Distressed Debt Report 2010 — The Wave That Was Cancelled

By 2009 the arithmetic looked unambiguous. The 2005–2007 leveraged finance vintage had been underwritten on assumptions that the crisis had destroyed — aggressive leverage multiples, thin covenants, growth forecasts that no longer applied. Those obligations matured on a schedule, and that schedule was known. It became known as the maturity wall.

The expected consequence was a default cycle of historic scale, and capital positioned for it. Distressed funds raised substantial commitments through 2008 and 2009 on that thesis. The logic was clean: over-levered companies with debt coming due into a damaged banking system cannot refinance, so they default, so debt trades at prices offering equity-like returns to whoever holds it through restructuring.

The wall was reached and largely not hit. Default rates rose from very low levels, peaked well below what the leverage statistics implied, and fell back. A great deal of the deployed distressed capital was returned or reallocated.

The explanation is not that the borrowers recovered. Many did not. The explanation is that they did not need to. The policy response described in the Global Investment Outlook 2009 — rates at the lower bound and asset purchases compressing yields — made refinancing available to credits that could not have refinanced at normal rates. A company unable to repay its debt but able to replace it with new debt does not default. It continues.

This is the single most consequential finding of the post-crisis credit period, and it generalises well beyond 2010. Default is not a mechanical consequence of unaffordable debt. It is what happens when unaffordable debt meets an unwilling or unavailable refinancing market. Remove the second condition and the first can persist indefinitely.

The consequences compounded quietly. Capital stayed committed to businesses that could not service their obligations from operations. Lenders traded covenant protections for extensions, beginning an erosion that ran for a decade. And distressed investing itself changed — from a cyclical strategy harvesting a predictable wave into a specialist one dependent on complexity, because the cycle it was built to harvest had been suppressed.

The wall that was real

The forecast was not foolish, and understanding why it was reasonable is necessary to understand what actually happened.

What had been underwritten in 2005–2007:

• High leverage multiples, justified by stable cash flows and available credit.
• Weakening covenant packages, already loosening before the crisis.
• Growth assumptions embedded in the ability to service and refinance.
• Bullet maturities concentrated in a small number of years, because the vintage was concentrated.

What the crisis did to those assumptions: cash flows fell, the growth forecasts became unattainable, the banking system that had provided the credit was impaired, and the securitisation vehicles that had bought much of the loan paper had stopped functioning.

So the analysis ran: a large stock of debt, underwritten on assumptions now invalid, maturing into a market with no capacity to refinance it. Default is the only available outcome.

Every step in that chain was correct except the last, and the error is instructive. The analysis treated refinancing capacity as a function of the banking system's health, which was genuinely impaired. It did not anticipate that the price of credit would fall far enough to bring an entirely different set of lenders into the market — investors reaching for yield because the risk-free alternative had been removed.

The wall was measured correctly and the ability to climb it was not. Refinancing capacity is a function of the price of credit, and the price was about to be set by policy rather than by the banks.

The Emerging Market Debt Report 2013 describes the same search for yield reaching a different asset class. It is one phenomenon with several destinations, and leveraged credit was the nearest one.

What zero rates do to a default rate

The mechanism connecting policy rates to corporate defaults is direct, and it is worth stating in steps because it is often compressed into an assertion.

Default requires two conditions, and both must hold:

1. The borrower cannot service or repay the obligation from operations or asset sales.
2. The borrower cannot replace the obligation with a new one.

Condition two is a market condition, not a borrower characteristic. It depends on whether any lender will provide new money and at what price. And that depends on the alternatives available to lenders:

• When the risk-free rate is meaningful, a lender comparing a stressed credit against a safe yield frequently prefers the safe yield. Weak borrowers cannot refinance.
• When the risk-free rate is near zero, the same comparison changes completely. A lender with a return requirement and no risk-free alternative will consider credits they would previously have declined.

So policy rates set the default rate through the refinancing channel, more or less directly, and the effect is largest for the weakest borrowers — the ones whose refinancing is most marginal.

Three consequences follow, and all three ran for a decade:

• The default rate stops reflecting corporate health. It becomes a joint statement about borrower fundamentals and credit availability, and the second dominates in accommodative conditions.
• Distress is deferred, not avoided. The debt still exists and still cannot be serviced from operations. The test is postponed to the next maturity, and postponed again.
• The eventual reckoning arrives with rates, not with a recession — which is why the correction described in the Global Investment Outlook 2023 followed a rate reset rather than an economic contraction.

This is the archive's deferral pattern in its purest form. Nothing was resolved. The test of whether these businesses could support their capital structures was postponed at each maturity, and each postponement was individually rational for every party involved.

What extend-and-pretend actually does

The practice acquired a dismissive name and deserves a precise description, because the alternative was not obviously better.

The mechanism. A loan approaching maturity that cannot be repaid is amended — the maturity extended, terms adjusted, sometimes a fee paid — rather than enforced.

Why the lender prefers it, and these are real reasons:

• Enforcement crystallises a loss. An extended loan continues to be carried at or near par; an enforced one produces a recovery that is usually well below it.
• Capital treatment is better. A performing loan consumes less capital than a defaulted one, and in 2010 bank capital was the binding constraint on everything.
• The recovery environment was genuinely poor. Selling collateral into a distressed market produces a bad outcome — the forced-sale problem the US Housing & Mortgage Report 2008 describes.
• The borrower might recover. Extension is a real option, and options have value.

Why the borrower prefers it is obvious, and worth noting: the alternative is losing the business.

So the arrangement is consensual and locally rational, which is precisely why it is so persistent and so hard to police.

What it does at the system level is different:

• Capital stays allocated to businesses that cannot service it, and is therefore unavailable to businesses that could use it — the misallocation cost, which is real and almost impossible to measure.
• Price discovery stops. An extended loan carried near par produces no market price, so nobody learns what the credit is worth. The absence of a mark is not the absence of a loss, the same point the US Venture Capital Report 2008 makes about private valuations.
• The problem becomes correlated with rates. A book of extended loans is a leveraged position on rates staying low, held by institutions that did not think of it that way.

The honest assessment: in 2010, with banking systems impaired and recovery markets broken, forcing recognition would have caused a second crisis. Extension was defensible as a bridge and became a habit, and the transition from one to the other has no identifiable date.

The zombie question

The persistence of firms unable to service their debt from operations became one of the defining structural questions of the following decade, and it originates here.

The definition that matters is not sentiment: a firm whose operating earnings persistently fail to cover its interest costs, which continues to exist because credit remains available.

Why they persist:

• Refinancing is available at rates their earnings can nearly cover, because the risk-free alternative is negligible.
• Lenders prefer extension to recognition, for the reasons above.
• No party has both the incentive and the ability to force resolution. The lender does not want the loss; the borrower does not want to fail; and in a syndicated structure, no single holder can act alone.

The cost is not primarily the failure of these firms — most did not fail. It is what their survival does to everyone else:

• Capacity that would have exited remains, suppressing prices and returns for healthier competitors in the same industry.
• Capital and labour stay locked in low-productivity uses rather than reallocating.
• New entrants face incumbents who are not earning their cost of capital and therefore price as though they need not.

A firm that cannot cover its interest is not merely a problem for its lender. It is a competitor that does not need to earn a return, and that is a problem for everyone selling the same thing.

The counterargument deserves weight. Forced liquidation in 2010 would have destroyed viable businesses along with unviable ones, because distinguishing them in a downturn is genuinely hard and recovery values were artificially low. Some forbearance preserved real value. The critique is about duration, not about the initial decision — and about the fact that nothing in the arrangement created a mechanism for ending it.

Where covenant erosion started

The decade-long weakening of lender protections is usually dated to the mid-2010s. Its origin is in the restructurings of 2009–2011, and the mechanism is specific.

The negotiating position. A lender holding a loan that will default without amendment has less leverage than the documentation suggests, because enforcement produces a poor recovery. The borrower's threat — to default and force enforcement — is credible precisely because the recovery environment is bad.

What lenders traded away in exchange for extensions:

• Maintenance covenants, tested periodically, replaced by weaker incurrence-based tests or removed.
• Restrictions on additional debt and on asset transfers, loosened.
• Information rights, reduced, which compounds later by making the next assessment harder.

Why the trade looked sensible at the time: the alternative was a default the lender did not want, and a covenant is only useful if you intend to act on it.

Why it mattered later, and this is the compounding part:

• Terms are sticky. Once a structure has been accepted in a restructuring, it becomes a precedent for new issuance.
• Competitive pressure ratchets one way. A lender insisting on stronger protections loses the deal, and there is no mechanism for restoring terms collectively.
• Weaker covenants delay the next cycle's recognition, because there is no test to breach. Distress arrives later and larger — the Private Credit Report 2016 and Private Credit Report 2024 both examine where that leaves the current market.

So the 2010 restructurings did more than defer the existing cycle. They weakened the machinery for detecting the next one.

What an allocator could act on

Model defaults as a function of credit availability, not just leverage. A default forecast built from leverage multiples and interest coverage is incomplete without a view on refinancing conditions, and in accommodative conditions the second dominates.

Treat a suppressed default rate as deferral, not health. Debt that cannot be serviced from operations remains a problem after it is extended. The test moves to the next maturity, and the cumulative stock grows.

Recognise that a book of extended loans is a rates position. An institution holding many amended credits has an undiversified exposure to refinancing conditions persisting. That is a macro position, and it is rarely described as one.

Ask what a strategy's return depends on before committing to it. Distressed capital raised for a cycle that policy suppressed had to deploy into something else or return capital. A strategy dependent on a variable set by policy has a capacity constrained by policy, not by opportunity.

Watch covenant terms as a leading indicator, not a documentation detail. Erosion begins in restructurings, spreads to new issuance by precedent, and determines how late the next cycle's distress becomes visible.

Discount unpriced positions rather than trusting them. A loan carried near par with no market and no test is an assumption. The absence of price discovery is itself the risk, and it is the common thread linking this report to private equity marks and to Chinese local government debt.

What 2010 established

• Default requires both unaffordable debt and an unavailable refinancing market, and policy controls the second directly.
• Extend-and-pretend converts a credit event into a duration extension — locally rational for every participant, and costly at the system level through misallocation and lost price discovery.
• Persistent non-earning firms impose costs on healthy competitors, not principally on their own lenders.
• Covenant erosion began in crisis restructurings, where lender leverage was weakest, and ratcheted one way for a decade.
• A strategy dependent on a policy-suppressed variable has a policy-constrained capacity, which distressed investing discovered by raising for a cycle that did not arrive.