How it works
A leveraged borrower in trouble has three honest options: pay, renegotiate with everyone, or file for bankruptcy. A liability-management exercise, or LME, is the fourth option, and it exists because loan documents written during a decade of lender enthusiasm turned out to have doors in them.
There are two classic forms, and almost everything else is a variation on them.
The drop-down moves assets out of reach. Most credit agreements let a borrower transfer some quantity of assets to a subsidiary that is not a guarantor of the existing debt — an "unrestricted subsidiary" — using baskets that were sized for ordinary business reasons and then never re-examined. The valuable collateral walks out of the box the lenders thought it was locked in, and the newly unencumbered subsidiary borrows fresh money against it. The original lenders keep their claim. The thing their claim was secured on is now somewhere else. The manoeuvre is named after J.Crew, whose intellectual property took precisely this walk.
The uptier reorders the queue. Credit agreements can usually be amended by a simple majority of lenders. So the borrower assembles a majority, and that majority votes to permit a new tranche of debt ranking ahead of the existing loan, then exchanges its own holdings into the new senior tranche at a favourable price. The lenders who were not in the room wake up holding the same instrument, now to the people who used to sit beside them. This one is named after Serta.
An illustration, with invented round numbers. A company owes 1,000 on a first-lien loan trading at 60 cents. It needs 200 of new cash. It finds holders of 55 per cent of the loan, who agree to lend the 200 as a new super-senior tranche and to exchange their existing 550 of old loan into a second-priority tranche at, say, 75 cents. They vote the amendment through. The remaining 450 is now third in line behind 200 of new money and roughly 410 of exchanged paper, on the same collateral, and promptly trades at 35. Nobody defaulted. Interest was paid on time. Twenty-five cents on moved from one set of lenders to another by document amendment.
The defensive response is the cooperation agreement: lenders sign contracts with each other promising not to be the majority that does this to the rest. That such agreements are now routine tells you how routine the practice is.
What makes all of it possible is the package. documents remove the maintenance tests that would have forced an early conversation, and permissive baskets supply the exits. The borrower is not cheating. It is using the contract the lenders signed.
Why it matters to this crash
An LME is a default that does not print as one. That matters enormously in a credit system whose health is assessed largely through counts of things that did print.
Default rates look calmer than the underlying stress, because a distressed exchange gets negotiated rather than filed. Recovery rates on first-lien loans get worse, because the whole point of the exercise is that first-lien no longer means first. And in , where the loan never trades and the lender chooses the , an LME can be booked as an amendment rather than an impairment. The borrower keeps servicing, often by switching to interest, so the loan never goes to either. The fund reports a performing asset. The asset is not performing. It has been rearranged.
This is why we treat clean-looking credit statistics with suspicion. The mechanism for converting a loss into a documentation event is now standard practice, widely lawyered, and cheap.
We have not yet covered a specific LME in our dispatches, so we are not attaching this to a named deal or a size. What we can say is that the term is going to appear, and that when it does it will be doing the work the word "default" used to do.
What would make this dangerous
Re-defaults are the clearest sign. A borrower that does a second LME within a year or two of the first has demonstrated that the exercise bought time rather than solvency, and the second round has less collateral to work with. A rising share of these among distressed situations would mean the technique is being used to postpone rather than to fix.
Second, recovery rates. If first-lien recoveries in completed restructurings settle durably below the historic norm for senior secured debt, the seniority that every credit model assumes is worth less than the models say, and that repricing hits and loan valuations across the board, not just at the borrowers involved.
Third, the private-credit tell: a fund reporting flat and flat through a period of visible restructuring activity. That combination is only possible if losses are being classified rather than recognised.
Fourth, cooperation agreements failing. They are contracts between lenders, and a lender offered a large enough incentive to break one will eventually break one. The first prominent case where the cooperating group fractures is the moment the informal defence stops working and the documents are all there is.