The American Securitisation Forum's annual conference, held in Las Vegas in January 2007, was the high-water mark of structured credit optimism. The rooms were full. The deals were large. The ABX indices — standardised credit default swap baskets referencing tranches of subprime mortgage-backed securities — had been trading since January 2006 and were widely used to hedge and express views on the subprime market.
What happened in the months surrounding that conference was the first visible signal of what was coming. The ABX HE 06-2 index, referencing 2006 vintage subprime mortgages, began selling off sharply. Not dramatically — with the persistent, grinding pressure that credit practitioners recognise as the market quietly re-pricing risk that had been misassessed.
The selling was being driven by a small number of funds that had done the forensic work. They had looked at the loan-level data in the mortgage pools. They had noticed that delinquencies on 2006 vintage subprime mortgages were running ahead of model expectations even before the loans had seasoned. They had identified that FICO scores were being calculated on the basis of the primary borrower only, while debt-to-income ratios included income that was never verified. They had found that the same properties were being appraised multiple times in the same quarter, each time at a higher value, by appraisers with a financial interest in the transaction closing.
The gap between what the models said and what the data showed was not a rounding error. It was a fundamental misrepresentation of the credit quality of the underlying assets, compounded by structures that amplified losses through leverage and waterfall mechanics that very few people fully understood.
The sequence that followed is documented history. But its structure — the slow realisation, the gap between public signals and private marks, the delay between the first visible stress and the crystallisation of losses — is the template against which the current private credit situation needs to be assessed.
II. The Timeline: From Las Vegas to Lehman
The following timeline is not offered as a prediction. It is offered as a structural map of how credit dislocations unfold, for the purpose of identifying where on that map the current situation sits.
The ABX indices begin their decline. The stress is visible only to those watching the specific instruments and is widely interpreted as technical rather than fundamental. Rating agencies have not moved. Bank earnings are strong.
HSBC announces a $10.6 billion writedown on its US subprime mortgage portfolio in February 2007. New Century Financial, one of the largest subprime lenders, files for bankruptcy in April. Bear Stearns launches a $3.2 billion rescue of one of its structured credit hedge funds in June. At each point, official commentary characterises the problems as contained. Ben Bernanke tells Congress in March 2007 that the impact of subprime is "likely to be contained."
What made this period so important, and so misunderstood in retrospect, is that two completely separate realities were operating simultaneously. Almost nobody outside institutional capital markets could see both of them at once.
In the world that retail investors and most financial commentators were watching, the stock market was volatile but functioning. Life, on the surface, appeared to be continuing.
I was sitting at an institutional trading desk in Bahrain, managing a large institutional Agency MBS, SBA (HQLA) and IG/EM credit book, and I can tell you that what we were watching was an entirely different reality.
The instrument I watched most closely was the LIBOR-OIS spread: the gap between the rate at which banks said they would lend to each other and the overnight index swap rate. In normal times, this spread sits at 7 to 10 basis points. When you are managing a fixed income book, you check it the way a doctor checks a patient's pulse: not because it is usually interesting, but because when it moves, something serious is happening.
In late July 2007, it started moving.
By August 10, 2007, the three-month LIBOR-OIS spread had hit nearly 50 basis points. By September it had passed 90 basis points. Before the crisis, it had been sitting at 7 to 9 basis points. The message was unambiguous to anyone reading it: banks had stopped trusting each other.
I watched short-end LIBOR spike not because of a policy rate move, but because banks were refusing to lend to each other at rates that assumed away counterparty risk. That is a categorically different signal from a normal rate move. It means the plumbing is seizing. The commercial paper market was experiencing the withdrawal of buyers who had previously treated it as risk-free. The repo market was beginning to price haircuts on collateral that had never carried haircuts before.
None of this was visible in the equity market. It was visible only to people sitting at institutional desks, watching the bids disappear from instruments that were supposed to have permanent liquidity. I was one of those people. I remember the feeling precisely.
On August 3, 2007, Jim Cramer appeared on CNBC's Street Signs and delivered what became one of the most viscerally honest pieces of financial television ever recorded. Pounding the desk, visibly shaken, he said:
"I have talked to the heads of almost every single one of these firms in the last 72 hours and he has no idea what it is like out there. None. My people have been in this game for 25 years and they're losing their jobs and these firms are gonna go out of business and he's nuts. They know nothing."
His demand was specific and institutional: that Bernanke open the Federal Reserve's discount window. The retail viewer saw an emotional television personality. I heard something different. I heard a former hedge fund manager describing in real time exactly what I was watching in the interbank market. He was not performing. He was genuinely panicked, and he was right. The Fed cut the discount rate on August 17th. It was not enough, and it was not fast enough.
BNP Paribas suspended redemptions from three investment funds on August 9, 2007, citing its inability to value assets in the US mortgage market. Every institutional desk in the world read that and drew the same conclusion: if BNP cannot value these assets, nobody can. The marks across the industry were fiction.
The retail investor did not know any of this was happening. I did. Everyone on a fixed income desk in any major financial institution did. The question we were all asking was not whether there was a problem. The question was how large it was, how deep it went into the balance sheets we could not see, and how long it would take for the surface to crack.
Merrill Lynch announces a $7.9 billion writedown in October 2007. Citigroup's Chuck Prince resigns after disclosing losses that would eventually reach tens of billions. The phrase attributed to Prince in a Financial Times interview shortly before the crisis deepened — "As long as the music is playing, you've got to get up and dance" — becomes the defining description of institutional complacency at a credit peak.
This phase is characterised by the sequential acknowledgment of losses that had been building for months or years. Each announcement surprises the market not because the losses are unexpected in retrospect, but because the speed and scale of acknowledgment is compressed into a short period.
Bear Stearns is acquired by JPMorgan Chase in March 2008 in a Fed-facilitated transaction at $2 per share. The US government places Fannie Mae and Freddie Mac into conservatorship in September 2008. Lehman Brothers files for Chapter 11 on September 15, 2008. Within days, the money market fund Reserve Primary "breaks the buck" after its Lehman paper becomes worthless, triggering a run on the money market fund industry that requires a Treasury guarantee to halt.
The crescendo is not the beginning of the crisis. It is the public acknowledgment of a process that began eighteen months earlier in the ABX indices and the loan-level data of 2006 vintage subprime mortgages.
III. Where Are We Now
The evidence suggests a position somewhere between the first signals and the denial phase of the template. The public signals — BDC discounts to NAV, rising PIK income, increased sponsor equity support for struggling portfolio companies — are present and observable. The private marks have not moved in a way that corresponds to these signals. The official commentary from managers and regulators continues to characterise the issues as manageable and idiosyncratic.
This is precisely the characterisation that was being applied to subprime mortgage stress in the first half of 2007.
Several structural features of private credit make the analogy imperfect. Subprime mortgage risk was amplified through CDO and CDO-squared structures that allowed the same underlying credit risk to be owned multiple times over. Private credit risk is more directly held. This limits the amplification mechanism. The Fed also has more room to ease than it did in 2006. What is structurally similar is the opacity, the valuation lag, the growth of the asset class beyond the universe of suitable borrowers, and the presence of retail capital in vehicles that carry liquidity mismatches.
IV. The Practitioner's Assessment
Managing a credit book through 2007 and 2008 produced a specific set of observations that are relevant to the current situation.
The first observation is that the early signals in credit markets are almost always correct in direction and wrong in timing. The ABX was right that subprime mortgages were overvalued by roughly eighteen months before the broader market acknowledged it. Investors who acted on the signal too early were right about the trade and sometimes lost money before making it. Investors who waited for confirmation acted too late.
The second observation is that counterparty risk is never fully mapped until it is tested. AIG's credit default swap book was not visible to the institutions depending on it as a hedge. Reserve Primary Fund's Lehman exposure was not visible to investors who treated money market funds as cash equivalents. The private credit market's distribution through insurance companies, pension funds, non-traded BDCs, and the balance sheets of banks in Japan, South Korea, and Europe is similarly opaque from the outside.
The third observation is that the phrase "this time is different" is most dangerous when it is most plausible. In 2006 and 2007, the argument that structured credit instruments had genuinely distributed and diversified risk was built on real theoretical foundations. It was wrong because the theory assumed that distributed holders of the risk would behave independently, when in fact they were correlated through their funding structures and their shared exposure to the same underlying asset class. The argument that private credit is different from subprime because it is senior secured, directly held, and managed by sophisticated institutions is similarly not absurd. It may also be similarly incomplete.
V. The Asia Lens
Japanese regional banks that reached for yield in US private credit and CLO structures during the period of domestic zero rates made decisions that were individually rational and collectively significant. The Bank of Japan's own financial stability assessments have flagged the opacity of these exposures. When those assets are eventually marked to something closer to their economic value, the implications for the capital adequacy and credit extension capacity of these institutions will be non-trivial.
For the equity investor in Japanese regional banks, this is the balance sheet question that the TSE governance reform conversation tends to obscure. A bank trading at 0.6 times book value may appear cheap. It may also be cheap because the book value overstates the economic value of the foreign credit portfolio sitting inside it.
The forensic discipline that private credit analysis requires — looking through the reported marks to the economic reality underneath — is the same discipline that credit practitioners apply to any balance sheet. It is the same discipline that this platform applies to Japanese, Korean, and Indian equities. The tools transfer.
Conclusion
The comparison between private credit today and subprime in 2006 to 2007 is not a prediction. It is a structural observation from someone who managed institutional capital through the period and watched the sequence unfold.
The similarities are real: rapid asset class growth, compressed underwriting standards, valuation opacity, structural liquidity mismatches in certain vehicles, early public signals being characterised as idiosyncratic rather than systemic, and retail capital in places where it probably should not be.
The differences are also real: the leverage amplification mechanisms are less extreme, the regulatory framework is more alert, and the Fed has more room to respond than it did in 2006.
Whether the outcome is a managed correction or a more disruptive dislocation depends on factors that are not yet determined. What is determined is the structure of the risk. Credit practitioners who lived through 2007 and 2008 are not being paranoid when they find the current sequence familiar. They are applying the most important lesson that markets teach: that the shape of a credit cycle is more consistent than the instruments that define it.
The instrument changes. The people involved change. The cycle does not.