What risk factors should banks look out for?

The world experienced 147 systemic banking crises from 1976 to 2009 —an average of one every 10 weeks. True, these crises came in waves: in early 80s, mid 90s, and then, after a relative calm of early 2000s, the crisis of 2007-08. However, the sheer number of crises has been unprecedented and speaks volumes about the way banks manage their risks.

Fundamentally, all banking crises are panic-driven exits from bank debt, i.e., the money banks owe to others, or their liabilities. This is almost a truism. The 1866 financial panic in England was triggered by run on Overend & Gurney, a London wholesale discount bank, which used short-term deposits to finance long-term speculative investments in ships and railroads. The 2008 crisis was also characterized by runs on repo, commercial paper, and prime brokerage accounts, which, together, form the plumbing of wholesale funding for banks.

What triggers an exit from bank debt is a loss of confidence in assets backing bank debt. Bank assets are credits extended to consumers, corporations, and governments for them to invest in real assets like factories, houses, cars, or to finance their current consumption. Anything that impacts borrower’s ability to repay has a direct bearing on the value of bank assets, for they are nothing but total repayments, adjusted for the time over which they flow in.

The loss of confidence and resulting crisis often appears precipitous, akin to a cardiac arrest. But as is often the case with cardiac arrest, the lifestyle leading to one spans years. Similarly, the factors leading to a banking crisis seldom develop suddenly; instead they build over time. These macro-financial risks — large-scale economic stresses and imbalances — can gestate for years before precipitating a crisis.

Examples abound: the 2008 crisis was preceded by an 85% increase in real US home prices from 1997 to 2006 along with an unprecedented growth of mortgage credit. Similarly, the origin of the 1991 Swedish banking crisis was the deregulation of financial sector in 1985, which gave strong incentives for companies and households to borrow at low and often negative interest rates creating excessive indebtedness. When real interest rates rose, the value of Swedish banks’ assets declined by almost 30%.

So one of the fundamental things that banks ought to do is to watch out for these tell-tale signs of crises. And this is precisely where they have failed. A natural question to ask is why?

Edifice of modern risk management is built on price indicators that mathematically model risk on market prices of a range of instruments like equities, bonds, insurance against defaults, and commodities. They serve well as long as the aim is to gauge current risks, but have proved to be of limited use as harbingers of banking crisis. The premium to insure against defaults of Bear Stearns and Lehman, for instance, actually fell from 2004 to 2007. Both defaulted in 2008. Northern Rock’s share price was at all-time high in early 2007; in September depositors ran on it. The switch from a benign environment to a hostile one is precipitous, and price-based indicators flash red just before that switch. So while necessary, they are by no means sufficient.

Macro-financial risks, on the other hand, are often ignored by banks.

Before 2007 how many American banks thought of “decline in expenditures on new housing construction” as factor to regulate real estate loans? Very few. However, in the last 60 years any decline of more than 10% was always followed by a recession. Expenditure on new housing construction started their downward spiral in the middle of 2006, ahead of decline in house prices, but it didn’t set alarm ringing for banks to reduce mortgage lending.

How many banks in the United Arab Emirates noticed that overall bank credit had increased by more than 50% in a span of just 15 months from December, 2006, and tried to curb this excess? Probably none. The result: real estate prices crashed by almost half, banks suffered major losses, and Dubai had to restructure $26 billion worth of borrowing in 2009.

Or for that matter, how many French banks lost sleep from 2003 to 2007, when Greece, Ireland, Portugal, and Spain each experienced substantial net financial inflows?

A simple yet powerful way to integrate macro-financial risks is through what bankers call risk appetite. Simply put, risk appetite is the amount of risk a bank is willing to accept; it answers questions like what types of customers and products should the bank target; which countries and industries to avoid; how much collateral to seek; how much deposit to ask for a mortgage; and so on.

Explicitly linking macro-financial indicators to trigger changes in risk appetite can go a long way in preventing buildup of financial vulnerabilities.

And it doesn’t have to be complicated. A handful of measures, for instance, can be used effectively: a big increase in debt-to-GDP ratio (as in China, where total debt stands at 280% of GDP) should trigger tightening of underwriting standards; swings in risk-return relationship, especially when risk is underpriced coupled with unusual increase in transactions (as has been the case for lowly-rated U.S. corporates in recent years) should inform banks to be cautious; and banks exposed to emerging markets should worry about Fed rate rise as Fed policy tightening has often been associated with a wave of emerging market crises, as it happened in 1981-82, in the late 1980s and, to a lesser degree, in mid-1990s. These are but a few examples.

Banks in past have missed the forest for the trees by overly relying on complicated risk models based on market prices. Time to get a few canaries in the coal mine.

The views expressed are solely of the author and not of the National Bank of Abu Dhabi.

To keep up with the Agenda subscribe to our weekly newsletter.

Author: Amit Tyagi is the Vice President and Head of Group Credit Risk and Portfolio Management at the National Bank of Abu Dhabi

Image: A man walks past buildings at the central business district of Singapore. REUTERS/Nicky Loh.