Skip to main content
CAIA Level II · Topic guide

Risk and Risk Management for Alternative Portfolios

Risk management connects market losses, liquidity needs, leverage, counterparties and operations. Use several measures and stress scenarios because a single volatility or VaR number cannot describe every way a portfolio can fail.

The key decisions, at a glance

  • 1. Market risk

    Measure exposures and repricing under adverse assumptions.

    Watch for: A historical estimate is conditional on its data and model.

  • 2. Liquidity risk

    Compare usable cash with calls, spending and collateral.

    Watch for: A solvent balance sheet can still face a cash shortage.

  • 3. Counterparty risk

    Assess default exposure, collateral and settlement terms.

    Watch for: A hedge creates obligations to or dependence on another party.

  • 4. Operational risk

    Test controls around valuation, authorization and reconciliation.

    Watch for: Investment performance does not validate the control environment.

Original worked case

A hedge reduces price risk but creates a cash call

Given: A portfolio has $12m usable cash, $5m spending and $4m expected private calls. A market shock produces a $6m collateral requirement before reliable inflows.

  1. 01 Baseline needs

    $5m + $4m = $9m; baseline surplus is $3m.

  2. 02 Add the shock

    $9m + $6m = $15m of immediate uses.

  3. 03 Liquidity gap

    $12m - $15m = -$3m.

Answer and interpretation

A position that helps economic exposure can still create a $3m funding shortfall. Assess payoff protection and liquidity together.

Change one assumption

If a credit facility is legally available in this scenario, include its usable capacity and costs, then test counterparty withdrawal and limits.

Independent teaching example with the stated assumptions. Not an official exam question, marking key or observed market trend.

Understand the method

Part 1

Distinguish a loss threshold from the severity beyond it

Value at Risk is a loss quantile for a stated horizon and confidence level under a specified model. It is not a maximum possible loss. Expected Shortfall describes the average loss in the defined tail beyond the quantile, with conventions needed for discrete samples. Always state the loss sign, horizon, confidence and method before comparing two reported figures.

For a deliberately simple set of twenty equally likely loss observations ranked from smallest to largest, define 90% historical VaR using the eighteenth ordered observation and define the tail as the two worst observations. If those three values are $4m, $6m and $10m, VaR is $4m and the average of the worst two losses is $8m. Other quantile interpolation conventions can produce different sample estimates.

Part 2

Use scenarios to expose interactions outside a single measure

A historical estimate can miss a shock that did not occur in its sample. A scenario instead specifies a combination of market moves and operational or funding conditions, then traces their effect on the portfolio. Keep scenario assumptions explicit and distinguish an exploratory stress from a probability forecast. The usefulness lies in revealing vulnerability and possible actions.

For alternatives, combine falling asset values with slower distributions, capital calls, higher collateral requirements and reduced sale proceeds. Those conditions can occur together. Stress tests that assume perfect liquidity precisely when markets fall can understate the need for cash. Also examine reverse stress: identify what combination of events would make the institution unable to meet a critical obligation.

Part 3

Choose a benchmark that matches the investment question

A benchmark provides a reference for evaluating decisions. A public index may be useful for opportunity cost, but differences in leverage, cash-flow timing, liquidity and valuation can make a simple return comparison misleading. State whether the benchmark tests policy allocation, manager implementation or the investor’s alternative use of capital. One benchmark may not answer all three.

Attribution then decomposes the result into specified sources, such as allocation and selection, factor exposures or cash-flow effects. Maintain consistency between gross and net returns, currency, period and valuation dates. A residual labelled alpha can contain omitted factors, stale pricing or model error. Explain the unexplained portion rather than treating it as proven skill.

Part 4

A hedge can reduce one risk while creating a cash requirement

A derivative hedge can offset changes in portfolio value but require daily collateral or margin. The hedged asset may be illiquid and unable to release cash when the hedge loses value. Economic risk reduction and funding resilience therefore need separate analysis. Basis risk also remains when the hedge instrument and the protected asset do not move identically.

A risk-management system should connect exposures to limits, monitoring, escalation and decision authority. Define what happens when a limit is breached and who can act. A dashboard that reports a problem without a response process is incomplete. Test data quality and reconciliation too: a precise risk number derived from missing positions is not a reliable control.

Each risk measure answers a different question

Volatility

Question answered
How dispersed are returns?
Key limitation
Does not describe every tail or funding event

VaR

Question answered
Where is a selected loss quantile?
Key limitation
Not the maximum or average tail loss

Expected Shortfall

Question answered
How severe is the modelled tail on average?
Key limitation
Depends on tail data and model assumptions

Funding stress

Question answered
Can dated cash obligations be met?
Key limitation
Depends on access assumptions and joint shocks

More worked practice

Original teaching cases. Use the stated assumptions; these are not official exam questions or marking keys.

Equal VaR can hide very different tail severity

Given: Two portfolios each have twenty equally likely one-period loss scenarios. Under the chosen nearest-rank 90% VaR convention, both eighteenth ordered losses are $4m. A’s two worst losses are $5m and $7m; B’s are $5m and $19m.

  1. 1The specified 90% VaR is $4m for both portfolios.
  2. 2Using the average of the two worst observations as the tail measure, A has (5 + 7) / 2 = $6m.
  3. 3B has (5 + 19) / 2 = $12m, twice A’s tail average.

Interpretation: The same quantile threshold does not imply the same severity beyond that threshold. State the sample convention and examine both tail losses and scenarios before comparing risk budgets.

Check your understanding

Answer first, then open the explanation. These checks are original practice and do not generate an exam score.

1. Does 99% VaR cap the next day’s loss?

No. It is a modelled quantile for the stated horizon; losses can exceed it.

2. Can reported alpha include measurement error?

Yes. Omitted exposures, valuation lags and inconsistent return bases can enter the residual.

3. Why test risk-limit escalation?

The system must turn a detected breach into an authorized decision. A limit without ownership and response procedures may not constrain risk.

Get the free Level II PDF by emailStudy CAIA with a free account

Common questions

Does lower measured volatility mean lower total risk?

No. Stale marks, nonlinear payoffs, leverage and illiquidity may be poorly represented by a simple volatility estimate.

Why add expected shortfall to VaR?

Expected shortfall describes average losses in the modeled tail beyond the quantile. It still depends on model and data assumptions.

Sources and scope

The sources below establish their stated framework or subject context. Calculations and cases are our own teaching material; these sources do not endorse them or establish complete exam coverage.

By MiloPrep editorial team. Updated 10 October 2026. Independent study notes, not affiliated with or endorsed by CAIA Association. How we prepare our content.

Notes

  • * Study scope: Selected study concepts, not complete curriculum coverage. Match your sitting to the official exam-year materials; CAIA’s current resource page now links to 2027. ↩ Return to content