separate from our own learners' answers, because anonymous visitors are a different population and mixing the two would bias the difficulty estimates.