Vulnerabilities in the international monetary and financial system

Speech by Mr Claudio Borio, Head of the Monetary and Economic Department of the BIS, at the OECD-G20 High Level Policy Seminar, Paris, 11 September 2019.

I am delighted to be here, and I would like to congratulate the organisers for finalising the revision of the code.

This session is about the possible vulnerabilities that capital flows can generate in new - or perhaps less illuminated - corners of the financial system. In addressing this question, I would like to take a broad view. The reason is that we cannot fully understand what the next vulnerabilities are and how best to address them without understanding what is wrong with the international monetary and financial system - or, as the late Tommaso Padoa-Schioppa used to say, "non-system".

Thus, in my remarks, I will briefly address three questions. What is the Achilles heel of the current international monetary and financial system (IMFS)? Where might the next pressure points or corners to watch be? And how can we address the Achilles heel? 

1.   The Achilles heel of the IMFS

There is a long intellectual tradition arguing that the IMFS's Achilles heel is its proclivity to generate large current account imbalances. Think, for instance, of Keynes' concern with asymmetric adjustments between debtors and creditors or of Bernanke's savings glut hypothesis.

Certainly, large current account imbalances can be a problem, not least because they give rise to protectionist pressures, as we painfully see today. But to my mind, the real Achilles heel has to do not so much with net capital flows, which is what current accounts represent, but with gross capital flows and the accumulated stocks.¹ It is here where the underlying weakness of the IMFS is most apparent.

That weakness is the failure to anchor effectively financial expansions and contractions, not just across borders but also within countries - cycles that can cause major macroeconomic dislocations and financial crises.

Such financial expansions and contractions take two forms.

First, expansions and contractions in gross capital flows and correlated asset prices, of which Hélène Rey's "global financial cycle" is the most famous incarnation. As recent BIS research indicates, the length of this cycle roughly coincides with that of business cycles.²

Second, their close cousin, "domestic financial cycles". These are best captured by strong joint expansions and contractions in domestic credit and, in particular, property prices. These cycles tend to be considerably longer than business cycles (sometimes twice as long) and to cause costly recessions and possibly outright financial crises, such as the recent Great Financial Crisis (GFC).

Importantly, while distinct, these two cycles interact and can amplify each other, especially around financial crises.³

Now, the IMFS fails to provide sufficiently strong anchors because of the way it shapes the interaction of domestic monetary and financial regimes.

On the one hand, domestic monetary regimes pay little attention to the build-up of financial imbalances, so that the easing bias spreads from the core economies to the rest of the world. It does so directly, through the extensive reach of international currencies - especially the dollar - beyond national borders. I will come back to this. And it does so indirectly, through policymakers' resistance to exchange rate appreciation, be these the result of concerns with price stability, financial stability or other considerations. That is, central banks keep interest rates lower than otherwise and/or intervene and accumulate foreign exchange reserves. This way, easing begets easing across the world.

On the other hand, the interaction of financial regimes reinforces and channels these effects through the free mobility of capital across both currencies and borders. As a result, external funding typically amplifies domestic credit booms and exchange rates move too far ("overshoot"). Hence the relevance of the concept of "global liquidity", ie the ease of global financing, which has been the focus of BIS research.⁴

What is the evidence for all this? For one, the surges in international credit ahead of episodes of serious financial stress. Think, for instance, of the GFC, of the Latin American crises in the 1980s, of the Asian crisis in the 1990s and of many similar episodes going further back in history, including under the gold standard - the first globalisation era. In addition, both pre- and post-GFC, we have seen historically accommodative monetary policy conditions, which have been a key source of unwelcome spillovers and which could encourage a further build-up in debt globally. 

2.   Corners to watch: the dollar and the asset management industry

So much for the big picture. Let me now focus briefly on two pressure points - or two corners to watch. One is old: the role of foreign exchange funding, and in particular that of the US dollar, to which I have already alluded. The other is newer: market-based finance and the role of asset managers in particular.

We know that many countries, notably emerging market economies (EMEs), have long relied on foreign exchange funding. And we know that, at the global level, the dollar reigns supreme. But unlike in most advanced economies, in emerging market economies such borrowing often gives rise to currency mismatches, mainly because of costly and sometimes limited hedging opportunities.

This, in turn, gives rise to the so-called "financial channel of the exchange rate". Currency appreciation flatters balance sheets and encourages further borrowing and appreciation - a self-reinforcing process. The opposite is naturally the case when the currency depreciates. All this blunts the shock absorber function of the exchange rate and makes EMEs especially vulnerable to boom-bust cycles in global capital flows - an aspect we developed in a special chapter of the latest BIS Annual Economic Report on monetary policy frameworks in EMEs.⁵

Against this background, the sharp increase in US dollar borrowing post-crisis is troubling. The corresponding debt has roughly doubled for EMEs' non-bank borrowers, to some $3.7 trillion. And this, let me stress, does not include borrowing via FX swaps, which is not covered in the statistics as it is off-balance sheet. That borrowing, according to our estimates, is even larger. Including also advanced economies, the equivalent figure to the on-balance sheet borrowing for EMEs is $11.8 trillion while the one for FX swaps is of the order of $14 trillion or more.⁶

It would not be surprising, therefore, if the next episode of financial stress had the US dollar segment at its epicentre, just as it was during the GFC.

In all this, the asset management industry is likely to play a substantially bigger role than in the past. The development of domestic currency markets was expected to insulate EMEs from the currency mismatches linked to borrowing in foreign exchange. The evidence so far suggests that, while clearly helping, it has fallen somewhat short of expectations. But, in any case, this would not solve the problem of currency mismatches in lenders' (investors') balance sheets. Whenever they invest on a foreign exchange unhedged basis to pick up extra return, investors would lose twice on their long-term bond holdings if the exchange rate depreciated as the domestic currency yields rose - a typical correlation, especially under stress. This would exacerbate the pressure to liquidate portfolios.

The rapid growth of market-based borrowing post-crisis, as many banks have retrenched, highlights these vulnerabilities.⁷ The share of bank loans in US dollar borrowing by non-banks outside the United States has fallen from around 60% to a bit below 50%. The bottom line: although banks are safer, we should watch closely the vulnerabilities linked to market-based finance.

3.   Policy implications

What are the policy implications? Let me just highlight two.

First, we need stronger anchors in domestic policy regimes - ensuring that one's own house is in order. At the BIS, we have argued for the need to put in place macro-financial stability frameworks, in which monetary policy, prudential (especially macroprudential) policies and also fiscal policy pay greater attention to the build-up of financial imbalances. In the BIS Annual Economic Report chapter on EMEs I mentioned before, we explain key aspects of this framework, notably the role of foreign exchange intervention in the spirit of a macroprudential tool alongside the active use of macroprudential measures. Here, I would stress, in particular, measures targeting currency mismatches and liquidity mismatches in foreign exchange. All this would reduce the scope for unwelcome spillovers - a natural outcome if domestic policies are already fully fit for purpose.

Second, we need stronger anchors internationally - ensuring that the global village is in order. Here, the degree of progress has differed. It has been substantial in prudential regulation and supervision, through international standards and a degree of reciprocity. Still, some gaps remain. In particular, we need to think harder about how to address the systemic risks that the asset management industry raises. A key strategy here would be to calibrate regulation from a macroprudential perspective, ie considering institutions not just on a standalone basis but explicitly as part of the system, in analogy with what has been done for banks. By contrast, progress has been very limited in monetary policy, for structural reasons. Here the key question is whether it is possible to go beyond enlightened self-interest. Finally, strengthening the global safety net would be an important complementary step, which would help contain stress when it cannot be prevented. Here, the need for funding in US dollars is critical. One question is whether it would be as forthcoming as during the GFC in the current political environment.