“I heard that Bank of East Asia was in trouble this morning but wasn’t so worried until I saw the queue. Something must have gone wrong.” ((Justine Lau and Tom Mitchell, “Malicious rumours spark run on Bank of East Asia” Financial Times September 25, 2008, page 1.))
These words express the essence of any financial crisis: the loss of confidence based on incomplete information and fearful inference. I spent much of this past week in Hong Kong and was treated to a mini-case example of financial crisis in the run that hit Bank of East Asia (BEA) here on Wednesday, the 24th. This was the first run-on-the-bank in Hong Kong since 1997. This week’s story of BEA illustrates the origins of a panic and what it takes to stop one.
Here’s a sketch of the case. Bank of East Asia is Hong Kong’s fifth-biggest bank by assets, mainly serving a regional clientele. It issued a quarterly earnings report in early September that showed a profitable and moderately-growing firm. Then in mid-month, two events occurred that challenged the image of stability. Turmoil in the American financial markets—especially the failures of Lehman Brothers and AIG—spawned concerns about the exposure of Hong Kong Banks to the collateralized debt obligations of the American firms. ((It turned out that BEA had outstanding exposures of HK$422.8 million to Lehman Brothers and HK$49.9 million to AIG.)) Also in mid-month, BEA issued a revision to its earnings report revealing that a rogue trader had accumulated losses of HK$93 million in equity derivatives that would reduce the latest quarter’s financial results.
On Monday, September 22nd, rumors began to circulate regarding the solvency of BEA. These rumors gathered momentum on Tuesday in a slew of broadcast text messages from an anonymous source. One message said, “There is a run on the BEA—if you have money there, go get it quickly.” ((Phyllis Tsang, “Panic sparked by electronic whispers,” South China Morning Post, September 25, 2008, page A3.)) Depositors showed up in large numbers on Tuesday afternoon at BEA’s branches. They expressed panic: “I just want the peace of mind when I go to sleep tonight. Nothing is safer than hard cash in my pocket…You know, this is a crazy time, banks are falling over…It’s not normal what is happening in the world. Who knows if what they are saying is true?…I just want my money. Everyone is queuing and so should I.” ((Barclay Crawford, Peter So, and Ambrose Leung, “Thousands waste hours in bank lines,” South China Morning Post, September 25, 2008, page A3.))
Management of BEA responded vigorously. First, they issued repeated public statements asserting the solvency of the bank and briefed the media on BEA’s financial condition. BEA’s CEO said, “I can confirm categorically these rumors are unfounded.” He shared data that showed BEA’s cash reserves were adequate to meet all withdrawals and capital adequacy ratio is twice as large as the required minimum. Second, they met all withdrawal requests and stayed open past normal closing time to do so. Third, they enlisted the help of the Hong Kong Monetary Authority who issued a statement confirming the solvency of BEA: “BEA’s capital adequacy and liquidity ratios are well above regulatory requirements.” ((“BEA and Monetary Authority’s Response” South China Morning Post September 25, 2008, page A3.)) The Monetary Authority also flooded the Hong Kong banking system with liquidity, nearly HK$3.9 billion. Fourth, local tycoon Li Ka-shing (and BEA’s Chairman of the Board) purchased shares of BEA, not unlike Warren Buffett purchasing a stake in Goldman Sachs, Constellation Energy and so on. Also the bank’s executives cast votes of confidence in the troubled lender by buying stock. Finally the commercial crime bureau of the Hong Kong police department embarked on an investigation of the source of the rumors.
By Thursday, September 25th, the bank run had ended. BEA announced that HK$2 billion in deposits had been withdrawn in the brief run—the withdrawal represented 0.67 percent of the bank’s total deposit base. On Saturday the 27th, the police announced the arrest of an individual who allegedly issued the mass-mailed text messages.
What are we to make of this episode? It illustrates the drivers of financial crisis that Sean Carr and I outlined in our study of the Panic of 1907: complexity, few safety buffers, an economic shock, all of which produce a sharp change for the worse in expectations. The thread that knits these factors together is what economists call an “information asymmetry,” roughly, the situation where a decision-maker doesn’t have all the vital facts and someone else does. In this case, the better-informed people might exploit the poorly informed. ((George Akerlof described this as the “lemons” problem in which the market for used cars features imperfect pricing. Michael Spence extended the insights drawing on the labor market. And Joseph Stiglitz focused on credit markets. Akerlof, Spence, and Stiglitz received the Nobel Prize in Economics in 2001 for their groundbreaking work.)) The poorly-informed people worry about this and begin to behave in ways that produce market panics. In an influential paper, Charles Calomiris, Gary Gorton suggested that bank runs could begin when some depositors observe negative information about the value of bank assets and withdraw their deposits. Then other depositors follow suit, being unable to discriminate perfectly between sound and unsound banks and observing a wave of withdrawals. A run begins. In a world of unequally distributed information, some depositors will find it costly to ascertain the solvency of their banks. Thus, runs might be a rational means of monitoring the performance of banks, a crude means of forcing the banks to reveal to depositors the adequacy of their assets and reserves. Calomiris and Gorton reasoned that if the information asymmetry theory were true, panics would be triggered by real asset shocks that cause the decline in collateral values underpinning bank loans. They found that bank panics originated following economic shocks and that typically the cause was a sudden decline in asset values. In particular, panics tend to follow sharp declines in the stock markets and tended to occur in the spring and the fall. They also reasoned that the resolution of a bank panic would be created by elimination of an important aspect of the information asymmetry: gaining clarity as to which banks were solvent and insolvent would stop the runs on solvent banks.
In contrast, bank runs could be triggered by random events that cause depositors to fear for the safety of their funds. I once heard a story ((Though plausible, this story is probably apocryphal. I and others have searched for the details on such a story and could find nothing. Nonetheless, some long-time residents of Hong Kong nodded knowingly when I related the story to them-perhaps it was a different year and/or a different city. I appeal to any interested readers of this blog for help in finding the facts of this case. Until then, one must take this story as part of the lore or mythology of bank runs rather than established fact.)) that in 1985, a panic hit a Hong Kong bank when a line extended out of the door of a nearby bakery and across the front of the bank. Depositors passing by simply feared for the solvency of the bank and began to withdraw their funds. This is the “stuff happens” theory of bank panics. ((In an influential paper published in 1983, two economists, Douglas Diamond and Phillip Dybvig suggested that bank panics are simply randomly occurring events. Bank runs occur when depositors fear that some kind of shock will force the bank into costly and time-consuming liquidation. To be last in line to withdraw deposited funds exposes the individual to the risk of loss. Therefore, a run is caused simply by the fear of random deposit withdrawals and the risk of being last in line.))
Which theory, information asymmetry or stuff happens, better explains the history of bank panics? Empirical research lends support to the asymmetric information theory. Studies have affirmed the relationship between information asymmetry and panics, by looking at how deposit losses predict panics, the yield spreads between low- and high-risk bonds peak at the panic, and real declines in the stock market are greater in panic years than non-panic years. In the case of Hong Kong’s BEA, the issue of information asymmetry seems to square with the facts.
My only issue with the simple explanation offered by the theory of information asymmetry is that it seems to suggest a very simple solution to panics: flood the market with information. Yet as our study of the Panic of 1907 showed, simply offering evidence of solvency of banks isn’t enough. It takes some serious leadership to organize the intense kind of collective action that it takes to stem the emotional tide of a panic. Virtually every panic contains deeper lessons about shaping a vision, enlisting others, acting with integrity, reading the mood of a crowd, communicating very well, and showing a bias for action–all of these are the attributes we associate with effective leaders.
To quell a panic requires ample liquidity, injection of capital, transparency about the facts of the situation, and vigorous leadership to organize collective action and assert the common good. The financial system in Hong Kong may have been lucky. The system was at risk; rumors were swirling about other institutions as well; banks in nearby Macau experienced some runs. As we are seeing in the U.S. today and saw in 1907, financial panic usually does not end so quickly. This mini-case emphasizes the importance of quick and decisive action in fighting the bank run.
Posted by Robert Bruner at 09/28/2008 04:42:42 AM