How does the focus change from Level I to Level II?

Level I builds the vocabulary and mechanics of alternative assets, including how their cash flows, leverage, and investment structures work. Level II asks you to apply those ideas to allocation, risk, access routes, manager selection, and ethical decisions.

Why should I examine both IRR and a cash multiple for a private fund?

A cash multiple compares value received with capital invested, while IRR also reflects when contributions and distributions occur. An early distribution can raise IRR without increasing the total proceeds, so inspect the dated cash flows alongside both measures.

Why might an alternatives portfolio be less diversified than it appears?

Different strategy labels can conceal shared exposure to credit conditions, leverage, or a need for liquidity. In a stress scenario, losses and funding demands may arrive together. Look through holdings and financing terms, and question correlations estimated from infrequent valuations.

CAIA Level I and II Study Guide: 2027 Overview

Start with concept 1 Choose a topic

Alternative investments challenge familiar portfolio habits: reported values may lag market conditions, cash can be tied up, and legal claims matter as much as quoted prices. This guide connects the foundations of CAIA Level I with the applied decisions of Level II for the March and September 2027 curriculum. You will examine how real assets, private capital, hedge funds, and digital assets earn returns and transmit risk, then use that understanding to assess allocation, liquidity, managers, and investment terms. The focus is on decisions you can explain: tracing cash flows, recognizing leverage, interpreting fund returns, testing diversification, and considering losses beyond a risk cutoff. Ethical duties run through both levels, from identifying conflicts to documenting a defensible choice.

Level I - CAIA Ethical Principles

1. Whose interests govern?

## Whose interests govern? Start an investment recommendation by identifying who supplied the capital and whose objectives the decision is meant to serve. That person or institution is the principal; an adviser or manager acting on its behalf is an agent. The agent brings expertise and may exercise discretion, but that authority does not replace the client's return needs, risk tolerance, liquidity needs or other agreed constraints. Employees, counterparties and communities may also be affected stakeholders. Their interests deserve attention, particularly when neglect could damage long-term value, yet they do not turn the agent's compensation into the client's objective. A useful decision sequence is to define the client's need, compare suitable choices on a consistent basis, identify benefits to the agent, and record why the recommended choice serves the client. The 2027 Level I companion guide identifies principals, agents and stakeholders within this topic.

Worked example: A client needs diversified private-credit exposure. Two funds have comparable portfolios, risk controls and access terms, but one charges the client 1.5% annually and the other 1.0%. On a hypothetical 1 million investment, that is 15,000 versus 10,000 per year before any compounding. The client is the principal; the manager's preference for the higher fee supplies no client benefit. Recommend the lower-cost suitable fund.

Watch the distinction: The manager's expertise does not make the manager the principal. Delegated discretion is a means of carrying out the client's objective, not a reason to favor the manager's revenue.

Reference: [2]

2. Conflicts in a fund recommendation

## Conflicts in a fund recommendation A conflict exists when an adviser's financial or personal interest could pull a recommendation away from the client's interest. Placement fees, referral arrangements and close manager relationships are common examples. The conflict matters even if the adviser believes the recommendation is impartial, because judgment can be affected without deliberate dishonesty. Identify who pays the benefit, when it is earned and whether it changes with the client's choice. Then compare the available funds on expected net return, risk, liquidity, terms and fit with the client's mandate. Disclose the conflict clearly enough for the client or decision maker to understand it, and use an independent review or remove the conflicted person when appropriate. Disclosure informs the client; it does not supply missing evidence of suitability or make an inferior recommendation sound.

Worked example: An adviser compares two otherwise similar funds. Fund A pays the adviser a hypothetical 2% placement fee; Fund B pays none. For a 1 million allocation, A would pay the adviser 20,000. The adviser should disclose that incentive, document a fair comparison and recommend A only if a client-centered reason supports it. If the funds serve the client equally well and A costs more, disclosure alone cannot justify choosing A.

Watch the distinction: A disclosed conflict is still a conflict. The client needs a recommendation supported by its own merits, not merely an explanation of how the adviser benefits.

Reference: [2]

3. Stewardship after investment

## Stewardship after investment Capital allocation begins an investor's responsibility; it does not end it. Stewardship means monitoring how a manager uses the capital and responding when decisions threaten the asset's durable value. Review operating results alongside leading indicators such as maintenance, employee turnover, customer retention and compliance issues. A rise in reported earnings can reflect genuine efficiency or the postponement of necessary spending. To distinguish them, ask what cost has been deferred, when it must be paid and what failure could cost. The investor can request evidence, engage the manager, vote where rights permit or reconsider the mandate. The response should match the investor's authority and the materiality of the issue. Stakeholder harm can become an investment loss through disrupted operations, repair costs or weaker demand, so stewardship joins financial analysis with practical oversight.

Worked example: A portfolio company postpones safety maintenance costing a hypothetical 2 million and reports earnings 2 million higher this year. If the delay could produce a 6 million repair next year plus lost operating time, the current increase is weak evidence of value creation. The investor should ask the manager for a maintenance schedule, assess the possible loss and press for a plan that protects the asset and its users.

Watch the distinction: Higher current earnings do not prove better management. Earnings may rise because a necessary expense was delayed, leaving investors exposed to a larger future loss.

Reference: [2]

4. Acting when a rule is silent

## Acting when a rule is silent Contracts cannot specify the right response to every judgment call. When an action is technically permitted, examine its purpose and effects before relying on the permission. First identify the clients and other parties affected. Next gather the best available evidence, consider reasonable alternatives and ask how each choice would change reported value or decisions based on it. Seek review from people with suitable independence and authority, and preserve the reasoning in a record. If someone pressures the investment professional to use a weak assumption, escalate the concern through the fund's governance process. A principles-based approach is especially useful in valuation, where estimates contain judgment and stale marks can influence subscriptions, redemptions or fees. The aim is a supportable estimate and fair communication of uncertainty, rather than a convenient reading of broad discretion.

Worked example: A fund's contract permits valuation discretion. A comparable transaction suggests a holding is worth 80 rather than its recorded 100, but a senior colleague asks to retain 100 until fundraising ends. Existing and incoming investors could transact on the mark. The analyst should test the comparison, seek a supportable fair value, disclose material uncertainty through the proper process and escalate pressure to keep an unsupported mark.

Watch the distinction: Contractual discretion is not permission to choose whichever estimate helps fundraising. The valuation still needs a defensible method, evidence and fair treatment of affected investors.

Reference: [2]

Level I - Introduction to Alternative Investments

5. What makes an investment alternative?

## What makes an investment alternative? The traditional and alternative labels are useful starting points, but an investment may share features of both. Classify it along four dimensions: the underlying asset, the return strategy, the legal or fund structure, and the investor's route to cash. Real estate, for example, can be owned directly, held through a private fund or accessed through a listed security. Each route gives exposure to property economics, while trading, governance and valuation differ. A listed share may trade each market day even when its buildings cannot be sold quickly; its price can also move with broad equity sentiment. A direct building requires a negotiated transaction and ongoing operating oversight. For each opportunity, ask what generates cash, what can impair that cash, how value is observed and how an exit would work. The Level I study guide is a useful companion for examining these dimensions.

Worked example: An investor compares a listed real estate company share with a direct office building. Both depend partly on rent and occupancy. The share can be sold through an exchange at the available market price; selling the building requires finding a buyer, negotiation and closing. Listing improves the investor's ability to trade the security, but it does not make the underlying property readily saleable.

Watch the distinction: Calling both holdings “real estate” does not make their liquidity or risk identical. The ownership route changes trading, control, costs and how quickly market prices respond.

Reference: [2]

6. The price of illiquidity

## The price of illiquidity An illiquid investment may offer a higher expected return because investors give up the ability to exit quickly. The difference is a potential illiquidity premium only after other differences have been considered. Compare loans with similar borrower quality, seniority, collateral, maturity, currency and covenant protection. Then adjust for origination and management fees, expected defaults and recovery, and the cost of holding cash for commitments. A private loan's quoted yield can exceed a traded loan's yield because its borrower is weaker or the loan carries more credit risk. Illiquidity also has a cost that depends on the investor: an owner with predictable long-term liabilities may bear a lockup more easily than one with near-term payments. Work from net expected cash flows and a realistic funding plan, rather than attributing the entire yield spread to one feature.

Worked example: A hypothetical private loan yields 9% and a traded loan yields 6%, a spread of 3 percentage points. Suppose closer analysis attributes 1 point to higher expected credit loss and 0.5 point to additional fees. The remaining 1.5 points is only a candidate for compensation for lockup and other differences; it is not a measured illiquidity premium until those differences are examined.

Watch the distinction: A three-point yield gap is not automatically a three-point liquidity reward. Credit quality, fees and loan terms can account for some or all of it.

Reference: [2]

7. Cash timing changes returns

## Cash timing changes returns Internal rate of return, or IRR, is the discount rate that makes the present value of an investment's cash flows equal zero. For an investor, contributions are outflows and distributions are inflows. Put each amount at its actual date, then solve for the rate that balances them. With one contribution and one final receipt, the calculation resembles compound annual growth. With several calls and distributions, earlier proceeds usually raise IRR relative to otherwise identical later proceeds, because less capital remains outstanding for as long. This makes timing essential in private investments. IRR alone does not reveal the total amount of wealth created, and interim IRRs may depend heavily on estimated residual values. Read it beside cash multiples and the cash-flow schedule. When comparing projects, also check scale and whether the cash-flow patterns are genuinely comparable.

Worked example: Invest 100 currency units now and receive 121 exactly two years later, with no interim flows or fees. Solve 100(1 + r)² = 121, so (1 + r)² = 1.21 and r = √1.21 − 1 = 0.10. The annual IRR is 10%. The total gain is 21 units, or 21% of the original 100, over the full two years.

Watch the distinction: A 21% total gain over two years is not a 21% annual IRR. Compounding and the dates of cash flows determine the annual rate.

Reference: [2]

8. Evergreen and drawdown funds

## Evergreen and drawdown funds An evergreen fund generally accepts subscriptions on a continuing basis and has no predetermined liquidation date. Investors usually pay the subscribed amount when they enter, then seek liquidity under the fund's dealing terms. Those terms can restrict withdrawals, particularly when assets are hard to sell. A drawdown fund instead accepts a commitment and calls portions of it as investments and expenses arise. The uncalled amount remains an obligation, so an investor must plan for calls even while that cash sits outside the fund. Compare the structures by mapping the timing of payments, available exit routes and how quickly the manager can deploy capital. A drawdown arrangement may reduce idle cash inside the fund, but it transfers cash-management work to the investor. An evergreen subscription may be simpler to fund while still exposing the investor to redemption limits.

Worked example: An investor commits 10 million currency units to a drawdown fund. A call for 30% requires 10 million × 0.30 = 3 million now, leaving 7 million callable under the commitment. If the investor instead subscribes 10 million to an evergreen fund upfront, the full 10 million is paid at entry. Future access to that money depends on the evergreen fund's stated redemption terms.

Watch the distinction: An uncalled commitment is not spare money free of obligation. The investor needs a credible source of cash when the manager issues a valid capital call.

Reference: [2]

9. Diversification beyond labels

## Diversification beyond labels Diversification depends on how holdings behave together, not on whether they sit in different product categories. Correlation summarizes how returns have moved together in a sample, while factor exposure helps explain why that relationship might change. A long-short fund can hold both long and short positions yet retain net equity exposure. It may also depend on credit conditions, funding markets or a particular industry. Examine holdings and estimated sensitivities, then ask what could make several strategies lose money at once. Normal-period correlation is a useful clue, but stress periods deserve separate attention because positions may become more alike when liquidity is scarce. Judge the proposed holding against the portfolio's existing risks and the role it is meant to play. CAIA practice questions can help you rehearse exposure-based reasoning.

Worked example: An equity index falls 20% in a stress period, while a long-short fund falls 12%. Its shorts softened the loss, but the fund did not protect the portfolio from that event. If its longs exceed its shorts and both positions respond adversely to stress, positive net equity exposure offers a plausible explanation. The investor should examine factor and position data before calling the fund a diversifier.

Watch the distinction: A different fund label does not establish a different risk source. Look for shared exposures and behavior during the events against which diversification is supposed to help.

Reference: [2]

Level I - Real Assets

10. Property income and value

## Property income and value Net operating income, or NOI, starts with property revenue and subtracts ordinary operating expenses needed to run the property. It is an unlevered operating measure: interest payments, principal repayments and the investor's income taxes do not belong in NOI. Check whether the rent figure allows for vacancy and whether the cost figure includes recurring items such as maintenance and property management. A capitalization rate turns a stabilized annual NOI into a simple value estimate: value equals NOI divided by the cap rate. The cap rate reflects the return required for the property's income and risk, so a higher cap rate produces a lower value for unchanged NOI. This shortcut assumes the income measure is representative and the selected cap rate fits the asset. It does not replace a cash-flow analysis when leases, renovation costs or growth vary materially over time.

Worked example: A property collects a hypothetical 1.2 million currency units in annual rent and incurs 0.4 million in annual operating costs. NOI is 1.2 − 0.4 = 0.8 million per year. At an assumed 5% capitalization rate, estimated value is 0.8 million ÷ 0.05 = 16 million. The estimate is before financing and depends on the stated income being sustainable.

Watch the distinction: Subtracting debt interest to calculate NOI mixes property operations with financing. Debt affects the equity investor's cash flow and value, but it is separate from unlevered NOI.

Reference: [2]

11. Commodity futures returns

## Commodity futures returns A long commodity futures position can gain or lose through changes in the futures price, and a strategy that maintains exposure must replace contracts as they approach expiry. The gap between the expiring contract and the next contract is called the roll spread. When the next contract costs more, buying it gives exposure at a higher futures price; if that price subsequently falls toward an unchanged spot price, the long position loses from convergence. Selling one contract and buying another does not by itself realize that entire gap as an immediate loss. The outcome depends on what prices do after the roll, along with collateral returns and trading costs. Separate these components when evaluating performance: spot movement, futures convergence or roll effect, collateral income and expenses. The Level I companion guide places commodities within real assets.

Worked example: A long investor closes an expiring contract at 100 currency units per unit and buys the next contract at 103. Assume spot stays at 100, the new contract converges to 100 by expiry, and collateral income and costs are zero. The new position loses 103 − 100 = 3 per unit. The 3 loss occurs through the subsequent futures-price decline, not merely from exchanging contracts.

Watch the distinction: A higher next-contract price does not create an instant cash loss equal to the price gap. It creates exposure to a potential convergence loss under the stated price path.

Reference: [2]

12. Contracted real-asset cash flows

## Contracted real-asset cash flows Infrastructure and intellectual property can both generate recurring payments, but their cash-flow drivers differ. An infrastructure investor should identify the physical asset, who pays to use it, how the tariff is set and who bears operating and renewal costs. A toll road may collect a payment per vehicle, making traffic volume important even if the concession lasts for years. An intellectual property investor should identify the legally supported right, the licensed use, the royalty formula and the period during which payments can be collected. A patent royalty may depend on units sold and could fall sharply when the right expires or a license ends. For either asset, forecast revenue as volume times price, then test what happens if demand changes or contractual protection weakens. Contract duration can support predictability, but it cannot create customers or eliminate operating costs.

Worked example: Suppose a road earns 2 currency units per vehicle and traffic falls from 1 million to 0.8 million vehicles a year. Toll revenue falls from 2 million to 1.6 million, before costs. A patent licenses a product at 1 unit per sale; 1 million sales produce 1 million of royalty revenue. If the relevant right expires and no replacement agreement exists, that contractual stream may end.

Watch the distinction: A long contract does not guarantee long-lived cash flow. Demand, operating obligations, enforceability and renewal rights determine whether payments continue and how much remains for investors.

Reference: [2]

Level I - Private Equity

13. Stages of private equity

## Stages of private equity Private equity strategies differ largely in the maturity of the company and the plan for creating value. Venture capital funds young businesses with uncertain products or markets; the capital often supports development and growth before steady profits exist. Growth equity usually backs a more established company that has evidence its business works and needs capital to expand. A buyout acquires control of a mature business, often using debt, and aims to improve operations, cash flow or the eventual sale price. These categories can overlap, so classify the particular deal by the company's condition, the use of proceeds, the control obtained and the source of the expected payoff. A high projected return does not itself make a deal venture capital. The loss pattern and funding needs can differ greatly across stages, especially when leverage adds fixed payment obligations.

Worked example: A profitable software company raises equity to enter two new markets and takes on little debt. Its operating business is established, while the money finances expansion, so growth equity is the best fit. A startup funding an unfinished product is closer to venture capital. Buying control of a mature company with substantial acquisition debt is a buyout, even if the buyer also plans growth.

Watch the distinction: All private-company equity is not venture capital. Company maturity, capital use, control and financing structure give a more useful classification than private ownership alone.

Reference: [2]

14. Leverage in a buyout

## Leverage in a buyout In a leveraged buyout, debt funds part of the purchase and equity supplies the remainder. At exit, the business sale proceeds first cover the debt claim; equity receives the residual. Holding debt constant makes the amplification easy to see: a fixed change in business value is spread over a smaller initial equity investment. If the business rises, equity's percentage gain exceeds the unlevered business gain. If it falls, the same mechanism magnifies the equity percentage loss and can eliminate equity altogether. Real deals require further adjustments for interest, principal repayments, fees, taxes and interim distributions. Debt also places demands on operating cash, which can force action before the intended exit. When assessing a proposed buyout, calculate both the expected equity payoff and a downside case in which lenders must still be paid.

Worked example: Buy a business for 100 currency units using 60 of debt and 40 of equity. Assume debt remains 60 and ignore interest, fees, taxes and interim cash flows. A sale at 120 leaves equity 120 − 60 = 60, a gain of 20 ÷ 40 = 50%. A sale at 80 leaves equity 20, a loss of 20 ÷ 40 = 50%.

Watch the distinction: A 20% change in enterprise value need not mean a 20% change in equity value. Debt is paid first, so the residual equity absorbs a larger percentage swing.

Reference: [2]

15. Order of fund distributions

## Order of fund distributions A distribution waterfall sets the sequence for dividing proceeds between fund investors and the manager. Begin with the fund agreement: it determines which contributions count as capital, whether investors receive a preferred return, when carried interest begins and how earlier distributions affect later calculations. In a deliberately simple case, proceeds first return invested capital. Remaining proceeds are profit, and a stated carry percentage gives the manager a share of that profit. More elaborate waterfalls can apply deal by deal or across the whole fund, add a hurdle and catch-up, or later reclaim excess carry. The order matters because a manager may receive cash before the final outcome of all investments is known. For calculations, label each tier, subtract the amount allocated there and check that all recipients' distributions sum to the available proceeds.

Worked example: A hypothetical fund invested 100 currency units and now distributes 120. Assume no fees, preferred return, earlier distributions or other waterfall tiers, and a 20% carry on profit after capital is returned. First return 100 to investors. Profit is 120 − 100 = 20; manager carry is 20 × 0.20 = 4. Investors receive the other 16, or 116 total. Check: 116 + 4 = 120.

Watch the distinction: Twenty percent carry on profit is not twenty percent of all proceeds under this rule. Returning the investors' 100 of capital comes before splitting the 20 profit.

Reference: [2]

Level I - Private Debt

16. Direct-lending capacity

## Direct-lending capacity A lender wants to know whether the borrower can make required cash payments when they fall due. Debt-service coverage compares cash available for debt service with scheduled interest and principal payments over the same period. Divide the available cash by those payments, keeping the units and period consistent. A ratio above one shows modeled cash exceeds scheduled service, but the size and reliability of the margin matter. Review how cash is defined: a strong accounting profit may coexist with weak cash collections or heavy maintenance spending. Then add the proposed loan's payments to existing debt service and repeat the calculation. Stress revenue, operating costs and interest costs to see how quickly the margin disappears. Collateral and covenants can improve a lender's position, but they do not generate operating cash to make routine payments.

Worked example: A borrower has a hypothetical 15 million currency units of annual cash available for debt service and 10 million of annual required payments. Coverage is 15 ÷ 10 = 1.5 times. A proposed loan adds 4 million of annual payments, giving 15 ÷ (10 + 4) = 1.0714, about 1.07 times. The modeled cushion falls from 5 million to 1 million before any cash-flow stress.

Watch the distinction: A borrower can appear profitable yet lack cash for scheduled payments. Coverage must use relevant cash flow and include the new loan's full required debt service.

Reference: [2]

17. Where private lenders rank

## Where private lenders rank Claim priority determines who receives value when a borrower cannot pay everyone. Senior debt generally has a contractual claim ahead of subordinated or mezzanine debt, often supported by stronger security. Mezzanine lenders accept a lower position and therefore depend more heavily on value remaining after senior claims are satisfied. Distressed debt describes the borrower's troubled condition or the debt's impaired trading situation; it is not itself a fixed place in the priority order. A distressed loan could be senior or junior. To analyze recovery, establish the cash available after relevant costs, apply it to claims in their agreed order and divide each payment by its claim to find its recovery rate. Check collateral and legal terms because priority can vary across assets and entities. A quoted yield is meaningful only beside expected recovery and the risk of delayed payment.

Worked example: Assume a liquidation leaves 70 million currency units for creditors, with 60 million of senior claims and 30 million of mezzanine claims, and no other claims or costs. Senior lenders receive 60 million, a 100% recovery. The remaining 70 − 60 = 10 million goes to mezzanine lenders. Their recovery rate is 10 ÷ 30 = 33.3%, leaving 20 million of their claim unpaid.

Watch the distinction: Distressed debt is not automatically junior debt. Identify both the borrower's condition and the instrument's contractual rank before estimating who bears a loss.

Reference: [2]

18. Tranches absorb loss in order

## Tranches absorb loss in order Securitization pools assets, such as loans, and divides the claims on their cash flows into tranches. In a simplified structure, the equity tranche absorbs the first collateral losses, the mezzanine tranche absorbs the next losses, and the senior tranche is affected only after the lower layers are exhausted. This subordination protects senior claims up to a point; it does not remove the underlying loan risk. Begin with the pool's original balance, calculate the dollar loss and apply it from the bottom upward. Then check the remaining pool value against the surviving tranche balances. Real structures may also include interest waterfalls, reserve accounts, triggers and differences in collateral timing. Those terms can change cash distributions before the ultimate principal loss is known. For a basic loss-allocation problem, the order and attachment points are the essential first steps.

Worked example: A hypothetical 100 million currency unit loan pool funds 10 million of equity, 20 million of mezzanine and 70 million of senior claims. A 15 million principal loss leaves pool value of 85 million. The first 10 million loss wipes out equity; the next 5 million reduces mezzanine from 20 to 15 million. Senior remains at 70 million. Check: 15 + 70 = 85 million of surviving claims.

Watch the distinction: A senior tranche can still lose if collateral losses pass through all subordinated protection. Senior status determines loss order, not an unconditional guarantee of repayment.

Reference: [2]

Level I - Hedge Funds

19. Long-short exposure

## Long-short exposure A long-short fund buys assets expected to rise and sells short assets expected to lag or fall. Express each side as a percentage of the fund's equity or net asset value. Gross exposure adds the absolute long and short amounts and indicates how much market value is at work. Net exposure subtracts shorts from longs and gives a first approximation of directional exposure. Both measures are needed: two portfolios can have the same net exposure while one carries far larger gross positions, financing needs and sensitivity to relative price moves. Net exposure is also an imperfect risk measure because longs and shorts may have different market betas or sector exposures. To understand a fund, calculate the arithmetic first, then inspect what the positions represent and how borrowing and short-sale costs affect outcomes.

Worked example: A fund with 100 million currency units of NAV holds 130 million in long positions and 80 million in shorts. Long exposure is 130% and short exposure is 80% of NAV. Gross exposure is 130% + 80% = 210%; net exposure is 130% − 80% = 50% long. If both sides had identical market sensitivity, a broad decline would tend to hurt this positive net position, before other effects.

Watch the distinction: The short book does not automatically neutralize the long book. Net exposure may remain positive, and unlike assets can respond differently to the same market move.

Reference: [2]

20. Market neutral is a risk claim

## Market neutral is a risk claim A fund may call itself market neutral because its long and short positions have equal dollar value. Equal dollars remove net dollar exposure, but market sensitivity depends on what is owned and sold. Estimate each side's beta, then compare sector, size, style and currency exposures. A portfolio long small technology firms and short large utilities can lose if technology weakens relative to utilities even when a broad index barely moves. Financing adds another dimension: short positions may incur borrow costs or be recalled, while margin demands can force sales. Test neutrality against the risk the fund claims to hedge and across plausible stress scenarios. A manager can aim to offset broad market beta while intentionally taking relative-value risks; the investor needs to know which risks remain and whether they fit the portfolio.

Worked example: A fund has 100 million currency units long in small technology companies and 100 million short in large utilities. Net dollar exposure is 100 − 100 = 0. Suppose technology falls 10% and utilities are unchanged, ignoring borrow costs and other effects. The long book loses 10 million and the short book gains zero, so the fund loses 10 million despite its zero net dollars.

Watch the distinction: Zero net dollars is an arithmetic fact, not proof of zero risk. Factor differences, sector moves and short financing can produce substantial gains or losses.

Reference: [2]

21. High-water-mark fees

A high-water mark sets a reference value that a fund must exceed before charging a new incentive fee. It keeps a manager from collecting a performance fee for merely recovering a loss that investors have already borne. First identify the relevant mark and the net asset value before the proposed incentive fee. Then calculate the amount above the mark and apply the stated fee rate only to that amount. Finally subtract the fee to find the investor's closing value. The order matters: charging a percentage of the entire gain from a depressed starting value would reward recovery twice. Actual fund terms may also specify management fees, hurdles, subscriptions, withdrawals, and how the mark is adjusted. Read those terms before applying a formula. The 2027 Level I companion guide includes high-water marks and hedge fund fee calculations.

Worked example: Suppose a fund's per-share net asset value begins at 90, its prior high-water mark is 100, and its value before an incentive fee reaches 108. Under a simplified rule charging 20% only above the mark, the eligible gain is 108 − 100 = 8. The fee is 0.20 × 8 = 1.6 per share. Closing value after that fee is 108 − 1.6 = 106.4 per share. The rise from 90 to 100 attracts no incentive fee.

Watch the distinction: Charging 20% on the full rise from 90 to 108 produces a fee of 3.6. That mistakenly treats recovery to the old high-water mark as newly earned performance.

Reference: [2]

Level I - Digital Assets

22. Ownership, keys and custody

A blockchain can record that an address holds a token, but a valid record alone does not give someone practical control. A transfer normally requires a transaction authorized with the relevant private signing key. The address identifies where the asset is recorded; the key lets its controller authorize movement under the network's rules. Ask three separate questions: what does the ledger show, who can sign, and what claim does the investor have against any custodian? Self-custody places key storage and recovery practices close to the investor. Custody through an intermediary can shift those tasks but adds operational and counterparty exposure. A lost key, compromised key, or failed custodian can therefore have different consequences even when the ledger itself works correctly. Assess backup arrangements, access controls, and the evidence supporting an intermediary's holdings without assuming that a blockchain entry resolves every ownership dispute.

Worked example: An investor's tokens remain visible at an address after the investor loses the only signing key. A block explorer can confirm the recorded balance, but it cannot sign a transfer. If no valid backup or recovery mechanism exists, the investor cannot move the tokens through the ordinary transaction process. If a custodian controlled the key instead, the investor would need to establish and enforce a claim against that custodian; the custody arrangement changes the problem.

Watch the distinction: A visible token balance does not prove that its intended owner can transfer it. Ledger accuracy, control of the signing key, and legal entitlement answer different questions.

Reference: [2]

23. Tokens and smart contracts

A token can be transferred by software while the economic right it purports to represent depends on agreements and events outside the blockchain. For a token linked to property income, identify the entity that owns the property, the investor's enforceable claim on that entity, and the rule for sharing cash after costs. A smart contract can execute a distribution rule once it receives inputs, but it cannot independently verify that tenants paid rent or that an expense was legitimate. That information must reach it through a trusted process or data provider, often called an oracle. Examine who maintains the ownership register, who may amend the code, and how errors or disputes are resolved. Faster transfers may improve recordkeeping, yet they do not by themselves improve the property's income, remove transfer restrictions, or guarantee that reported cash matches money actually received.

Worked example: A property vehicle promises token holders 10% of distributable rent. It collects 100 units of rent and pays 30 units of permitted property costs, leaving 70. If the agreement grants the tokens 10% of that amount, holders collectively receive 7 units. The software can divide 7 among eligible wallets only after reliable records establish rent, costs, and token ownership. A token transfer alone cannot verify those inputs.

Watch the distinction: Calling a token a property share does not establish a direct property right. The governing documents, ownership records, cash-flow definition, and data supplied to the contract determine what holders receive.

Reference: [2]

24. Sizing a crypto allocation

Portfolio weight converts an asset's price move into its direct contribution to portfolio return. For a position with initial weight w and return r, that contribution is w × r, assuming no trading during the period. This arithmetic is a useful first limit on price-loss exposure, but it is only one part of sizing. Check how the asset behaves when the rest of the portfolio is stressed, whether a position can be sold near its quoted price, and how keys or custodial arrangements are protected. A volatile asset can also grow beyond its intended weight after a rally, so a rebalancing rule matters. Compare the potential loss with the owner's spending needs and capacity to hold through a drawdown. The initial weight does not cap losses from other holdings or operational failures, and correlations can change when liquidity becomes scarce.

Worked example: A portfolio begins at 100 units, with 2 in crypto and 98 elsewhere. Crypto falls 60%, leaving 2 × 0.40 = 0.8; other holdings stay at 98. Total value is 98.8, a loss of 1.2 units or 1.2% of the starting portfolio. This measures direct price impact under the stated assumptions. It says nothing about whether the remaining tokens can be transferred or sold promptly if a custodian fails.

Watch the distinction: A small allocation limits its direct arithmetic contribution to a price loss; it does not make custody, trading access, or simultaneous losses elsewhere in the portfolio disappear.

Reference: [2]

Level II - CAIA Ethical Principles

25. An ethical decision in writing

For a client conflict, build the response around the decision, not a list of admirable words. State who the client is and what information the client needs to make or assess the investment decision. Identify the person applying pressure, the parties affected, and the uncertainty that must be explained. Then choose an action that preserves accurate reporting, document its basis, and use the appropriate escalation channel if pressure continues. A delayed valuation matters because investors may rely on the reported value for allocation, withdrawals, or oversight. Present a provisional value as provisional; explain its method and limits instead of allowing an old figure to appear current. Keep a record of the request, the evidence reviewed, and the communication made. The Level II study guide is the relevant companion for practicing an applied written response.

Worked example: A trustee asks an adviser to omit a fund's delayed appraisal from a client report. The adviser should identify the client as the party relying on the report, describe the last available valuation and its date, and explain that the current value remains uncertain. The adviser records the trustee's request, seeks a supportable update, and escalates pressure to conceal the delay. Reporting the stale figure without qualification would mislead the client about its reliability.

Watch the distinction: A truthful number can still mislead when its date and uncertainty are hidden. The duty is to communicate the valuation's status clearly, not merely to avoid fabricating a figure.

Reference: [3]

26. Independence under manager pressure

A manager-selection process can be distorted by benefits that arrive outside the fund's stated fees. Gifts, preferred research access, introductions, or career opportunities may create an incentive to favor one candidate even if the evaluator believes their judgment is unaffected. Identify who receives the benefit, whether it depends on selection, and whether the committee knows about it. Then assess the manager against criteria set before reviewing the offer: strategy fit, risk, fees, operations, and evidence behind performance. Record the offer and disclose it to the decision makers so they can manage the conflict. Refusing or declining to use a benefit may be necessary where independence cannot otherwise be protected. Disclosure is part of the response, but it does not turn a weak candidate into a suitable one. The recommendation must still stand on evidence relevant to the investor.

Worked example: Two managers seek a mandate. One promises the evaluator exclusive research access if selected. The evaluator records that conditional benefit and tells the selection committee before recommending either manager. The committee compares both using the same mandate, performance analysis, risk controls, and fees. If the offering manager scores worse on those measures, its extra access cannot justify selection. Removing the incentive from the decision protects the client's choice.

Watch the distinction: An evaluator's confidence in personal objectivity does not settle a conflict. A benefit tied to selection should be disclosed and managed while the candidates are judged on the client's criteria.

Reference: [3]

27. Stewardship with competing interests

Stewardship continues after an investment is made. When a portfolio company proposes a larger distribution by postponing maintenance, examine the whole path from today's cash saving to future repair costs, downtime, legal exposure, and harm to workers or users. Ask whether the work is optional, when it must be done, and what failure would cost. Compare the proposed distribution with the expected effect on the asset's durable cash flow and sale value. Investor interests can align with stakeholder safety because neglected equipment may destroy future value, but the analysis should identify both effects explicitly. Engage the manager or company for evidence and a funded maintenance plan before approving the payout. If estimates are uncertain, test a severe but plausible case. A short-term earnings increase is meaningful only after the associated future obligation has been recognized.

Worked example: A company can distribute 5 million now by deferring a safety repair estimated at 4 million. Suppose delay creates a 25% chance of a further 12 million loss, in addition to the 4 million repair. The expected extra loss is 0.25 × 12 = 3 million, and the work still costs 4 million, for 7 million of expected future outflow. The investor should challenge the 5 million payout and request a repair plan, while examining harm beyond this simplified estimate.

Watch the distinction: Higher current earnings do not prove higher investment value. Deferred maintenance can shift costs and safety risk into later periods, so distributions must be tested against those obligations.

Reference: [3]

Level II - Institutional Asset Owners

28. Endowment spending across generations

An endowment that aims to serve future beneficiaries must judge returns after spending and inflation. Start with its opening value, apply the investment return, subtract the year's payout, and then express the remainder in opening-period purchasing power. A quick approximation is nominal return minus spending rate minus inflation, but exact results depend on the value used to calculate spending and the timing of cash flows. Fees and new gifts also change the result. A portfolio can report a positive return while its real capital shrinks if payouts and inflation consume more than it earned. Spending rules often smooth payments across market cycles, which can make the current payout differ from a simple percentage of today's assets. Liquidity matters alongside the long-run return target: grants need cash even in years when private holdings cannot be sold readily.

Worked example: Suppose an endowment starts with 100 million, earns 7% during the year, pays 4 million at year-end, and faces 3% inflation. Its value becomes 100 × 1.07 − 4 = 103 million. In starting-year purchasing power, 103 ÷ 1.03 = 100 million, so real capital is unchanged before fees. If the year-end payout were 5 million instead, real capital would be 102 ÷ 1.03 ≈ 99.03 million, a decline of about 0.97 million.

Watch the distinction: A 7% investment return is not a 7% increase in resources available to future beneficiaries. Spending and inflation both reduce what remains in real terms.

Reference: [3]

29. Pension assets and liabilities

For a defined benefit pension, funded status compares assets with the present value of promised benefits. The funded ratio is assets divided by liabilities; the surplus or shortfall is assets minus liabilities. Both sides move. Investment gains raise assets, while changes in discount rates, expected payments, longevity, or benefit terms may change the measured liability. In particular, a lower discount rate generally raises the present value of future payments, which can worsen funded status even if asset returns are positive. Examine the dates and valuation assumptions behind both figures before comparing ratios. Also consider when benefits come due: a plan can have substantial long-term assets yet need liquid holdings to make near-term payments. Liability-aware investing seeks to manage the gap between the behavior of assets and obligations, rather than maximizing asset return in isolation.

Worked example: A plan reports assets of 900 million and liabilities of 1,000 million. Its funded ratio is 900 ÷ 1,000 = 90%, and its shortfall is 900 − 1,000 = −100 million. If assets remain at 900 million while measured liabilities rise to 1,100 million, the ratio falls to 900 ÷ 1,100 ≈ 81.8% and the shortfall becomes 200 million. No asset loss was needed for the plan's measured position to deteriorate.

Watch the distinction: A positive asset return does not establish that a pension improved. Compare the change in assets with the change in the present value and timing of promised benefits.

Reference: [3]

30. Purpose shapes sovereign funds

A sovereign wealth fund's investment policy follows the job its assets must perform. A stabilization pool cushions a budget against volatile revenue, so it needs holdings that can be converted to cash when that revenue falls. A savings pool transfers wealth from a finite resource or current surplus to future generations and may tolerate a longer horizon. A development pool may finance domestic projects, which raises questions about project selection, concentration, and public accountability. A reserve-related pool has its own currency and liquidity needs. Begin by identifying the fund's inflows, expected withdrawals, currency of spending, and governance. Then match asset duration and liquidity to those cash flows. A high expected return from an illiquid investment can be unsuitable if the government may need to sell it during the same downturn that cuts tax or export receipts.

Worked example: A commodity exporter places 100 units in a stabilization reserve and expects that a revenue shock could require 25 units for the budget within a month. Holding all 100 in projects that take years to sell would leave the reserve unable to perform its purpose. At least the potential near-term draw needs a credible source of timely cash, allowing for market stress. A separate savings pool with no comparable withdrawal demand can consider longer-lived investments.

Watch the distinction: A sovereign label does not imply one investment horizon. The stabilization, savings, reserve, and development purposes create different cash needs and tolerances for illiquidity.

Reference: [3]

31. The family balance sheet

A family office should assess an alternative investment against the family's entire economic position, including a private business, property, expected earnings, debts, and future spending. A concentrated operating company may already dominate exposure to an industry, region, or currency. A new fund with a different manager can still increase the same underlying risk. Map the business's revenue drivers and financing needs alongside those of the proposed investment. Then ask how much liquid wealth must remain available for living costs, taxes where applicable, business support, and succession transfers. An asset can appear attractive on its own yet be a poor addition because it cannot be sold when the family most needs cash. Decision rights also matter: an investment plan should remain workable when ownership passes to multiple heirs with different needs and risk tolerances.

Worked example: A family has 100 million of wealth: 70 million in its manufacturing business and 30 million in other assets. It considers putting 10 million of the liquid assets into a private fund concentrated in suppliers to that same industry. If both the business and suppliers decline in an industry downturn, the family's common exposure increases to as much as 80 million. The new manager adds a vehicle, but little economic diversification; only 20 million remains outside those linked positions.

Watch the distinction: Counting managers or funds is not enough to measure diversification. Look through each holding to the businesses, economic drivers, and cash demands shared with the family's existing wealth.

Reference: [3]

Level II - Asset Allocation

32. Objectives before allocation

Asset allocation starts with what the owner must accomplish. Write down spending or liability payments, the desired preservation of purchasing power, the investment horizon, and the size of losses the owner can endure without abandoning the plan. Translate these into requirements for cash, expected long-run return, inflation sensitivity, and limits on concentration or illiquidity. Separate a return objective from a cash-flow requirement: a promising private investment may help long-run growth but cannot fund a payment due next month. Distinguish internal constraints, such as governance capacity or spending policy, from external ones that apply to the owner. Only then compare candidate allocations and test them under weak markets and delayed private distributions. The 2027 Level II companion guide covers asset-owner objectives, constraints, and investment policy statements.

Worked example: A foundation has a 100 million portfolio, must pay 5 million each year, and wants to maintain real capital. It needs a dependable way to supply the next 5 million without forced private-asset sales. If inflation is assumed to be 3% for a planning illustration, earning 8% before fees only approximately covers 5% spending plus inflation; cash-flow timing can change the exact result. Those requirements should be specified before choosing a private-asset percentage.

Watch the distinction: Setting a private-asset target from an expected-return table skips the owner's obligations. A feasible allocation must finance payments and tolerate adverse timing as well as pursue growth.

Reference: [3]

33. Covariance drives portfolio risk

Two-asset portfolio risk depends on each asset's volatility and how their returns move together. With weights w₁ and w₂, volatilities σ₁ and σ₂, and correlation ρ, variance equals w₁²σ₁² + w₂²σ₂² + 2w₁w₂ρσ₁σ₂. Take the square root to obtain volatility. When correlation is below one, the combined volatility can be less than the weighted average of standalone volatilities because some fluctuations offset. A zero correlation removes the cross term; it does not remove either asset's own variance. Check that weights sum to one and use returns and volatilities in consistent decimal units. Historical correlation is an estimate, not a promise. Private valuations that update slowly can make measured correlation look artificially low, so examine underlying economic exposures and stress behavior before relying on the apparent diversification.

Worked example: Suppose two assets each receive half the portfolio. Their annual volatilities are 10% and 20%, with assumed correlation zero. Variance is 0.5²(0.10²) + 0.5²(0.20²) + 2(0.5)(0.5)(0)(0.10)(0.20) = 0.0125. Annual volatility is √0.0125 ≈ 0.1118, or 11.18%. The weighted average standalone volatility is 15%, so imperfect co-movement lowers measured portfolio volatility under these assumptions.

Watch the distinction: A zero correlation does not produce a riskless portfolio. Each holding still fluctuates, and the estimated relationship between them may change during a market shock.

Reference: [3]

34. Rebalancing with illiquid assets

When public prices fall, an unchanged private valuation can cause private holdings to occupy a larger share of the reported portfolio. Calculate current weights using current reported values, then ask whether the private mark is timely enough for the decision. Restoring the target through an immediate private sale may be slow or costly. New contributions, public-asset purchases, reduced private commitments, and the use of incoming distributions can move the mix gradually. Keep enough liquid assets for spending, collateral, and future capital calls before directing every available unit to rebalancing. A secondary sale is another option, but compare its price with a defensible estimate of value and the urgency of the cash need. Rebalancing is a decision about economic exposure and liquidity, not simply a mechanical response to reported percentages.

Worked example: A portfolio starts with 60 million in public assets and 40 million in private assets, a 60%/40% split. Public assets fall to 40 million while private assets remain marked at 40 million. Reported total value is 80 million, so private weight is 40 ÷ 80 = 50%. With no change in the private mark, 20 million of new contributions directed to public assets would restore 60 million public and 40 million private, or 60%/40%. Check the private mark before acting.

Watch the distinction: A larger reported private weight does not require an immediate private sale. New cash and public purchases may rebalance more cheaply, especially when private marks lag market conditions.

Reference: [3]

35. Fragile allocation estimates

An optimizer treats expected returns, volatilities, and correlations as inputs, then seeks the best allocation under its stated objective. Small errors in those inputs can produce large changes in its preferred weights, particularly when two assets look similar or constraints are loose. A one-point change in a private equity return forecast is not automatically evidence for a major allocation shift; the estimate itself may be uncertain and its reported volatility may be smoothed. Test the proposed weight under lower returns, delayed exits, and adverse correlations. Apply practical limits based on liquidity, concentration, governance capacity, and future commitments. A range of plausible allocations often informs judgment better than one precise output. If an optimizer's answer changes sharply with a modest assumption change, treat that sensitivity as a finding to investigate before committing capital.

Worked example: A model recommends 15% private equity when its assumed return is 8%, then 45% when that assumption rises to 9%, with other inputs unchanged. The recommendation jumps 30 percentage points for a one-point forecast change. If the investor caps private equity at 25% because of cash needs and unfunded commitments, rerun the model within that limit and test returns below 8%. The original 45% is a sensitivity warning, not a mandate.

Watch the distinction: A mathematically optimal weight is conditional on estimated inputs. It should not override a liquidity limit merely because the software reports a precise percentage.

Reference: [3]

Level II - Risk and Risk Management

36. Value at risk and its boundary

Value at risk, or VaR, describes a loss cutoff for a specified horizon and confidence level under a stated model or historical sample. A one-day 95% VaR of 2 million means the cutoff is set so that modeled losses exceed 2 million in roughly 5% of one-day outcomes. It does not describe how large those worse losses become. Before comparing two VaR figures, check that they use the same portfolio, horizon, confidence level, valuation approach, and assumptions about market moves. Liquidity can make realized losses exceed a modeled mark when a position must be sold. A short historical sample may also miss regimes that matter for private assets or leveraged strategies. Use VaR alongside stress scenarios, concentration analysis, and a measure of losses beyond the cutoff. It is a summary of a distribution, not a guarantee about tomorrow.

Worked example: Assume a model gives a portfolio a one-day 95% VaR of 2 million. Across 100 equally likely modeled trading days, about five would be expected to show losses greater than 2 million. One of those losses could be 3 million or 15 million; the VaR figure alone does not say. If cash obligations force sales during those days, execution costs could add to the modeled loss. The stated horizon and model are essential to the interpretation.

Watch the distinction: VaR is not the largest possible loss. Its cutoff deliberately leaves a tail of worse outcomes, whose size requires separate analysis.

Reference: [1]

37. Losses beyond the cutoff

Expected shortfall summarizes the average loss within a specified worst fraction of outcomes. It addresses a question that VaR leaves open: how severe are losses after the cutoff has been crossed? With 100 equally likely scenarios and a 5% tail, sort losses from worst to best and average the five worst observations. State whether losses are positive magnitudes or signed returns before calculating; mixing conventions causes errors. The answer still depends on the scenario set and on how positions are valued under stress. Rare events absent from the sample cannot affect its calculated average, however damaging they might be. Also inspect the individual tail losses, because a single extreme result may dominate the mean. Expected shortfall is useful for comparing tail severity, but it should sit beside scenario analysis and plans for meeting cash demands during a loss.

Worked example: Suppose the five worst losses among 100 equally likely scenarios are 3, 4, 5, 6, and 12 million. Their sum is 30 million, so the average loss in the worst 5% is 30 ÷ 5 = 6 million. The 12 million scenario contributes one-third of that sum. The calculation shows severity within the defined tail; it does not predict the worst loss possible outside these 100 scenarios.

Watch the distinction: Using only the fifth-worst loss reports a cutoff, not the average tail loss. Expected shortfall includes the more severe outcomes beyond that boundary.

Reference: [1]

38. A liquidity stress sequence

Liquidity stress often begins with a price fall and then spreads through cash obligations. Lower collateral values can trigger a margin call; private funds may call committed capital at the same time; sales into a weak market can lock in losses or cause further collateral pressure. Build a dated cash schedule: starting usable cash, each payment's amount and deadline, reliable inflows, and assets that can actually be sold or pledged by then. Separate quoted value from cash realizable after spreads, settlement time, and any restrictions. If a gap remains, compare available borrowing with asset sales and the consequences of each. Test whether using a credit line today merely moves the problem to its maturity. The critical decision is whether the portfolio can meet obligations on time while preserving enough flexibility for another adverse move.

Worked example: A portfolio has 12 million in usable cash. A margin call of 7 million is due tomorrow, and a private-fund capital call of 8 million is due the same day. Total demand is 15 million, leaving a 3 million shortfall. The manager checks which public holdings can settle in time and what cash their sale would produce after costs. A private interest quoted at 20 million is no immediate solution if a negotiated sale takes weeks.

Watch the distinction: Adding up asset values does not establish liquidity. The relevant amount is cash available by each deadline after settlement, restrictions, and plausible sale costs.

Reference: [1]

39. Hidden common exposures

Portfolio risks can overlap even when holdings have different strategy labels. A buyout fund's companies may rely on refinancing, while a long-short fund may own credit-sensitive shares or finance positions with borrowing. Widening credit spreads can hurt both through different routes. Identify each holding's sensitivity to common drivers such as equity markets, interest rates, credit spreads, currencies, and funding conditions. Then estimate portfolio exposure by combining positions and their factor sensitivities, allowing for direction and scale. Do not assume an offset from a fund's short positions until their factor exposure is known. Check what happens when correlations rise and lenders tighten terms together. Report both the factor view and the underlying business or instrument paths, since a single statistic can hide nonlinear losses. Diversification comes from different economic responses, not different names on a holdings list.

Worked example: Suppose a portfolio has 30 million in a buyout fund and 20 million in a long-short fund. In a hypothetical credit-spread shock, estimated losses are 15% and 8%, respectively. The combined loss is 30 × 0.15 + 20 × 0.08 = 4.5 + 1.6 = 6.1 million. The funds' structures differ, yet both contribute to the same stress loss. The estimates should be checked for financing and valuation effects.

Watch the distinction: A hedge fund label does not guarantee protection against a buyout fund's risks. Look through both strategies to their shared economic drivers and financing needs.

Reference: [1]

Level II - Methods and Models

40. Alpha and beta in a return model

A simple factor model separates a return associated with market movement from a remainder: predicted return equals intercept plus beta times market return. Beta measures how strongly the asset's return tends to move with the chosen market factor in the model. The intercept is the model's fitted return when that factor is zero; an actual return minus the prediction is the period's residual. Do not call one favorable residual proof of skill. It may reflect omitted factors, random variation, stale prices, or a benchmark that does not fit the strategy. Check the return period, currency, fee basis, and factor definition before comparing numbers. With several relevant factors, a one-factor residual may shrink or change sign. Use the model to ask which exposures explain performance and what remains to investigate, rather than treating its labels as causal findings.

Worked example: Assume a model has a 2% intercept and market beta of 0.5 for the period. If the market returns 10%, predicted fund return is 2% + 0.5 × 10% = 7%. The fund actually returns 8%, leaving a residual of 8% − 7% = 1 percentage point. That point is unexplained by this one-factor prediction; it is not automatically manager skill. Another relevant factor or measurement error could account for it.

Watch the distinction: A positive one-period residual is not conclusive alpha. It measures what this particular model failed to explain and depends on its benchmark, factors, and data.

Reference: [1]

41. Stale marks and smooth returns

Private assets often lack frequent transactions, so reported values may depend on appraisals, comparable sales or models. An appraisal that incorporates new information slowly can spread one economic price change across several reporting periods. The resulting return series may look unusually steady: measured volatility falls, and measured correlation with publicly traded assets can fall because the private mark reacts later. That apparent diversification may disappear when both assets are valued at the same point in a downturn. Compare valuation dates, assumptions, comparable assets and later revisions before using reported returns in a risk model. The 2027 CAIA Exam Handbook identifies Methods and Models as a Level II topic; a smooth reported series still needs economic scrutiny.

Worked example: Listed property comparables fall 20% in a quarter, while a private property fund reports an unchanged appraisal. If both began at 100, the listed measure ends at 80 and the fund still reports 100. The 20-point gap does not establish a 20-point difference in underlying performance: lease terms, leverage and property quality may differ. Investigate those differences and the appraisal date before treating the flat mark as protection.

Watch the distinction: A zero reported return means zero economic risk. A delayed appraisal may record a loss in a later quarter, making contemporaneous volatility and correlation look artificially low.

Reference: [1]

42. Private-fund multiples and IRR

Divide distributions by paid-in capital to obtain DPI, the cash already returned per unit contributed. Divide residual value by paid-in capital to obtain RVPI, the value still held per unit contributed. Their sum is TVPI, the total reported value multiple. These measures separate realized proceeds from value that depends on an estimate and a future exit. Internal rate of return answers a different question: what annual discount rate equates dated contributions with dated distributions and residual value? An early distribution can lift IRR even when the total multiple is modest. Conversely, a high multiple earned over many years can have a lower annualized return. Check fund fees, valuation dates and remaining commitments when comparing funds. The related CIPM study guide provides further performance-measurement context.

Worked example: A fund has paid-in capital of 100 million, distributed 40 million and reports residual value of 90 million. DPI = 40/100 = 0.4; RVPI = 90/100 = 0.9; TVPI = (40 + 90)/100 = 1.3. Of the 130 million in reported total value, 90 million remains unrealized. No unique IRR follows because the dates of contributions, distributions and valuation are missing.

Watch the distinction: TVPI of 1.3 does not mean investors have received 130% of their capital in cash. Here they have received 40%; the rest is a reported holding value.

Reference: [1]

43. Scenario models need assumptions

A simulation generates many possible paths from specified return distributions, relationships among assets and rules for contributions, distributions and spending. Its shortfall frequency is conditional on those inputs, not an observed law of nature. For an alternative-investment portfolio, timing assumptions matter especially: a private fund may call capital while distributions from other funds slow. Correlations may also rise during stress, leaving fewer assets available for sale at acceptable prices. Read a reported probability by asking what creates cash, what consumes it and which events the model permits to occur together. Then vary assumptions that could change the decision, such as delayed exits, lower sale proceeds or simultaneous margin calls. A useful result explains how robust the plan is across plausible adverse paths.

Worked example: A simulation shows cash shortfalls in 50 of 1,000 paths, or 5%, while holding private-fund distributions steady during recessions. If recessions delay exits as well as reduce public-asset values, the portfolio could have less cash precisely when it needs more. The revised shortfall rate cannot be calculated from the 5% alone. Rerun the paths with linked distribution delays and inspect the new frequency and shortfall size.

Watch the distinction: A modeled 5% shortfall probability is not a guarantee that cash will suffice 95% of the time. Changing the cash-flow assumptions can materially change the estimate.

Reference: [1]

Level II - Accessing Alternative Investments

44. Direct deals or pooled funds

A direct investment gives the owner greater influence over asset selection, financing and exit, but concentrates exposure and demands staff able to source, assess and oversee the asset. A pooled fund delegates those tasks to a manager and can spread capital across several holdings. Its costs include management fees, potential performance fees and less control over individual decisions. Choose an access route by first defining the desired exposure and maximum tolerable single-asset loss. Next assess internal expertise, decision speed, governance capacity and the cash needed for follow-on investment. Finally compare total fees with the value of diversification and delegated execution. A fund's asset count is only a starting point: holdings exposed to the same demand, regulation or financing conditions can still lose value together.

Worked example: An owner has 10 million to invest in infrastructure. One direct toll-road deal would place the full 10 million in a single asset requiring active oversight. A fund could, hypothetically, allocate the owner's share across five projects at 2 million each. If the owner lacks deal staff and cannot bear one project's failure, choose the fund, then check whether its projects share the same traffic or financing risks.

Watch the distinction: Direct ownership is not automatically cheaper after accounting for diligence, legal work and ongoing oversight. Nor does a pooled fund guarantee diversification merely because it holds several assets.

Reference: [1]

45. Buying a fund interest later

A secondary buyer acquires an existing investor's fund interest, often with exposure to assets already selected. Start with the purchase price relative to the latest reported net asset value, but treat that NAV as an estimate dated to a specific reporting point. Then review changes since that date, the portfolio's likely exits and the terms governing the transfer. Remaining unfunded commitments are a separate cash obligation: a discount on today's NAV does not make future capital calls disappear. Compare the price paid plus potential calls with expected distributions, timing and risk, rather than calling every discounted purchase a bargain. A large discount may reflect weak assets, stale valuations, a seller's need for liquidity or several of these at once. Transfer restrictions and manager consent can also affect execution.

Worked example: A fund interest has reported NAV of 100 million, costs 85 million and carries 20 million of uncalled commitments. The headline discount is 100 - 85 = 15 million, or 15% of reported NAV. The buyer pays 85 million now and must plan for up to another 20 million if all remaining capital is called. Total possible cash outlay is 105 million; future calls should be evaluated with the assets they finance.

Watch the distinction: A 15% discount to NAV is not a locked-in 15% gain. NAV can change, distributions take time, and the buyer inherits the obligation to meet future calls.

Reference: [1]

46. The co-investment tradeoff

A co-investment is a separate position alongside a fund's investment in a particular company or asset. It may have lower incremental fees and offer more choice over the exposure, but it adds concentration where the investor may already participate through the fund. Measure that overlap before comparing fee schedules. Also check who controls the deal, whether the co-investor receives the same economic terms, and how follow-on funding or an exit decision is made. Co-investment invitations can arrive on a short timetable, so the owner needs a repeatable process for valuation, legal review and risk limits. A fee saving has value only after expected returns are adjusted for extra risk and the cost of making the decision. The fund's own allocation is the first input to that calculation.

Worked example: Assume a 20 million fund commitment is fully invested and 10% of the investor's fund exposure sits in Company A. That is 2 million of indirect exposure. A separate 5 million co-investment in A raises the combined position to 7 million. The new investment may save fees on 5 million, but it more than triples this investor's exposure to A, from 2 to 7 million.

Watch the distinction: Calling the co-investment a second vehicle does not create a second independent risk. If it owns the same company as the fund, the two positions overlap.

Reference: [1]

Level II - Due Diligence and Selecting Managers

47. Strategy fit before performance

Define the job of a manager before judging its return record. An owner seeking equity-crash protection needs evidence about behavior during equity losses, the source of any hedge and the cost of carrying it. A high average return, by itself, answers none of those questions. Examine holdings and factor exposures, then compare the manager with a benchmark that represents the proposed role. Ask whether reported gains came from equity beta, credit risk, leverage or a distinct process likely to behave differently when needed. Stress-period observations are informative but sparse, so combine them with position-level analysis and scenario tests. The Level II study guide can sit alongside this role-first approach: articulate the mandate, test fit, then assess performance and terms.

Worked example: An owner seeks protection during equity crashes. A hedge fund returned 12% annually in a favorable sample, but in one equity selloff the index lost 25% and the fund lost 15%. Its loss was 60% of the index loss in magnitude, since 15/25 = 0.60. This observation does not prove future behavior, but it contradicts a claim of reliable crash protection. Investigate its equity exposure before appointing it.

Watch the distinction: A strong absolute return does not establish that a fund fills a defensive portfolio role. A fund can earn well in rising markets while adding losses in downturns.

Reference: [1]

48. Operational controls matter

Investment diligence asks whether a strategy can make money; operational diligence asks whether assets, transactions and reported results can be trusted. Trace who can authorize trades and cash movements, who holds the assets, who independently reconciles positions and who calculates NAV. These tasks should have checks that prevent one person or firm from both creating and verifying a record. Review service-provider agreements and obtain evidence directly where possible, rather than relying only on a manager's description. For hard-to-price holdings, examine valuation policies, the source of model inputs, exception handling and oversight of overrides. Investigate breaks between manager, administrator, broker and custodian records. Even a persuasive strategy can fail investors through misappropriation, incorrect pricing or an inability to meet redemptions. Controls should be tested as processes, not merely named on an organization chart.

Worked example: A manager calculates fund NAV internally and can move cash from the same account holding fund assets. The investor should verify that a separate custodian confirms holdings and movements, and that an independent administrator reconciles records and calculates or checks NAV. If one party can alter positions, cash and valuations without effective review, a strong performance history does not resolve the operational risk.

Watch the distinction: An external auditor's name alone does not show that daily cash controls or each interim NAV is independently checked. Identify what each service provider actually verifies.

Reference: [1]

49. A track record in context

Evaluate a manager against the risks actually taken, over the same dates and after relevant fees. Begin with a benchmark representing the strategy's investable opportunity, then inspect leverage, market sensitivity, sector tilts and liquidity. A return above an unleveraged index can still be poor for a portfolio carrying much more market risk. A simple factor model estimates the return attributable to broad exposures; the residual is a clue for further investigation, not automatic proof of skill. Check whether exposures were stable, whether the sample spans adverse conditions and whether results include all relevant accounts rather than selected successes. Distinguish realized gains from model-based marks. Finally, ask whether the process and team that produced the history are still in place. Past performance supports an assessment only when its source is understood.

Worked example: A fund earns 12% while a relevant index earns 10%. Its 2-percentage-point lead looks attractive until you learn it maintained twice the index's market exposure. Under a deliberately simple model with zero intercept and a constant beta of 2, the market-linked return would be 2 × 10% = 20%. The fund's 12% is 8 points below that model value, before asking whether the model fits.

Watch the distinction: Beating an index by two points is not necessarily positive alpha. The comparison must account for the amount and type of risk used to earn the return.

Reference: [1]

50. Terms that shape incentives

Fund documents determine what investors pay, when they can get cash and what happens if the people responsible for the strategy leave. Read management fees, incentive fees, expenses and fee offsets together: each affects the investor's net return and the manager's incentives. Read redemption frequency alongside notice periods, gates, suspensions and asset liquidity; a stated dealing date may not ensure payment on that date. A key-person clause may pause new investments, trigger a vote or provide another specified remedy, depending on its wording. It does not automatically create a right to redeem. During diligence, map each important event to the relevant clause and identify who can exercise a right, by when and with what vote. This turns contract language into an actionable response plan.

Worked example: Suppose a hypothetical fund agreement permits temporary redemption suspensions and says that a portfolio manager's departure pauses new investments until investors vote. The manager leaves just as the fund suspends redemptions. Investors can use the key-person voting process concerning new deals, but cannot assume that vote releases existing capital. They must read the separate redemption and suspension clauses to identify any exit right and its timing.

Watch the distinction: A key-person provision does not necessarily let investors withdraw immediately. Its remedy may govern new investments or governance votes, while redemption rights sit in different clauses.

Reference: [1]

Level II - Volatility and Complex Strategies

51. Paying for convex protection

A purchased put gives its holder the right to sell an asset at a stated strike by expiry. At expiry, its payoff per unit is the greater of strike minus asset price and zero. Subtract the premium to find net profit. As the asset falls further below the strike, the put's payoff grows, while the buyer's maximum loss on the option itself is the premium paid. This shape can cushion a portfolio against a sharp fall, though strike, maturity and position size determine how much protection it provides. Protection has a recurring cost: if the feared fall does not happen, each expired put can consume its premium. Compare that carrying cost with the portfolio losses the put is intended to offset, and include any change in option value before expiry when assessing an early exit.

Worked example: A one-unit put has a strike of 100 and costs 4. If the asset expires at 80, payoff = max(100 - 80, 0) = 20, and net profit = 20 - 4 = 16 units before other costs. If the asset finishes above 100, payoff is zero and the buyer loses the 4-unit premium. Buying the put repeatedly would incur another premium each time.

Watch the distinction: A put that protects against a large fall is not free insurance. Small or absent declines can leave the buyer with repeated premium losses.

Reference: [1]

52. Short volatility and tail loss

Selling a put brings in a premium immediately but creates an obligation if the underlying asset falls below the strike. At expiry, the seller's net result per unit is premium minus the greater of strike minus asset price and zero. Many quiet periods can therefore produce modest gains, while one severe decline creates a much larger loss. The distribution is asymmetric; a high fraction of profitable trades says little about the damage in the losing tail. Evaluate the position using plausible stress prices, collateral needs and whether losses on several positions could occur together. Option premiums may compensate for bearing risk, but a strategy's realized premium history cannot establish that the risk has vanished. CAIA practice questions can help reinforce the payoff calculation.

Worked example: A trader sells a one-unit put with strike 100 for a premium of 3. At expiry price 70, the buyer exercises for a payoff of 100 - 70 = 30, so the seller's net loss is 30 - 3 = 27 units. Nine earlier puts that each expired worthless earned 9 × 3 = 27 units. This single loss cancels those nine premiums before costs.

Watch the distinction: Nine winning trades followed by one losing trade do not imply a profitable strategy. The size of the tail loss matters as much as the win rate.

Reference: [1]

53. Convertible arbitrage exposures

A convertible bond combines a bond claim with an option to exchange it for the issuer's shares under stated terms. A manager may buy the convertible and short the shares to reduce sensitivity to small equity-price moves. The share hedge is imperfect and usually changes as the stock moves because the convertible's equity sensitivity changes. Other risks remain: a wider credit spread can lower the bond's value, falling option value can hurt the position, and borrowing shares or financing the bond can become costly or unavailable. Analyze each exposure separately before treating the trade as neutral. Estimate how much of a bond-value change comes from equity, credit and volatility, then test the cost of maintaining the hedge during stress. The short stock is a hedge for one component, not for the entire instrument.

Worked example: Over a short interval, suppose a convertible loses 4 units from its equity component and another 12 from a credit-spread widening. A stock short sized for that initial equity move gains 4. Combined change is -4 - 12 + 4 = -12 units before financing and borrow costs. The equity move was offset, but the issuer's deteriorating credit still caused a loss.

Watch the distinction: Shorting the issuer's shares does not make a convertible position riskless. Credit deterioration, changing equity sensitivity, option value and share-borrow costs can all affect returns.

Reference: [1]

Level II - Universal Investment Considerations

54. Returns after tax

Investors consume or reinvest cash after taxes, so a strategy's value depends on the investor's own tax treatment and the timing of cash flows. For a single taxable gain, subtract the tax on that gain from proceeds before computing return on initial capital. More complex strategies may distribute income, realize gains at different times or retain unrealized appreciation; those patterns can change the after-tax comparison even when pretax returns match. Work from the investor's applicable assumptions rather than applying one rate to every cash flow. Then distinguish tax deferral from tax elimination: a tax paid later may improve reinvestment capacity, but it may still be owed. Include fees and any taxes triggered by portfolio turnover when the problem provides them. Without a stated jurisdiction and taxpayer facts, a numerical tax rate belongs to the hypothetical problem only.

Worked example: Assume an investor puts in 100 currency units, receives the original 100 plus a 10-unit taxable gain after one year, and owes tax equal to 30% of that gain. Tax is 0.30 × 10 = 3 units. Net proceeds are 110 - 3 = 107, so after-tax gain is 7 and after-tax return is 7/100 = 7%. Fees and other taxes are excluded.

Watch the distinction: Multiplying the whole 110-unit proceeds by 70% taxes the returned principal in this example. The stated 30% rate applies only to the 10-unit gain.

Reference: [1]

55. Sustainability claims and evidence

A sustainability label is a claim about investment decisions that should be tested against holdings and behavior. Begin with the fund's stated objective: exclusion, lower portfolio emissions, engagement with high-emission companies, or another measurable goal. Each calls for different evidence. Check selection criteria, baseline data, coverage of material holdings, voting or engagement records and how results are reported over time. A high-emission asset need not invalidate a transition strategy if the manager has a credible reason to own it and a monitored plan for change. Equally, calling an asset transitional does not establish progress. Compare promised actions with actual capital spending, operating changes and outcomes, while allowing for data limitations and changes in portfolio composition. The useful question is whether the manager's choices follow its stated method.

Worked example: A climate-focused fund owns a high-emission industrial asset but has never reviewed its emissions, upgrade needs or engagement plan. The holding might fit a transition mandate, yet the current evidence cannot show that. Ask for a baseline, specific actions, decision milestones and subsequent results. If the manager cannot connect the holding to its stated criteria, treat the claim as unsubstantiated rather than assuming the label describes the portfolio.

Watch the distinction: One high-emission holding does not automatically disprove a transition strategy, and a climate label does not validate it. Judge the holding against the fund's stated process and evidence.

Reference: [1]

56. Currency changes the result

A foreign asset gives a home-currency investor two linked returns: the asset's return in its local currency and the change in that currency's value against the home currency. Multiply their gross returns, then subtract one: home return = (1 + local asset return)(1 + foreign-currency return) - 1. Simply adding the two percentages omits the interaction term. Define the exchange-rate direction before calculating; a foreign-currency loss means each foreign unit buys fewer home units. A currency hedge can reduce that exposure, but its cost and effectiveness belong in the comparison. Currency can also affect operating cash flows and asset prices, so translating a reported return is only the first step in a full risk analysis. Use the investor's actual spending currency to judge the economic result.

Worked example: A foreign asset rises 10% in its local currency while that currency loses 5% against the investor's home currency. Starting with 100 home-currency units and assuming an initial exchange rate of one to one, the asset becomes 110 foreign units. Each foreign unit then buys 0.95 home units, giving 110 × 0.95 = 104.5. Home-currency return is 4.5%, before fees and taxes.

Watch the distinction: Adding 10% and negative 5% gives 5%, which misses the currency effect on the larger ending foreign value. Multiply gross returns to obtain 4.5%.

Reference: [1]

Level II - Emerging Topics

57. Testing alternative data

A novel data set can appear predictive because it captures information early, or because its history contains bias that will not persist. First establish the right to obtain and use the data, including restrictions on collection, sharing and material nonpublic information. Then identify what each observation measures, when it became available and which entities are missing. A backtest that uses data revised later, or selects only firms with surviving records, can overstate performance. Test the signal on periods and entities held out from discovery, account for transaction costs and examine whether coverage changes in stress. Compare the signal with simpler public information to see whether it adds distinct value. Funding a strategy requires an executable decision rule and reliable data delivery, not merely an attractive historical chart.

Worked example: A vendor's store-traffic series covers 100 of 200 locations, and the covered stores predicted past sales well. Coverage is only 100/200 = 50%. If missing stores are smaller or less successful, the historical relationship may not represent the full chain. Check why records are absent, use only data available at each forecast date, and test later periods before committing capital.

Watch the distinction: A successful backtest on the stores with records does not validate forecasts for stores without records. Missing observations can create coverage bias, while later revisions can create look-ahead bias.

Reference: [1]

58. Tokenized fund claims

A token can serve as a digital record or transfer mechanism for an interest in a fund, but the technology alone does not define the holder's legal claim. Examine the governing fund documents and ownership register to establish what rights transfer with the token, who may hold it and whether the transfer is recognized by the fund. A market in tokens can offer an exit to an individual holder if another buyer appears, yet the fund's underlying assets still take time to sell and value. A quoted token price may diverge from reported NAV, especially when buyers demand compensation for uncertainty or limited redemption rights. Separate three questions: can the record transfer, can the holder enforce the associated rights, and can the underlying portfolio produce cash? Each has a different source of risk.

Worked example: A token represents an interest in a private property fund and changes hands daily. The buildings are still sold through negotiated transactions, so daily token trading does not make them liquid. Suppose a token with reported NAV of 10 units trades for 9: the buyer pays 10% below reported NAV, since (10 - 9)/10 = 10%. That discount may reflect uncertainty and does not guarantee prompt fund redemption.

Watch the distinction: Fast token transfers do not make property sales fast. The token's trading liquidity, the fund's redemption rights and the buildings' sale liquidity are separate matters.

Reference: [1]

59. Transition risk in real assets

Changes in technology, customer preferences, energy costs or operating requirements can alter both a real asset's income and the capital needed to preserve it. Current rent or operating earnings should therefore be projected alongside maintenance and upgrade spending. Identify the trigger: will tenants leave, pay less or demand improvements? Estimate when spending is required, which party bears it and whether higher future income or avoided vacancy repays it. Discount incremental cash flows at a rate suited to their risk; an undiscounted sum can make a distant benefit look too attractive. Also test what happens if the upgrade is delayed or fails to retain users. An asset can show strong current income yet have lower value once necessary investment and lost operating time are recognized.

Worked example: A building earns 10 million annually but needs a 20 million upgrade to retain tenants. Suppose upgrading preserves 10 million of annual income for two additional years, while without it the building would earn nothing in those years. The 20 million of extra income merely equals the 20 million cost before discounting, taxes and disruption. On those assumptions alone, the upgrade does not create positive present value; a longer benefit is needed.

Watch the distinction: Current income of 10 million is not necessarily durable free cash flow. A required 20 million upgrade can absorb future proceeds even when today's occupancy is strong.

Reference: [1]

60. New access, old liquidity mismatch

A fund may offer periodic redemption opportunities while owning assets that cannot be sold on the same timetable. Assess its cash buffer against plausible requests, capital calls and operating needs, then read the terms governing notice, gates, suspensions and borrowing. A cash reserve covers withdrawals only until it is used; repeated requests may require asset sales at unfavorable prices. Selling loans quickly can crystallize a discount that affects investors who remain. Borrowing bridges timing only if capacity exists and repayment is credible. Model both the first redemption period and the fund's position afterward. A promise to process requests is meaningful only within the documented limits and the liquidity of the portfolio. Distinguish an investor's ability to submit a request from the fund's ability to pay it promptly.

Worked example: A semi-liquid fund reports 80 million in private loans and 20 million in cash. Redemption requests total 30 million, so cash falls short by 30 - 20 = 10 million before other needs. If loans can be sold for 90% of reported value, the fund must sell about 11.11 million of reported loan value to raise 10 million, crystallizing about 1.11 million of loss. Check whether its terms instead limit or delay payments.

Watch the distinction: Accepting daily or periodic redemption requests does not mean the fund holds enough cash to pay them all. Payment depends on reserves, sale proceeds, borrowing and contractual limits.

Reference: [1]

Official references and further reading

All study guides

Choose your exam version

CAIA Level I

CAIA Level II

Level I - CAIA Ethical Principles

1. Whose interests govern?

2. Conflicts in a fund recommendation

3. Stewardship after investment

4. Acting when a rule is silent

Level I - Introduction to Alternative Investments

5. What makes an investment alternative?

6. The price of illiquidity

7. Cash timing changes returns

8. Evergreen and drawdown funds

9. Diversification beyond labels

Level I - Real Assets

10. Property income and value

11. Commodity futures returns

12. Contracted real-asset cash flows

Level I - Private Equity

13. Stages of private equity

14. Leverage in a buyout

15. Order of fund distributions

Level I - Private Debt

16. Direct-lending capacity

17. Where private lenders rank

18. Tranches absorb loss in order

Level I - Hedge Funds

19. Long-short exposure

20. Market neutral is a risk claim

21. High-water-mark fees

Level I - Digital Assets

22. Ownership, keys and custody

23. Tokens and smart contracts

24. Sizing a crypto allocation

Level II - CAIA Ethical Principles

25. An ethical decision in writing

26. Independence under manager pressure

27. Stewardship with competing interests

Level II - Institutional Asset Owners

28. Endowment spending across generations

29. Pension assets and liabilities

30. Purpose shapes sovereign funds

31. The family balance sheet

Level II - Asset Allocation

32. Objectives before allocation

33. Covariance drives portfolio risk

34. Rebalancing with illiquid assets

35. Fragile allocation estimates

Level II - Risk and Risk Management

36. Value at risk and its boundary

37. Losses beyond the cutoff

38. A liquidity stress sequence

39. Hidden common exposures

Level II - Methods and Models

40. Alpha and beta in a return model

41. Stale marks and smooth returns

42. Private-fund multiples and IRR

43. Scenario models need assumptions

Level II - Accessing Alternative Investments

44. Direct deals or pooled funds

45. Buying a fund interest later

46. The co-investment tradeoff

Level II - Due Diligence and Selecting Managers

47. Strategy fit before performance

48. Operational controls matter

49. A track record in context

50. Terms that shape incentives

Level II - Volatility and Complex Strategies

51. Paying for convex protection

52. Short volatility and tail loss

53. Convertible arbitrage exposures

Level II - Universal Investment Considerations

54. Returns after tax

55. Sustainability claims and evidence

56. Currency changes the result

Level II - Emerging Topics

57. Testing alternative data

58. Tokenized fund claims

59. Transition risk in real assets

60. New access, old liquidity mismatch