Stay informed with free updates
Simply sign up to the Global Economy myFT Digest — delivered directly to your inbox.
Invest long, borrow short and leverage up as much as possible. That is the way to make money in finance. It is how banks have always made their living. But we also know very well that this story can end in panic-stricken runs for the exit and financial crises. That is what happened in the great financial crisis (GFC) of 2007-09. Since then, as the Bank for International Settlements explains in its latest Annual Economic Report, the financial system has changed a great deal. But this central characteristic has not.
Moreover, notes Hyun Song Shin, economic adviser to the BIS, “despite the fragmentation of the real economy, the monetary and financial system is now more tightly connected than ever”. If this sounds like an accident waiting to happen, you are quite right. Central banks must be prepared to ride to the rescue.
The story the BIS tells is an intriguing one. Thus, the aftermath of the GFC did not make the system fundamentally different. It just changed who was involved. In the run-up to the crisis, the dominant form of lending was to the private sector, particularly in the form of mortgages. Afterwards, lending to the private sector levelled off, while credit to governments exploded. The pandemic accelerated that tendency.
That was not surprising: if people want to save and lend, someone else has to borrow and spend. That is macroeconomics 101. In addition to the change in direction came a change in intermediaries: in place of the big banks have come global portfolio managers. (See charts.)
As a result, cross-border bond holdings have increased enormously. What matters here are changes in gross, not net, holdings. The latter are relevant to long-term sustainability of macroeconomic patterns of saving and spending. The former are more relevant to financial stability, because they drive (and are driven by) changes in financial leverage, notably cross-border leverage. Moreover, notes Shin, “the largest increases in portfolio holdings have been between advanced economies, especially between the US and Europe”. The emerging economies are relatively less involved in this lending.
How then does this new cross-border financial system work? It has two fundamental characteristics: the leading roles of foreign currency swaps and non-bank financial intermediaries.
The biggest part of this cross-border lending consists of the purchase of dollar bonds, particularly US Treasuries. The foreign institutions buying these bonds, such as pension funds, insurance companies and hedge funds, end up with a dollar asset and a domestic currency liability. Currency hedging is essential. The banking sector plays a key role, by enabling the market for foreign exchange swaps, which provide these hedges. Moreover, a forex swap is a “collateralised borrowing operation”. Yet these do not appear on balance sheets.
According to the BIS, outstanding forex swaps (including forwards and currency swaps) reached $111tn at the end of 2024, with forex swaps and forwards accounting for some two-thirds of that amount. This is vastly more than cross-border bank claims ($40tn) and international bonds ($29tn). Moreover, the market’s largest and fastest-growing part consists of contracts with non-dealer institutions. Finally, some 90 per cent of forex swaps have the dollar on one side of the transaction and over three-quarters have a maturity of less than one year.
As the BIS notes, this highly non-transparent set of cross-border funding arrangements also affects the transmission of monetary policy. One of the propositions it makes is that the greater role of non-bank financial intermediaries, notably hedge funds “may have contributed to more correlated financial conditions across countries”. Some of this is quite subtle. Given the large-scale foreign ownership of US bonds, for example, conditions in the owners’ home markets can be transmitted to the US. Again, exchange rate movements that affect the dollar value of holdings of emerging market debts can trigger adjustments in their domestic prices.
What are the risks in this new system of finance? As has been noted, banks are active in the market for forex swaps. They also provide much of the repo financing for hedge funds speculating actively in the bond market. Moreover, according to the BIS, over 70 per cent of the bilateral repo financing from banks is at zero haircut. As a result, lenders have very little control over the leverage of the hedge funds active in these markets. Not least, non-US banks are active in providing dollar funding for firms engaged in these markets.
What does all this imply? Well, we now have tightly integrated financial systems, especially among high-income countries, even as the countries are moving apart, politically and in terms of their trade relations. Moreover, much of the funding is in dollars on relatively short maturities. It is easy to imagine conditions in which funding dries up, perhaps in response to large movements in bond yields or some other shock. As happened in the GFC and the pandemic, the Federal Reserve would have to step in as lender of last resort, both directly and via swap lines to other central banks, notably those in Europe. We assume that the Fed would indeed come to the rescue. But can that be taken for granted, especially after Jay Powell is replaced next year?
The system the BIS elucidates has much of the fragility of traditional banking, but even less transparency. We have a vast number of unregulated businesses taking highly leveraged positions, funded on a short-term basis, to invest in long-term assets whose market values may vary substantially even if their capital values are ultimately safe. This system demands an active lender of last resort and a willingness to sustain deep international co-operation in a crisis. It should work. But will it?
Follow Martin Wolf with myFT and on X