2021-vintage private equity funds: what to do with companies bought at peak multiples

The short answer: earnings growth and cash, not the multiple

Companies bought in 2021 were often priced at high multiples of EBITDA and financed with debt that was cheap at the time; where that debt floats, interest costs may now be well above what the deal model assumed. A 2021-vintage fund that wants to return capital on those deals cannot count on selling at the same multiple, so it has to grow EBITDA, release cash and keep debt costs under control. The levers that work fastest without new capital deserve the first look.

The wider market explains why. McKinsey's Global Private Markets Report 2026 says multiple expansion and cheap leverage, which accounted for 59 percent of PE returns between 2010 and 2022, have faded, leaving operational value creation as the likely primary source of returns. The same report notes that firms have more than doubled their operating groups since 2021, which is the year many of these deals closed.

What does 2021 vintage actually mean?

A fund's vintage year is usually the year of its first close or first capital call, depending on the data provider's convention. A 2021-vintage fund therefore invested across several years, not only in 2021. The pressure concentrates in companies closed during 2021 and early 2022, when prices and leverage were highest; later deals in the same fund may have been bought on better terms.

By late 2026, a 2021-vintage fund is around the end of its investment period and entering the stage where LPs judge it on cash returned rather than paper value. The private equity fund lifecycle guide maps what changes for the operating team at that point. Four metrics frame the conversation:

Distribution pressure is industry-wide, not only a 2021 problem. Bain's Global Private Equity Report 2026 reports distributions as a percentage of NAV below 15% for four years running and about 32,000 unsold companies worth $3.8 trillion. Those figures cover all vintages. We do not quote vintage-specific DPI here; your fund administrator or LP reporting gives the real numbers for your own fund.

Why does a peak entry multiple change the plan?

Equity value at exit is roughly EBITDA times the exit multiple, minus net debt. If buyers pay a lower multiple than the fund did, EBITDA has to grow just to stand still, and higher interest costs absorb cash that would otherwise reduce debt.

Bain's Global Private Equity Report 2026 makes a similar point across the market: it says a deal that needed 5% EBITDA growth a decade ago now needs about 12% to reach a 2.5x return over five years.

That arithmetic sets the playbook for these companies: protect and grow EBITDA, convert more of it into cash, and find sources of cash that do not need fresh capital from a fund with little left to invest.

Which levers fit, by speed and capital need?

Sort the options by how fast they produce cash or earnings, what they cost, and whether a buyer will count the result in run-rate EBITDA.

Software companies bought in 2021 face an extra pricing question as AI changes how many seats customers need; see seat compression and AI. For fund-level liquidity, the explainer on what a NAV loan is covers how those facilities work and what they mean for portfolio companies.

Where does a one-time data license fit?

For a qualifying company, a data license turns operational records it already keeps into a one-time payment from AI labs and data buyers, with no new capital, no dilution and no change of ownership. The company keeps its data, licenses it for an agreed term (deals are typically exclusive for AI training), and agrees one all-in price before anything is signed. The company is usually paid within about 60 days of the invoice, once a buyer has chosen the data.

Three features make it relevant to a peak-multiple company:

• It is capital-light. The main cost is management time to build a data inventory and answer rights questions.
• It does not depend on the exit market. A company can pursue it while waiting for a better sale window.
• Its cash can go where the plan needs it, within the credit agreement: paying down a revolver, funding an add-on or rebuilding liquidity. Moving proceeds up to the sponsor is a separate question governed by restricted payments covenants; see whether license proceeds can be paid to the sponsor.

What it does not do: raise run-rate EBITDA. It will not close the multiple gap, so present it as non-recurring in board packs, lender reporting and any sale process.

A company bought in 2021 can have the depth buyers look for: years of operations before the sponsor arrived and several more since. Archived and retired systems help because they extend the history, so ask what was migrated or switched off along the way. The baseline is a US company that reached 50+ full-time employees at peak (contractors excluded), has several years of documented operations, holds the rights to license its records, and has an owner, CEO, CFO or authorized representative able to sign; SourceX makes the qualification call. The who qualifies page has the detail.

How to screen a 2021-vintage portfolio in one pass

1. List every company closed in 2021 and 2022 with its entry multiple, current leverage and covenant headroom.
2. For each one, note peak headcount, years of operating history and the main systems: CRM, ticketing, ERP, engineering tools, shared drives, Slack or Teams.
3. Strike any company that fails on rights or on whether the records survive: client-owned material without consent, a dataset dominated by consumer or patient information, histories already purged, or an existing AI-training license on the same data.
4. Ask the CEO and CFO of each remaining company whether they would accept exclusivity for AI training over a fixed term in return for a one-time payment.
5. Introduce the willing ones to SourceX through your referral link or the referral form. SourceX runs qualification, pricing, buyer review, contracting and delivery, and the company completes its own data inventory; you never handle the data.
6. Record a result for every company, including those that did not qualify, so the screen is documented and can be repeated each year.

How does the partner reward work?

Partners earn 25% of the eligible platform fees SourceX actually collects from the referred company's licensing deals, capped at $100,000 per referred company. Rewards become payable only after the buyer pays and SourceX receives its fee; a lead, meeting or signed agreement alone does not trigger payment, and no reward is guaranteed. Because the reward is a share of SourceX's fee, it never reduces what the portfolio company receives. Ask your compliance team whether fund documents or firm policy limit fees tied to portfolio companies before registering; the referral program for operating partners explains the sponsor's side.

Limits and open questions

• No vintage-level performance data here. Judge a specific fund by its own DPI, TVPI and LP reporting, not by its vintage label.
• A license does not repair a capital structure. If covenants are tight, the CFO should talk to lenders before counting on any one-time cash.
• Exclusivity has to fit the exit plan. A future buyer of the company will read the license, its term and any continuing obligations.
• Accounting needs an answer early. Ask the company's auditors how the license will be recognized before it is signed.
• Not every company qualifies. Agencies and outsourcers whose records mainly concern their clients usually fail on rights.

This is general information, not legal, tax or financial advice. Confirm with your own counsel, tax adviser or professional body before acting.

Next step

Pull the list of 2021 and 2022 closings this week and run the six-step screen; the network opportunity finder is a quick way to think about which portfolio executives you know well enough to call. Once a company passes, register as a partner so the introduction is credited, then send its CFO your referral link to sourcex.si/apply.