What good are risk disclosures anyway?

Large banks release thousands upon thousands of pages of information about their risk exposures in shareholder reports and periodic regulatory filings. Scouring these disclosures, as Risk Quantum is wont to do, reveals little about the threat posed by Archegos Capital Management. The family office of former New York hedge fund manager Bill Hwang, which blew up in late March, may inflict losses of up to $10 billion on its roster of bank counterparties, according to JP Morgan analysts. 

Archegos itself is out of reach of US regulators. As a family office, it is exempt from most disclosure requirements. That’s a problem in itself, as it impedes analysts’ and watchdogs’ ability to pinpoint the source of certain highly leveraged positions. The UBS Global Family Office 2020 report says 87 of the largest family offices have assets totalling $142.4 billion – a huge blind spot for regulators.

Banks, meanwhile, are subject to a wide range of disclosure requirements. Still, the risk reports of those caught up in the Archegos implosion – Credit Suisse, Nomura, Goldman Sachs, Morgan Stanley and Deutsche Bank being chief among them – shed little light on their exposures to highly leveraged clients.

Yes, it’s possible to track their aggregate derivatives holdings. Among US banks, for instance, regulatory filings show that Goldman Sachs and Morgan Stanley had the largest portfolios of equity swaps – the instruments at the center of the Archegos calamity. But the counterparty credit risk (CCR) posed by derivatives isn’t broken down, making it hard to point to one portfolio over another as especially high risk. 

Goldman Sachs, for instance, does not break out CCR exposures related to its trading activities from those linked to traditional lending in its Pillar 3 report. Neither does it segment these by probability of default (PD), like JP Morgan, Citi, and Bank of America. Furthermore, the subsidiary that houses its prime brokerage operations – Goldman Sachs & Co LLC – does not file separate risk reports.

Credit Suisse, in contrast, does break down its CCR by PD – but the data it discloses gives no clue that it had billions in swaps exposure to a highly leveraged investor. As of end-December, the bank said 90% of its CCR exposure to financial institutions covered by its internal risk model – $13.8 billion worth – had a PD of less than 0.15%. Given the size of its projected Archegos loss, it seems highly likely that this position was classified in this 0.15% bucket. Using just the Pillar 3 data, then, it would be impossible to predict that a huge blow-up was likely in the near future.

It may be too much to ask banks to produce more granular risk reports. Concerns that doing so would inadvertently give away the identity of clients and other business secrets are certainly valid. But the Archegos debacle begs two questions. One, are banks accurately filling in their Pillar 3 disclosures? And two, if these disclosures essentially gloss over risks that result in billions of dollars in losses, what good are they anyway?