A first restructuring recovers 70 to 90 cents. The average is 50, and repeats are why.

What routes exist before a filing?

Five, and they are not sequential options that remain available while an owner decides between them. They expire in order. A consensual going-concern sale, run as a compressed process with the lender informed. A refinancing or replacement of the facility, which requires a lender willing to advance against today's numbers. A forbearance or amendment, which buys time against a plan. A secured party sale under Article 9 of the Uniform Commercial Code, which is the lender's route rather than the owner's. And a filing, in or out of Subchapter V depending on the debt.

The distinction that matters is who controls the timetable. In the first three the owner is still the counterparty setting the pace, even if the leverage has gone. In the fourth and fifth, somebody else is. Owners consistently understate how quickly the transition happens, because nothing visible occurs on the day it does.

The market context is that lenders are increasingly choosing route four for themselves. Direct lenders foreclosed on 24.2 billion dollars of principal across 2025 and a further 15.2 billion in the first quarter of 2026 alone, against 13.6 billion across the preceding three years combined, with nearly 75 percent of those change-of-control transactions relating to 2021 and 2022 vintage loans (Lincoln International, as at 31 March 2026). Advisers now report private credit lenders commissioning carve-out-style hundred-day transition plans before taking control (FTI Consulting, reported by Octus, 29 July 2026). The lender is not deciding whether to act. It is preparing to.

Why do the best routes expire first?

Because each of them requires something the passage of time removes. A consensual sale requires a business that a third party can underwrite, which requires trailing financials that are not yet catastrophic and a management team that has not left. A refinancing requires a new lender willing to advance against current earnings, which is a function of coverage rather than of the story. A forbearance requires a lender who believes the plan, and belief is spent once.

The extension market is where this is now visible in published data. Amend-and-extend volume set records in the first half of 2026, and yet the credit quality of who is being extended rose sharply: 30 percent of 2026 amendments were rated BB minus or higher at the issuer level, up from 11 percent in 2025, while the B minus share fell to 27 percent from 44 percent (PitchBook LCD, 17 July 2026). Record volume and a collapsing weak-credit share together mean the weakest cohort is being cut out of the extension market rather than served by it.

The second time is worse than the first. Between 30 and 40 percent of direct lending deals maturing in the next two years have already extended once, and the lender's own framing puts an incremental extension and a restructuring on the same fork (Lincoln International, 11 February 2026). And where an out-of-court liability management transaction has been done, the historical record is that roughly one in four of them ends in a Chapter 11 filing anyway, on a tracked population of 148 transactions producing 37 bankruptcies (Covenant Review and LevFin Insights, research dated 14 January 2026).

Finally, the backdrop is not improving. The share of the leveraged loan index trading below 80 cents stood at 6.87 percent in June 2026 against 3.06 percent a year earlier, and the rating agency publishing the headline default rates has said in terms that the first-half decline was driven by base effects rather than credit improvement, with default activity expected to accelerate in the second half of 2026 (PitchBook LCD, 10 July 2026; Fitch Ratings, US Corporate Distressed and Default Monitor: July 2026, published 17 July 2026). An owner reading a falling default rate as evidence that there is time is reading an arithmetic artefact.

What does each route actually recover?

The recovery evidence is genuinely contested and it splits along a line that is useful to an owner. Across private credit restructurings, average recoveries run at approximately 50 cents on the dollar, well below the 70 percent commonly cited by managers, and the cause named by the firm measuring it is repeat restructurings: each successive restructuring of the same credit destroys more principal (Octus, 11 May 2026).

But the same source publishes the counter-example. Two consensual restructurings in a non-disrupted healthcare services sector in the second quarter of 2026 produced first-lien debt reduction of 45 to 65 percent with first-lien recoveries of 70 to 90 percent, per the lenders' own filings (Octus, private credit coverage, second quarter 2026). Consensual, first-time, single-lender-group restructurings in sectors that are not being structurally disrupted still clear at the top of the range. The 50-cent average is dragged down by the repeats and by the disrupted sectors.

At the other end, the liquidation evidence is stark. Excluding asset-based recoveries, par-weighted first-lien recoveries on claims against retail companies with at least 100 million dollars of first-lien debt averaged approximately 49 percent across 2016 to 2025, liquidations remain common with nearly half of those bankruptcies resulting in liquidation, and term loan lenders recovered nothing at all in three named cases (Fitch Ratings, US Retail Bankruptcies Show Weak Recoveries, High Liquidation Rates, 10 June 2026). In an inventory-dependent or asset-light business, first lien is worth roughly half of par on average, and can be worth zero.

One thing nobody publishes. There is no primary quantitative series for 2026 giving Section 363 sale counts or distressed M&A volume; law-firm surveys and trade coverage discuss the trend without a countable number. Any figure quoted for distressed deal volume this year should be treated as an estimate, and none is used here.

Is your lender a neutral counterparty?

No, and the honest version of that is not adversarial. A lender in a loan-to-own posture has a legitimate economic interest in acquiring the business at the value of its debt rather than being repaid at par, and a lender who has already written the hundred-day plan has decided which outcome it prefers before the meeting.

The structural feature that matters in a private credit workout is information rather than procedure. There is no steering committee, no publicly traded price and no agent-published amendment. Valuation gaps of nearly 40 points have been observed between funds holding the same stressed loan, and non-accrual reporting varies enough across funds that adjusting the universe for commonly held loans raises the sector figure materially (Octus, 11 May 2026). Two lenders in the same facility can hold irreconcilable views of what the business is worth before a negotiation starts.

The practical instruction that follows is unglamorous and decisive: establish what every lender in the structure has the loan marked at before proposing anything. A proposal pitched at a value one lender has already written down to is a different proposal to the lender who has not.

The second is to separate the roles. The lender's advisers are not the company's advisers, a turnaround manager introduced by the lender is not neutral, and a sale process run to the lender's timetable is a sale process run for the lender's recovery. None of that is improper. It is simply not the owner's mandate.

What should an owner do first?

Build a thirteen-week cash forecast and keep it current. Every route above is decided on liquidity runway, and an owner who cannot say with confidence which week the facility becomes binding has no basis for choosing between routes and no credibility in the conversation where they ask for time.

Second, read the documents before the lender does. Which covenant is tested, on what definition, on what date, with what cure rights and what cushion. Typical leverage covenant cushions run 25 to 35 percent to the borrower's own model and springing tests are commonly triggered above 40 percent revolver usage (Sidley Austin, 24 March 2026). The difference between a covenant that is tested quarterly on a trailing basis and one that springs on a draw is often the difference between having a quarter and having a week.

Third, decide the objective before the meeting, because the routes optimize for different things. Maximum proceeds points toward a consensual going-concern sale run early. Continuity of employment and customer relationships points toward a restructuring that keeps the entity intact. Protection of personal guarantees points somewhere else again, and it is the one owners often will not say out loud, which then quietly drives every decision they make.

The general shape of the market supports moving early rather than waiting. Commercial Chapter 11 filings rose 28 percent year on year in the first half of 2026 and Subchapter V elections rose 50 percent, while the headline loan default rate fell, which means stress is concentrated below the rated universe in exactly the size of business this piece is about (Epiq AACER and the American Bankruptcy Institute, 8 July 2026). Those filings are the population that ran out of routes.