EDB Capital Research · Portfolio Construction

The EDB Capital Guide to Building an Uncorrelated Private Investment Portfolio

A practical framework for combining multiple modest return engines whose underlying sources of repayment—and primary causes of loss—are meaningfully different.

The premise: diversification by cause, not label

Traditional diversification starts with labels: stocks, bonds, real estate, private credit, private equity. The weakness is that several labels can still depend on the same economic conditions. A recession can hit equities, leveraged private companies, commercial real estate, equipment borrowers, and ordinary middle-market credit at the same time. Rising rates can pressure both long-duration bonds and highly leveraged businesses.

EDB Capital working definition

Two investments are meaningfully diversified when the real-world event that causes one to stop paying is substantially different from the event that causes the other to stop paying.

That changes the portfolio-construction question. Instead of asking whether two assets have historically reported low correlation, ask: what actually produces the cash flow, what can interrupt it, and could the same event impair several positions simultaneously?

The goal is not zero risk. The goal is to assemble several return engines where each exposure is small enough, understandable enough, and structurally protected enough that no single economic outcome determines the portfolio.

Four principles before selecting any investment

Protect principal firstYield is considered after collateral, seniority, payment control, insurance, reserves, and recovery rights.
Prefer repeatable cash flowContractual payments and short duration are generally easier to underwrite than dependence on future valuation.
Limit shared dependenciesManager, servicer, bank, insurer, borrower, geography, leverage, and liquidity can create hidden portfolio correlation.

A portfolio with ten strategies can still be concentrated if all ten rely on refinancing, the same bank, the same manager, or the same economic cycle. True diversification requires separate limits for the return engine and for the infrastructure supporting that engine.

Ten possible independent return engines

The table below is an analytical map—not a recommended portfolio. The return ranges are deliberately broad and represent illustrative gross underwriting targets, not current quotations, forecasts, or guaranteed returns. Actual opportunities vary materially by structure, duration, credit quality, fees, leverage, and market conditions.

Return engineWhat produces the returnPrimary permanent-loss riskTypical liquidityIllustrative gross target
Short-term U.S. TreasuriesGovernment interest and principal paymentsSovereign / inflation / reinvestment riskHigh3.5%–4.5%
Essential equipment financeBorrower lease or loan payments plus collateral recoveryBorrower default, poor collateral recovery, fraudMedium6.5%–9.0%
Trade finance / verified receivablesPayment of short-dated invoices by account debtorsObligor default, invoice fraud, dilution / disputeShort6.0%–9.0%
Commercial insurance-premium financeInstallment payments plus potential recovery of unearned premiumServicing, legal mechanics, carrier / insured issuesShort6.5%–9.5%
First-lien farmland lendingBorrower interest and principal backed by productive landAgricultural cash-flow stress, land-value decline, water / environmental riskMedium-long6.0%–8.5%
Higher-attachment catastrophe bondsInsurance premium spread plus collateral yield when trigger is not hitSpecified catastrophe event / model riskMedium7.0%–11.0%
Insurer-backed contractual paymentsDefined future payments from contractual obligationsLegal title, insurer credit, duration / illiquidityLong6.0%–9.0%
Ground leases / infrastructure easementsRent or access payments tied to hard-to-relocate infrastructureTenant credit, duration, location obsolescenceLong5.5%–8.0%
Seasoned royalty streamsUsage of intellectual property, licenses, or contractual royaltiesRevenue decay, concentration, legal rightsMedium-long7.0%–12.0%
Tax-lien / municipal claim strategiesStatutory interest or redemption paymentsTitle, legal procedure, worthless collateral, jurisdiction riskUncertain5.0%–9.0%

Illustrative ranges are used only to demonstrate portfolio construction. They are not an indication that EDB Capital has investments available at these returns or that such returns will be achieved.

The six core engines we would investigate first

For a portfolio focused on lower probability of permanent loss rather than maximum upside, the strongest starting group is narrower than the full ten-engine menu:

  • Essential equipment finance: contractual payments plus identifiable collateral and potential secondary-market recovery.
  • Trade finance / receivables: short-duration payment obligations tied to specific completed commercial transactions.
  • Commercial insurance-premium finance: short duration with policy-cancellation and unearned-premium mechanics where permitted by applicable law.
  • First-lien farmland lending: conservative leverage against productive agricultural real estate.
  • Higher-attachment catastrophe bonds: principal risk driven primarily by defined catastrophe events rather than corporate earnings.
  • Insurer-backed contractual payments: defined payment streams where title, transferability, counterparty strength, and legal enforceability can be verified.

These are not equally safe and they are not perfectly independent. The attraction is that their dominant risk drivers are meaningfully different enough to justify further underwriting.

Stress the portfolio by event, not by asset class

A useful portfolio should survive several different worlds. The matrix below is qualitative. It illustrates the questions we would ask before allocating capital.

ScenarioMost exposedLess directly exposedWhat still can connect them
Deep recessionEquipment borrowers, trade sellers / obligors, some premium-finance borrowersCat bonds, Treasuries, many insurer-backed paymentsLiquidity, bank / servicer stress, counterparty failures
Inflation / higher ratesLong-duration fixed payment streams and long leases without escalatorsShort-duration receivables, premium finance, rolling TreasuriesBorrower debt-service pressure and refinancing costs
Major hurricane / earthquakeCat bonds tied to the affected peril and regionMost receivables, equipment, contractual payments outside the regionGeographic concentration and insurer exposure
Public-market crashAssets needing immediate sale or mark-to-market financingShort contractual private cash flows that can be held to maturityMargin calls, leverage, fund redemptions, shared banks
Insurer distressContractual insurer payments; certain insurance-linked structuresFarmland loans, equipment, ordinary trade receivablesConcentration in one carrier or insurance group
Fraud / servicer failureReceivables, equipment pools, premium finance if controls are weakExchange-traded Treasuries; independently custodied securitiesShared originator, administrator, lockbox, or custodian

Portfolio question: If one recession, hurricane, fraud, insurer failure, legal change, or liquidity squeeze occurs, how many positions can it realistically impair at once?

A hypothetical $1 million independent-return-engine portfolio

This example is intentionally conservative and is shown only to demonstrate the construction method. It is not a recommendation and would need to be adapted to the investor’s liquidity needs, taxes, access, legal structure, and risk tolerance.

EngineHypothetical allocationWhy it is present
Short-term Treasuries20%Liquidity, dry powder, and a non-private-market anchor
Equipment finance12%Contractual business payments plus hard collateral
Trade receivables12%Short duration and specific account-debtor payment risk
Premium finance12%Short-duration insurance-related cash flow
First-lien farmland loans12%Low-leverage real-asset secured credit
Higher-attachment cat bonds10%Natural-catastrophe risk distinct from ordinary business credit
Insurer-backed payment streams8%Contractual counterparty payment risk
Ground leases / easements6%Location-dependent contractual income
Seasoned royalties4%Usage / licensing-driven cash flow
Tax-lien / municipal claims4%Jurisdiction-specific statutory repayment mechanics
Illustrative gross target: ~5.7%–8.4%

Using the broad underwriting ranges above, the hypothetical allocation produces a simple weighted range before fees, losses, taxes, legal costs, cash drag, and manager expenses. It is not a forecast. The purpose is to show that a portfolio can target a respectable overall return without requiring every sleeve to seek double-digit returns.

The larger point is behavioral: a portfolio of modest, unrelated outcomes can be more valuable than one apparently superior yield that exposes the investor to a common tail risk.

The six forms of hidden correlation

The label on an investment is only the first layer. We would maintain separate concentration limits for:

  • Manager: several strategies controlled by one manager can become one operational bet.
  • Servicer / originator: fraud, weak controls, or bankruptcy can impair otherwise unrelated assets.
  • Counterparty: one insurer, bank, customer, or guarantor may sit behind multiple positions.
  • Liquidity: too many locked assets can force a sale elsewhere at the wrong time.
  • Leverage: financing can make independent assets correlate during stress through margin calls or covenant breaches.
  • Economic cycle: equipment, receivables, and generic private credit can all deteriorate together when business cash flow falls.

This is why “private” does not automatically mean diversified and why smooth quarterly marks should never be confused with low economic risk.

Implementation: build originator relationships, not isolated deals

For several of these strategies, the practical route is not to find one attractive loan. It is to identify a repeat originator with a documented history and define a narrow credit box for future transactions. That can include maximum advance rates, minimum seasoning, collateral standards, payment-control requirements, concentration limits, first-loss retention, and stop-funding triggers.

In other categories—particularly catastrophe bonds—the more sensible route may be a specialist manager with diversified institutional access rather than direct deal-by-deal selection.

Our preference is to know in advance what qualifies, what automatically fails, who services the asset, where cash is collected, how fraud is detected, and how the position can be exited or allowed to mature.

Use one underwriting framework across every engine

The strategies differ, but the questions should be consistent: What pays us? What protects principal? Who controls the cash? What breaks the structure? What would recovery look like? How long is capital locked? Who else is exposed to the same risk?

Read the EDB Capital 25-Question Specialty Finance Underwriting Framework.

Research Methodology

How this research is developed

EDB Capital research prioritizes primary regulatory, government, and established industry sources. The analytical framework focuses on source of repayment, collateral, payment control, concentration, liquidity, servicing, legal rights, and downside protection. Illustrative examples are analytical tools—not forecasts or promises of return.

Sources & Further Reading
Bryan Leighton, founder of EDB Capital
About the Author

Bryan Leighton

Bryan Leighton is the founder of EDB Capital. For detailed professional background and experience, view his LinkedIn profile or the site’s About page.

For informational purposes only. This material describes EDB Capital's general investment interests and analytical framework. It is not investment, legal, tax, or accounting advice; an offer to sell securities; or a solicitation of outside investment capital. Any opportunity is subject to independent due diligence, documentation, legal and regulatory review, and final approval.