Start with the exam you are actually preparing for
The FRM consists of two parts, Part I and Part II. Both are multiple-choice examinations sat in computer-based sessions. Questions test quantitative techniques and their interpretation, as well as the application of risk frameworks to market, credit, operational, liquidity and investment risk scenarios. That means preparation has to train more than recall: you need to select a method, set it up correctly, calculate efficiently and explain what the result means for a risk decision.
Time pressure makes small errors costly. A correct formula used with the wrong units, sign, horizon or assumptions can still produce a poor answer. Practise deciding what the question asks before calculating, and learn to identify when a model’s assumptions do not fit the situation.
Use the official GARP website for current curriculum details, registration information, exam dates and any exam-specific instructions. The subject emphasis and administrative details can change; do not base your plan on an old syllabus or an unofficial summary.
A useful first step is to collect the current official materials for the part you are sitting, then turn each topic into a section on a working revision board. A board on this topic ends up looking like this:
Build a board from the curriculum
Keep Part I and Part II distinct in your notes, even where a concept recurs. Use the official learning objectives as the checklist for what belongs on each board. Do not infer that a topic is unimportant because it appears briefly in a study guide, or assume a fixed weighting unless the current official materials state one.
For each learning objective, create a compact note that answers three questions: What is the concept or model? What assumptions and inputs does it need? What would its result change in a risk decision? Add a worked example where calculation is involved. For a framework, include the mechanism, use case and limitation rather than just its name.
A board should help you retrieve and apply information, not become a second textbook. Keep formulas beside definitions of their variables and units. Put common traps beside the relevant method: for example, confusing a confidence level with a probability of loss, or applying a time-scaling rule without checking its assumptions. Mark sections for review when you miss questions or cannot explain a result without looking at your notes.
Here is a compact cross-part board structure. In your own revision, split it into the relevant part and official learning objectives rather than treating it as a substitute syllabus. A board on this topic ends up looking like this:
- Define the quantity: a loss threshold over a stated horizon and confidence level, under a specified model.
- VaR does not state how large losses are once the threshold is exceeded.
- State portfolio, horizon, confidence level and return or loss convention.
Identify the risk factor, position, horizon and liquidity conditions. A mark-to-market loss estimate may not capture the cost or time needed to exit a position in stressed markets.
The square-root-of-time rule gives VaR_T ≈ VaR_1 × √T only under restrictive assumptions, including stable volatility and appropriately independent returns. Check whether the model and data support those assumptions before applying it.
| Check | Question to ask |
|---|---|
| Units and scale | Do inputs and output use compatible units and horizons? |
| Distribution | Does the assumed distribution reflect the relevant tails? |
| Interpretation | What decision or control follows from the estimate? |
| Limitations | What is omitted, and when could that matter? |
- Separate probability of default, exposure at default and loss given default.
- Expected loss is commonly expressed as PD × EAD × LGD, with compatible horizons and definitions.
- Do not treat expected loss as a measure of unexpected tail loss.
Make a short core checklist
Once you have notes, reduce each board to a small “must not miss” checklist. This is not a replacement for the curriculum: it is a prompt to recover the essential setup under time pressure. Include a handful of definitions, relationships and checks that you repeatedly need. Avoid memorising a formula without the conditions that make it useful.
For quantitative work, practise writing the given information and the quantity asked for before choosing a formula. Define symbols, check units and state assumptions. For conceptual questions, identify the risk and the decision at stake, then connect the relevant framework to its limits. A compact core checklist might look like this:
Practise questions as a decision process
Question practice should diagnose how you think, not just whether you recognised a term. After each question, classify any error: knowledge gap, wrong model choice, misread quantity, arithmetic slip, unit or scaling error, or failure to interpret. Record the reason beside the relevant board section. If you guessed correctly, still verify the reasoning; a lucky selection is not evidence that you can repeat the method.
For a calculation, use a consistent sequence: Given and asked; assumptions and setup; method; calculation; interpretation and sanity check. Under time pressure, this prevents premature calculation and makes it easier to spot an implausible answer. In a conceptual or case-style question, identify the main risk mechanism, select an appropriate framework, connect it to measurement and mitigation, then state a limitation or trade-off.
Use timed sets to practise triage. If a question is taking too long, eliminate options using units, direction of effect or an invalid assumption, make a deliberate choice and move on if needed. Return to difficult items if time remains. After the set, spend at least as much attention on why an option was tempting as on why the correct answer works.
A marked practice response can show whether a sound calculation is supported by a sound interpretation. FRM is multiple choice; this example is a short written practice task for learning to explain a method, not a claim about the exam’s response format. A marked response on this topic ends up looking like this:
A portfolio has a one-day 99% VaR of $2.0 million. Estimate its 10-day 99% VaR using square-root-of-time scaling, and state what the estimate means and one important limitation.
3/4Using square-root scaling: $2.0m × √10 = about $6.32m. This is an estimated 10-day 99% VaR. The result depends on the assumptions behind the scaling rule.
The setup, arithmetic and interpretation of the quantile are sound. The answer recognises that scaling has assumptions but does not name them or explain what VaR leaves out.
Missed
−State that scaling assumes conditions such as stable volatility and independent returns; serial dependence or changing volatility can make the estimate unreliable.
−Clarify that VaR does not measure the size of losses beyond the 99th-percentile threshold.
Model answerUsing the square-root-of-time approximation, 10-day VaR ≈ $2.0m × √10 = $6.32m. This is an estimated loss threshold at the 99% confidence level over ten days, under the model’s assumptions. The approximation requires conditions such as stable volatility and independent returns; it may be unreliable when these fail. VaR does not quantify the severity of losses beyond the threshold, so a tail measure such as expected shortfall may add information.
That is a MySummaries paper, filled with FRM material. Yours is written from your own notes. Start free
Review weak areas, then listen for the logic
Do not respond to a weak score by rereading everything. Find the board section behind the error, repair the specific gap and test it again with a new question. If you repeatedly miss a topic, write a brief explanation from memory, compare it with your source notes and update the board with the missing assumption or distinction. Keep a separate record of recurring execution errors, such as forgetting to convert a horizon or mixing percentage and decimal inputs.
Audio review can be useful for revisiting a method away from the screen, but it should reinforce active practice rather than replace it. Listen for the sequence of reasoning: what is being estimated, what the inputs mean, which assumptions matter and how the output affects a decision. Pause after a worked step and predict the next one. A lecture drawn from a board should keep those details together. A lecture on this topic ends up looking like this:
Transcript · tap any word to jump there
Start by naming the quantity. Value at Risk is a loss quantile for a stated portfolio, horizon and confidence level, under a particular model. A 99% one-day VaR of two million dollars is not a promise that losses cannot exceed two million. It says that, according to the model and its assumptions, the loss threshold is exceeded in about one per cent of outcomes over that horizon.
Suppose a question asks for an approximate ten-day VaR and invites square-root-of-time scaling. The calculation is two million multiplied by the square root of ten, or about 6.32 million dollars. Before accepting it, check what has been assumed: volatility is treated as stable, and returns are treated as independent in a way that supports this scaling. If volatility changes or returns are dependent, the approximation may not represent the ten-day risk well.
Finally, interpret the estimate. It is a threshold, not an estimate of the average loss in the worst one per cent of outcomes. It does not describe how severe a tail loss could be. State the assumptions, keep the horizon and units visible, and ask what additional measure or control would help decision-makers understand the risk that VaR omits.
A repeatable weekly routine
Set your routine around the hours you can sustain and the official material you need to cover. In each study session, combine retrieval with problem solving; avoid spending the whole session highlighting or watching explanations passively. A workable 90-minute session could be:
- 10 minutes: retrieve key definitions or formulas from memory.
- 35 minutes: study one learning objective and add concise notes to the board.
- 30 minutes: answer related questions, including at least one calculation if relevant.
- 15 minutes: review errors, update the core checklist and schedule a later retest.
At the start of each week, choose objectives from your current part’s official materials. Across the week, revisit earlier sections as well as adding new ones. Reserve a longer session for a timed mixed set when you have enough coverage to make that useful, then use its results to choose the next week’s priorities. As the exam approaches, spend more time on timed retrieval and reviewing patterns in your errors; do not stop checking the official instructions.
The exact number of study weeks and hours depends on your background, available time and the materials you need to cover. Make a plan you can measure: objectives studied, questions attempted, errors diagnosed and weak areas retested. A score without that error review tells you less than it seems.
How MySummaries helps
MySummaries can turn your own FRM notes and study materials into revision boards, flashcards, written practice papers marked against your notes, and audio lectures. Use the board to keep formulas beside assumptions and risk interpretations; use marked practice to find repeated calculation or reasoning gaps, then review those areas again. Start at portal.mysummaries.app.