Light needs shedding on the dollar swaps black hole
Does it matter if you lose track of $80tn? That is a question that global regulators urgently need to ponder, according to the Bank for International Settlements.
The BIS’s economists have recently delved into the entrails of the bank’s foreign exchange survey to track the trend in dollar swaps — the derivatives deals that enable investors to exchange dollars for other currencies, for a specific period, before swapping these back.
A layperson might assume this is a transparent corner of finance, given that the dollar underpins so much of the global industry and these swaps are used by most large companies and investment groups.
Not so. As the latest BIS quarterly report notes, forward dollar payment obligations “do not appear on balance sheets and are missing in standard debt statistics”, because they are not classified as a “loan” in accounting systems. They are off-balance sheet — in other words, a black hole.
And that hole is astonishingly big. The BIS detective work suggests that there are “$80tn-plus” in outstanding obligations to pay US dollars in FX swaps — or forwards — and currency swaps. This has doubled in the past decade, with a $5tn churn each day.
Some $39tn of this involve non-American banks, and another $26tn non-American, non-bank entities such as hedge funds and insurance companies. Moreover, the non-bank category has grown rapidly, dwarfing on-balance sheet risks, and is often associated with maturity mismatches.
Thankfully, there is no sign of current stress. But history shows that FX swaps markets are “vulnerable to funding squeezes”, the BIS notes. This is because when shocks hit, investors worry about how to repay these dollar obligations, creating wild price gyrations.
And while the Federal Reserve has quelled panic in the past by letting other central banks supply dollars, the black hole means that the dollar swaps market was “backstopped in 2008 and 2020 by central banks acting on little information about who owed the debt”. The BIS is now (quite sensibly) begging central bankers to plug the data gap.
I hope the Fed and others will heed the call. Regulators have already created a data hub for globally systemic banks, which tracks some of these flows. But policymakers should not stop there. One dirty secret of global finance is that this $80tn black hole is far from being the only place where we need more sunlight.
The sphere of private credit is also startling murky. So is the behaviour of the Chinese credit cycle; as a new paper from the Fed notes. This matters because Chinese credit conditions have big spillover implications for global asset markets, including the dollar-based one.
But another area of surprising data fog is Treasuries which (like dollar swaps) underpin so much of the financial system. During the dramatic market turmoil of 2020, it became clear that secondary market trading structures have big vulnerabilities that were not understood — or reported — before. And while the Securities and Exchange Commission and US Treasury is trying to fix this, progress is slow.
The primary market, meanwhile, has its own transparency problems, as the economists Alexandra M. Tabova and Francis E. Warnock have recently argued. Public data shows overall bids in bond auctions and the broad patterns of foreign purchases. But it does not reveal which buyers are buying which bonds, or their price sensitivity.
This matters, they argue, given the US government’s $18tn negative foreign asset position. It needs to know “which investors will step in and buy Treasuries” against the looming spectre of higher Treasury yields while at the same time, the Federal Reserve is reducing its Treasury portfolio.
Tabova and Warnock have scoured private data sources and concluded that while non-Americans own around half of all Treasuries, their behaviour is poorly understood. To cite one example: while official statistics claim that foreign private investors were buying $3.5tn more Treasuries each year than foreign governments between 2005 and 2019, they think the latter actually bought $1tn more than the former. That is important, since foreign government purchases have recently slowed, and they have acted in a strikingly “price inelastic way” before, with much shorter portfolio durations than private investors.
The good news is that the digitisation of private data is making it easier than ever to do this type of detective work. Better still, after the 2008 global financial crisis, the US Treasury created the Office of Financial Research, which is supposed to take a holistic view of markets.
But the bad news is that OFR has faced calls to cut its budget, and it cannot hope to get better data on dollar swaps or Treasuries without help from central banks.
So let us hope that the American government (among others) treats the revelations about this $80tn black hole as a wake-up call. After all, if there was ever a moment when central bankers needed to understand the real vulnerabilities arising from dollar-linked debt, it is when the world faces geopolitical risks, economic pain and political challenges to the supremacy of the dollar. That is precisely what 2023 could bring.