Alternative investments are assets or strategies outside public stocks, bonds, and cash. The label includes private equity, private credit, hedge funds, real assets, insurance-linked securities, royalties, commodities, and digital assets. It says where an investment sits on the shelf—not whether it will diversify a portfolio.
Diversification is a property of the whole portfolio. An alternative earns that role when its return comes from a genuinely different cause, its risk is measured honestly, and its vehicle remains usable under stress. Novelty, illiquidity, and a quarterly mark prove none of those things.
what alternative investments actually are
The category is defined mostly by what it is not. CFA Institute groups investments outside public equity, fixed income, and cash under alternatives, while noting their longer holding periods, reduced liquidity, specialized underwriting, and more complex structures.
That broad definition is useful for filing products. It is too broad for allocating capital. A buyout fund, a gold bar, a catastrophe bond, and a music royalty do not share a return engine merely because they occupy the same consultant deck.
The practical question is not, “Is this an alternative?” It is, “What will make this position pay when the rest of my portfolio does not?”
five tests for a diversifier
Run these tests in order. A weak answer does not automatically make an investment bad. It tells you which portfolio job the investment can—and cannot—honestly perform.
1. name what actually generates the return
Start with the causal driver: the event or activity that creates the cash flow or gain.
“Private markets,” “scarcity,” and “manager skill” are categories or assertions, not causes. A private equity return may come from operating growth, leverage, multiple expansion, and an eventual sale. A royalty may come from prescriptions filled or songs streamed. A catastrophe bond earns income for bearing a defined peril and can lose principal if its trigger is met.
Write the driver as a complete sentence: This position pays because ___ happens. Then ask whether that event is truly different from the forces already deciding the rest of the book. If the blank ultimately says GDP growth, cheap refinancing, rising exit multiples, or investor risk appetite, the engine may be familiar even when the wrapper is not.
2. ask whether the observed correlation is marked, lagged, or appraisal-smoothed
Correlation calculated from reported returns is only as honest as the marks behind it.
Daily market prices are noisy, but they reveal changing information. Private valuations may update quarterly, rely on appraisals, or move only when a financing or sale forces recognition. That can make reported volatility and correlation look lower without changing the underlying economics.
Cliff Asness calls this volatility laundering: non-marking and illiquidity can turn economically volatile assets into unrealistically smooth reported return series. His objection is not that every private investment is poor. It is that not marking an asset does not make it low risk or low correlation.
Ask for the valuation policy, frequency, model inputs, stale-price treatment, and performance during actual realizations. Compare the asset with relevant public-market drivers on a lag, not only with same-month reported net asset value.
Also treat the correlation itself as non-stationary. A quiet decade does not bind the next crisis. In a forced-seller quarter, pairwise correlations often move toward one — the same assets that looked independent on a full-period statistic fail together. If the diversification case depends on a historical correlation holding in the tail, say so, and say what happens if it does not.
3. name who pays and why that payment survives stress
Income needs a payer. Appreciation needs a future buyer willing to pay more. Name that party before discussing yield.
For a loan, the payer is the borrower. For a royalty, it may be a licensee or distributor passing through end-customer payments. For an insurance-linked security, an insurer or reinsurer pays premium to transfer a defined risk, often through a collateralized structure. Each case still requires questions about credit, concentration, legal enforceability, termination rights, and the reason the obligation remains valuable in a downturn.
A contract is stronger than a story, but a contract is not cash. Test the payer’s incentive and capacity under the exact stress in which you expect diversification. If payment depends on discretionary budgets, refinancing, one platform, one customer, or a healthy exit market, say so.
4. test liquidity, leverage, and the wrapper under forced selling
The asset and the vehicle are separate underwriting decisions.
A sound cash flow can sit inside a fund with mismatched redemption terms, leverage, capital calls, gates, concentrated counterparties, or fees that absorb the premium. A liquid strategy can still be forced to cut risk after margin moves. An illiquid strategy can avoid panic selling while leaving an investor unable to meet obligations elsewhere.
Map the full chain: underlying liquidity, fund liquidity, financing terms, collateral calls, redemption rights, transfer restrictions, valuation authority, and the likely buyer in a sale. Then model what happens when several investors seek cash at once. If the answer relies on the manager having time, the market staying open, or lenders extending terms, the wrapper has reintroduced the common risk.
5. state which regime breaks the thesis
Every diversifier is conditional. Name the condition that makes it stop working.
Credit strategies can break under defaults and refinancing stress. Trend strategies can suffer in rapid reversals and directionless markets. Commodities can disappoint in disinflation or demand destruction. Royalties can fail through obsolescence, concentration, litigation, or simple overpayment. Event-risk securities can lose exactly when the specified event occurs.
Separate a diversifier from a hedge here. A diversifier may improve the portfolio across a chosen horizon without making money in every drawdown. A hedge should have a named exposure, trigger, horizon, and expected payoff when that exposure loses. “Alternative” is neither. And because correlations are non-stationary, a thesis that only works if the next crisis looks like the last sample is already broken — write the tail-correlation failure in this box, not as a footnote.
how familiar alternative investments score
The scores below judge the diversification claim, not expected return or investment quality. “Strong” means the answer is reasonably observable and distinct. “Mixed” means structure and underwriting decide. “Weak” means the common alternative-investment pitch does not clear the test. “Only with” names the missing condition.
| candidate | 1. return driver | 2. marks | 3. payer | 4. forced selling | 5. breaking regime | diversification read |
|---|---|---|---|---|---|---|
| private equity | mixed: operating results, leverage, and exit multiples often retain equity-market exposure | weak: model and appraisal marks can lag public markets | mixed: customers fund operations; a future buyer often realizes the gain | weak: lockups prevent daily selling, but capital calls, borrowing against fund portfolios, and secondary discounts still matter | recession, expensive debt, or a closed exit market | only with genuine operating improvement or manager alpha; do not call the category uncorrelated |
| private credit | mixed: borrower cash flow plus spread, still tied to the credit cycle | weak: model marks, amendments, and interest paid with more debt rather than cash can delay recognition | mixed: contractual borrower payments survive only if the business does | weak: illiquid loans become fragile inside leveraged or redemption-bearing vehicles | defaults, refinancing pressure, covenant erosion, or recovery shortfalls | Primarily an income and illiquidity sleeve, not an independent engine |
| managed futures | strong: systematic long and short exposure to trends across markets | strong: exchange-traded positions are marked frequently | mixed: gains come from price moves rather than a contracted payer | mixed: liquid markets help, while leverage, margin, and crowded exits can force de-risking | whipsaw, rapid reversal, or prolonged range-bound markets | A credible diversifier when mandate, leverage, collateral, fees, and implementation hold |
| catastrophe bonds and insurance-linked securities (ILS) | strong: a specified insured event, outside corporate earnings | mixed: tradable marks exist, but models, event estimates, and extensions introduce uncertainty | strong: sponsor premium and dedicated collateral make the payment chain legible | mixed: specialist liquidity and fund terms still matter after an event | the covered peril, model error, trigger mismatch, event clustering, or trapped collateral | A genuinely different engine for a named peril, not a low-risk bond substitute |
| royalties | mixed: prescriptions, streams, production, or other licensed activity can differ from capital-market cycles | mixed: cash receipts may be observable while the royalty asset remains modeled and illiquid | strong only with: enforceable rights, clean collection data, and diversified counterparties | weak: thin resale markets and fund leverage can dominate the stream | demand decay, platform change, legal challenge, concentration, or overpayment | Potentially useful when the contract and consumption driver are stronger than the acquisition story |
| gold and broad commodities | mixed: scarcity, inventories, supply, demand, and monetary regimes | strong: public markets provide frequent marks | weak: gold has no payer; futures returns also depend on curve and collateral mechanics | strong only with: controlled leverage, roll, custody, and vehicle risk | disinflation, rising real yields, demand destruction, or an adverse futures curve | A regime diversifier, not a contracted paycheck or universal crisis hedge |
Three conclusions follow.
First, alternative investments do not share a common diversification benefit. The category contains corporate risk, credit risk, event risk, consumption streams, systematic trading, and non-income stores of value.
Second, a smooth return series can be the weakest evidence in the file. Honest daily volatility may be preferable to a calm NAV that discovers its price only when capital must move.
Third, different does not mean safe. Catastrophe risk can be genuinely distinct and still produce a large loss. Managed futures can diversify and still whipsaw. The purpose of the five tests is correct role assignment, not a higher score at any price.
score nature with the same card
Nature gets no exemption from the framework. It also needs to be split into two different positions: working nature, which sells or leases productive output, and pure protection, which preserves standing ecological function.
| candidate | 1. return driver | 2. marks | 3. payer | 4. forced selling | 5. breaking regime | diversification read |
|---|---|---|---|---|---|---|
| working nature: farmland and timberland | mixed: biological growth, leases, harvests, and commodity sales provide a partly distinct engine alongside operating and market risk | weak to mixed: appraisal-based valuations can smooth and lag economic changes | mixed: tenants and commodity buyers are identifiable, but terms and credit vary | weak: the land is illiquid; leverage, fund terms, and sale timing can dominate | drought, fire, disease, input inflation, commodity weakness, high rates, or forced sale | A credible real-asset diversifier in some structures; not automatically uncorrelated and not a crisis hedge |
| pure protection: intact or restored ecosystems | only with a contract: flood attenuation, water quality, heat reduction, habitat, and other standing functions are candidate drivers, but no return exists until payment is tied to them | weak today: condition can be measured, but that does not establish return correlation or a track record | only with: a contracted, underwritten beneficiary such as a utility, municipality, company, insurer, foundation, or pooled balance sheet | only with honest duration: the work is long-lived and illiquid; a claim should not promise daily exit against that reality | ecological damage, payer failure, measurement mismatch, legal weakness, or a shock that harms both the ecosystem and the wider portfolio | Weak as a financial diversifier without a payer and evidence; potentially strong as a causal risk reducer that lowers the underlying loss |
That last distinction is the point. Working nature has return histories because farms and forests sell output. Pure protection does not inherit those histories merely because both involve land.
Pure protection may reduce a shared physical dependency beneath a portfolio—for example, by lowering flood exposure or supporting water reliability. That is valuable, but risk reduction is not the same as a positive return during a market drawdown. An ecological shock can damage protected systems and conventional holdings at the same time.
ensurance is working on the missing financial plumbing. RealValue is the protocol's natural capital accounting engine — ecosystem service value per acre per year measured against ecological condition. A certificate is a recorded claim on an agent's protection funding, not an insurance policy and not a payout promise. Proceeds are the value routed to that agent's onchain wallet to fund the work. None of those mechanisms establishes market sensitivity, proves low correlation, or substitutes for a track record. The diversification case begins only when the payer, term, rights, measurement, loss allocation, valuation, and exit are underwritable.
For a foundation, keep the two tools apart. A program-related investment (PRI) is a charitable-purpose tool that can accept a below-market return. A mission-related investment (MRI) seeks a market-rate return while advancing the mission. Either can widen the set of viable payers; neither is evidence of low correlation.
the decision rule
Before approving an alternative, require a one-page card with five complete answers:
- The return is generated by ___.
- The reported correlation depends on marks that are ___.
- The payer is ___, and payment survives stress because ___.
- In forced selling, the asset and wrapper behave as follows: ___.
- The thesis breaks when ___.
If the manager cannot fill those blanks plainly, the committee is being asked to buy a label. If the answers are clear, the investment may deserve a role—even when that role is income, inflation sensitivity, or risk reduction rather than “uncorrelated.”
Use the five allocator hunts to name the portfolio desire. Read is nature an asset class yet? for the working-nature versus pure-protection split, or the fine print of the free lunch for the broader diversification argument. Then browse markets to apply the card to actual instruments.
If you are defining a portfolio role rather than shopping for a label, talk through the underwriting questions. When the card is filled and you want the holdable claim: the position that funds the thing that shrinks the loss.
This framework is educational, not an offer, recommendation, or assurance that diversification will prevent loss. The governing documents, legal rights, tax treatment, valuation policy, and actual portfolio exposures of any vehicle control.