A focused FRM flashcard deck

FRM questions test more than whether you can recall a formula. You need to identify what a quantity means, choose assumptions that fit the problem, calculate without unit or sign errors, and explain what the result implies for risk decisions. That applies across both parts of the multiple-choice, computer-based examination.

Use this deck as a short, active study session. It focuses on core risk-measurement ideas that connect quantitative methods with practical interpretation. It is not a substitute for your current syllabus or notes: check those for the full set of topics and definitions you are responsible for.

How to work through the cards

Take each prompt in order. Read the question, answer from memory, then reveal the answer. Grade your recall honestly: did you retrieve it promptly and accurately, or did you need a hint or correction? A familiar-looking formula is not a successful recall if you cannot explain its assumptions or interpret the result.

For calculation-related cards, say the units and sign convention aloud. For conceptual cards, add one sentence on how the fact might change a risk decision. Mark cards for review when your answer is incomplete, even if you recognised the topic. MySummaries presents a deck from a candidate’s revision board in this format:

A useful FRM deck on market, credit and model risk looks like this:

Cards — Risk measurement: VaR, ES and credit risk18 due

What does a one-day 99% VaR of $5 million state, under the loss convention?

Over the one-day horizon, the loss is not expected to exceed $5 million with 99% confidence, under the model and assumptions used. It does not say the maximum possible loss is $5 million.

All 18 cards
What does a one-day 99% VaR of $5 million state, under the loss convention?Over the one-day horizon, the loss is not expected to exceed $5 million with 99% confidence, under the model and assumptions used. It does not say the maximum possible loss is $5 million.
For zero-mean, normally distributed portfolio returns, what multiplier gives one-sided 99% VaR in standard deviations?Approximately 2.326 times the portfolio standard deviation, for the stated horizon and under the normality and zero-mean assumptions.
What is the parametric VaR formula for a zero-mean normal loss over a stated horizon?VaR at confidence level c is z_c times the portfolio standard deviation over that horizon; z_c is the one-sided standard-normal quantile.
When is square-root-of-time scaling of VaR from one day to T days justified?As an approximation when returns are independent and identically distributed with finite variance; it can fail with autocorrelation, volatility clustering or changing distributions.
In historical simulation, how is VaR estimated?Apply historical changes in risk factors to the current portfolio, order the resulting simulated losses, and take the loss quantile for the chosen confidence level.
What does expected shortfall at confidence level c measure?The average loss in the worst tail beyond the VaR threshold at confidence level c, using a consistent tail-averaging convention.
What is the expected-loss formula for a credit exposure?Expected loss = PD × LGD × EAD, with probability of default, loss given default and exposure at default expressed consistently.
If PD is 2%, LGD is 40% and EAD is $10 million, what is expected loss?0.02 × 0.40 × $10 million = $80,000 over the period and on the definitions used for those inputs.
How is loss given default related to recovery rate?LGD = 1 − recovery rate, when both are expressed as fractions of the relevant exposure.
What is the first-order duration approximation for a bond price change?ΔP/P ≈ −D_mod × Δy, where modified duration is D_mod and the yield change Δy is in decimal units.
What does the minus sign in the duration approximation indicate?For a conventional fixed-rate bond, price generally moves inversely to yield: a yield rise implies an approximate price fall, all else equal.
What does a bond’s convexity adjustment add to the duration approximation?A second-order term, approximately ½ × convexity × (Δy)², to improve the price-change estimate for larger yield moves, subject to the model’s conventions.
What is a key limitation of VaR as a tail-risk measure?VaR identifies a loss quantile but does not describe the size of losses beyond that threshold.
Why can normal-distribution VaR understate market risk?If returns have heavier tails or more extreme outcomes than the normal model allows, the estimated tail quantile may be too small.
Why can diversification fail to reduce VaR as expected?Dependence and correlations can change in stressed markets; moreover, VaR is not subadditive for all loss distributions.
What does a stress test add to a VaR estimate?It examines portfolio losses under specified severe scenarios, including events or combinations of moves that a routine statistical estimate may not represent well.
What is the difference between market risk and liquidity risk?Market risk is the possibility of loss from adverse market-price movements; liquidity risk includes being unable to trade promptly or without a material price impact.
Why should a risk estimate be reported with its horizon, confidence level and assumptions?Those details define what the number represents; without them, comparisons and decisions based on the estimate can be misleading.

That is a MySummaries deck, filled with FRM material. Yours is written from your own notes. Start free

Review the facts that carry the most weight

Once you have worked through all 18 cards, sort your misses into three categories: formula setup, interpretation, and assumptions or limitations. This tells you what kind of practice to do next. If you recalled the VaR multiplier but used a two-sided quantile, revisit the definition and convention. If you calculated expected loss correctly but could not explain its period or inputs, practise stating what each input represents.

A short “must not miss” list helps you check the structure of the deck before repeating it. MySummaries can keep that checklist beside the cards, built from the same material:

A compact core checklist for this deck would read:

Key pointsRisk measurement: VaR, ES and credit risk
VaR: state horizon, confidence level, loss convention and model assumptions; VaR is a quantile, not a maximum-loss estimate.
Normal VaR: for zero-mean returns, use the one-sided quantile multiplier; at 99% it is approximately 2.326 standard deviations.
Credit expected loss: PD × LGD × EAD; keep input definitions, units and time period consistent.
Model discipline: test distributional and dependence assumptions; interpret the estimate, note tail limits, and use stress analysis where appropriate.

Use the checklist to diagnose, not to replace retrieval. Cover one item and explain it in your own words, including a relevant limitation. Then return to the corresponding card and check whether your answer was precise enough. For numerical items, work through the arithmetic with units attached. In the expected-loss example, 2% is 0.02 and 40% is 0.40; multiplying each by a $10 million exposure gives $80,000. That answer still needs its period and input definitions to be meaningful.

Turn missed marks into the next review

A repeated miss usually has a cause. Perhaps you know the equation but have not learnt when its assumptions fail; perhaps you can calculate but omit the decision implication. After a practice question, write down the exact step where your answer stopped earning credit. Convert that step into a single, answerable prompt—not a broad instruction such as “revise VaR”.

For instance, if you repeatedly describe VaR as the worst possible loss, the repair card should ask what VaR does and does not claim. MySummaries can put a targeted card into a remediation tray after a mark is lost twice:

A useful repair prompt for that error is:

Remediation tray

You lost this mark twice: what does a one-day 99% VaR quantify, and why is it not the maximum possible loss?

Add cardDismiss

Finish the session by answering the repair prompt without looking at the deck. Then schedule another retrieval rather than rereading the answer immediately. When the cards feel easy, test the same ideas in a mixed question: identify the requested quantity, state assumptions, set up the formula, calculate with units, and interpret the result. That sequence mirrors the habits this deck is meant to practise; it does not replace timed practice with the question styles in your current FRM materials.

How MySummaries helps

MySummaries turns your own PDFs, slides and photographed notes into revision boards, then creates flashcards and spaced-repetition reviews from those materials. For FRM preparation, you can keep a deck like this linked to your own definitions, formula conventions and practice-question corrections. Open MySummaries.