BIS Quarterly Review, September 2018 - media briefing

BIS speech  | 
23 September 2018

Claudio Borio's remarks | Hyun Song Shin's remarks

Claudio Borio

Divergence is the name of the game. During the review period, while US financial markets powered ahead, emerging markets faced mounting pressure. One could say that, on average, global financial markets continued to do well, extending the previous vintage year. But the average was not particularly meaningful. It was a bit like that proverbial person whose temperature, on average, was fine, except that their head was on fire and their feet freezing.

Why this divergence? Odd as it may appear, the reasons are not hard to understand. Despite the maturity of this expansion, the US economy, if anything, speeded up further. Yet inflation, as it tended to converge towards target, posed no threat. Under these conditions, the Fed could afford to proceed along its very predictable and gradual path to normalisation. For better or for worse, nothing jolted US markets out of their comfort zone. Against this favourable backdrop, the US dollar, which had finally turned in February and gathered speed in April, continued its ascent. Inevitably, emerging market economies (EMEs) felt the pain. After all, as noted before, and examined again more closely in the current issue, US dollar lending to non-bank EME residents has actually more than doubled since the Great Financial Crisis (GFC), to some $3.7 trillion. And these figures do not include any dollar borrowing through FX swaps, which could easily be of a similar order of magnitude. As the Highlights document, this surge has been part of a broader expansion international credit to non-banks, a key indicator of global liquidity, which rose from 33% to 38% of global GDP between the first quarters of 2015 and 2018. Within that total, bank loans lost ground to securities, which are held largely by non-bank investors. 

Looking more closely, we see that, remarkably, US financial conditions continued to be very easy from a long-term perspective; in fact, if anything, they eased further. Equity prices reached new historical peaks, buoyed in particular by the technology sector and share repurchases. Volatility stayed low. Term premia remained highly compressed. In particular, while edging up slightly, corporate credit spreads were unusually compressed - with high-yield ones even hovering around the levels that prevailed just ahead of the GFC. As discussed in a box, and in line with the low spreads, the leveraged loan market was red-hot. Volumes ran high, as banks off-loaded their loans onto an eager investor base. Some were off-loaded via collateralised loan obligations (CLOs), close cousins of the infamous mortgage-based collateralised debt obligations (CDOs) of GFC fame. Covenants weakened further and prices often had to be adjusted upwards to clear the market following preliminary auctions. The allure of variable rate loans proved irresistible, not least as investors looked to further increases in bond yields and shunned potential capital losses - a prospect no doubt reinforced by the major fiscal stimulus so late in the expansion. The search for yield proceeded unabated. The one sign of investors' wariness or doubt was a relatively high, and rising, price of insurance against an equity market drop.

Worrisome as these developments are, it was EMEs that took centre stage. The initial rumblings, first heard as the US dollar started to appreciate earlier in the year, became considerably louder. Crises erupted in Argentina and Turkey - two countries weakened by long-standing domestic vulnerabilities. Other economies, primarily those with larger current account deficits, smaller FX reserve cushions and higher inflation, saw their currencies fall considerably, especially if the political backdrop was unfavourable. After an unprecedented 16 consecutive months of inflows into EME investment funds, the tide suddenly reversed in May. Since late March, the cumulative valuation losses on many EME asset classes and currency exposures have been substantial. In some respects, they have exceeded those during the taper tantrum in 2013 and have come close to those in the wake of the renminbi depreciation in August 2015. That said, sovereign spreads have yet to revisit their previous highs on either local currency or dollar bonds.

The EME turbulence reflected the confluence of three forces. One, as already mentioned, was the combination of US monetary policy tightening and US dollar appreciation. The second was the escalation of trade tensions. While in advanced economies asset prices quickly recovered whenever adverse news hit the headlines, their EME peers failed to bounce back. The third, not entirely unrelated, force was the re-emergence of signs that the Chinese economy might be weakening. The country has become critically important for commodity producers and EMEs - in fact, for the world at large. Add in country-specific vulnerabilities and the resulting pattern is not hard to explain.

Fortunately, the EME turbulence did not spread widely. Contagion remained limited, investors largely continued to discriminate; and there were no stampedes among portfolio managers. Even so, not all advanced economies got off scot-free. The repercussions reached as far as the euro area, notably its weaker periphery (see box). As Turkey wobbled, risk aversion heightened and global investors pulled back, the share prices of the banks most exposed to the country took a knock. In addition, some sovereign spreads widened. Italy suffered most: political uncertainties had already shaken markets in May and banks there remain burdened with non-performing loans. Moreover, if anything, the sovereign-bank nexus, so much in evidence during the euro area debt crisis, has grown tighter since then.

If we take a further step back, the bout of volatility engulfing EMEs should not come as a surprise. As noted in the BIS Annual Economic Report, these developments are symptoms of a broader malaise. The highly unbalanced post-GFC recovery has overburdened central banks. The powerful medicine of unusually and persistently low interest rates has served to boost economic activity, but some side effects were inevitable. The financial vulnerabilities that we now see are, to some extent, one such example. The market ructions are akin to a patient's withdrawal symptoms.

What happens next is, as always, hard to tell. Will the patient continue to mend, as looked likely until the first quarter of this year, or will there be a relapse? What one can say is that the patient's full recovery will not be smooth. On the financial side, things look rather fragile. Markets in advanced economies are still overstretched and financial conditions still too easy. Above all, there is too much debt around: in relation to GDP, globally, overall (private and public) debt is now considerably higher than pre-crisis. Ironically, too much debt was at the heart of the crisis, and now we have more of it - although, fortunately, banks have reduced their leverage thanks to financial reform. With interest rates still unusually low and central banks' balance sheets still bloated as never before, there is little left in the medicine chest to nurse the patient back to health or care for him in case of a relapse. Moreover, the political and social backlash against globalisation and multilateralism adds to the fever.

Policymakers and market participants should brace themselves for a lengthy and eventful convalescence.

Hyun Song Shin

Cryptocurrencies do not serve as money, and they raise concerns regarding fraud and market manipulation as well as money laundering and illicit activities. But what should be the policy response?

Cryptocurrencies are often thought to operate beyond the reach of individual national authorities, but the evidence suggests that prices, transaction volumes and user bases around the world react strongly to news about possible interventions by individual national authorities. In "Regulating cryptocurrencies: assessing market reactions", Raphael Auer and Stijn Claessens examine the evidence.

The authors classify news about possible interventions into a number of categories. Interventions that affect their legal status (ranging from outright bans to pronouncements that cryptocurrencies fall under securities law) have the greatest adverse impact on cryptocurrency prices. The negative effect is large and persists over time.

A second type of news relates to guidance on the treatment of cryptocurrencies under anti-money laundering and terrorist financing regulations. These interventions have also had a significant and lasting negative impact on prices, though not as large as the first category.

In contrast, general warnings about cryptocurrencies, including about the risk of loss, have had little discernible effect on prices. In this sense, moral suasion alone does not seem to work in stemming speculative markets.

These findings suggest that, far from being beyond the reach of individual national authorities due to their cross-border nature, cryptocurrencies are susceptible to effective interventions by individual jurisdictions. The findings may reflect how cryptocurrencies piggyback on the conventional financial system, and especially how they are intertwined with the bank-based payment system. No doubt a globally coordinated policy response would be more effective, but the absence of such coordination need not be an impediment to effective intervention. National authorities are not powerless to act if they so choose.

In "Fintech credit markets around the world: size, drivers and policy issues", Stijn Claessens, Jon Frost, Grant Turner and Feng Zhu take stock of the development in non-bank lending intermediated through online financial technology platforms. Such "fintech credit" has grown rapidly, notably in China, the United States and the United Kingdom (which are the top three fintech credit countries, in that order), although China has seen a marked slowdown in the past year or so.

The platforms and business models differ, but most operate by screening borrowers and matching them with investors - the lenders. The size of an economy's fintech credit market is positively related to its income level, and negatively related to the competitiveness of its banking industry and the stringency of its banking regulation. The negative relationship with regulatory stringency may reflect common elements in the stringency of regulations applied to banks as well as fintech firms. It does not, however, support the commonly heard argument that fintech firms spring up as a response to tougher banking regulation.

As fintech grows, it could bring benefits in terms of greater financial inclusion. But there are accompanying risks to do with the erosion of lending standards and the greater procyclicality of credit, as recent operational failures and conduct problems in some jurisdictions highlight. Since fintech credit has not been tested over a full economic cycle, we need to learn more, including how best to address the challenges for regulators in ensuring adequate consumer and investor protection and overall financial stability.

In "The rise of zombie firms: causes and consequences", Ryan Banerjee and Boris Hofmann examine the prevalence, causes and consequences of unprofitable firms that still manage to survive. The term "zombie" has become fashionable to denote such firms, but the authors refine the definition to focus the question more sharply. They examine firms whose interest payments exceed their profit, while at the same time having a share price that is lagging their sector median, as measured by the ratio of market equity to book equity (Tobin's q).

Of the 32,000 firms in their sample, from 14 advanced economies, roughly 6% fall into the zombie category under this definition as of 2016. The authors find that the prevalence of zombie firms has ratcheted up since the late 1980s, tending to rise in recessions but not fully falling back to previous levels during recoveries. Among the possible explanations, the authors find an important role for reduced financial pressure to meet debt payments arising from longer periods of lower interest rates. They also show that a greater prevalence of zombie firms is associated with lower overall corporate productivity.

In "Term premia: models and some stylised facts", Benjamin Cohen, Peter Hördahl and Dora Xia provide a review of so-called "term structure models" of the bond market that aim to separate the expectations and term premium components of bond yields. Applied to 10-year government yields in the United States and the euro area in recent years, the models produce different estimates for the levels of the term premia, but broadly agree on the trends and dynamics. US term premia have fallen since the GFC, possibly reflecting reduced short rate uncertainty (thanks to forward guidance and the zero lower bound) and central bank asset purchase programmes.

Interestingly, the rise in US yields since late 2016 appears to be mostly about expected real rates, not changes in term premia or expected inflation. Real risk premia have even fallen somewhat in 2016-18. Euro area yields have also fallen since the GFC, with similar possible causes.