Could an Earthquake Cause the Next Financial Crisis?

On January 17, 1995, an earthquake struck the Japanese port city of Kobe, killing 6,500 people and destroying $100 billion worth of property and capital. Five weeks later, Barings Bank was declared insolvent after admitting losses of $1.3 billion, twice the amount of its capital. Was there any link between the two events?

Barings had no exposure to Japanese borrowers. Its losses were generated by the large portfolio of derivative securities managed by a single trader, Nick Leeson. Though he had been losing money for years, Leeson managed to conceal his true results from internal auditors and supervisors.

In early 1995 he placed his last big bet on the Nikkei 225 – the main Japanese stock market index. Leeson expected the Nikkei to remain stable over the following weeks. Had he been right, the bet would have paid off nicely, perhaps even offsetting all his previous losses. It is at this point that the earth began to shake in Kobe…

Uncertainty about the extent of the damages caused by the earthquake and its impact on the Japanese economy frightened investors, who panicked. Volatility of the Japanese stock market sharply increased. It was a brief spike, but enough to destroy Leeson’s hopes. On February 23, he revealed the amount of his losses in a memo to his superiors.

Three days later Barings was declared insolvent; it was later sold to ING, the Dutch financial group, for the grand sum of £1.

Perhaps surprisingly, Barings’ demise was the only lasting economic consequence of the Kobe earthquake. By the end of 1996, in fact, manufacturing output in Greater Kobe was back at 98% of its pre-earthquake trend, all department stores and 79% of the city shops had reopened, and import trade through the port was fully recovered while export was only 15% below its 1994 level according to George Horwich in a CIBER Working Paper for Purdue University from 1997.

Also, Barings’ collapse was an isolated incident. In today’s parlance, the bank was not “systemically important.” Its insolvency did not materially affect the wider financial sector or the “real” economy. But AIG and Lehman Brothers have proved that some banks and insurers are indeed big enough to cause mayhem if they fail. So one may ask, could another Kobe earthquake bring down the financial markets, and the real sector with them?

The Panic of 1907

Before dismissing the question as ridiculous, you may want to consider an intriguing paper by Kerry Odell and Marc Weidenmeier, two US economists. Odell and Weidenmeier trace the origins of the financial crisis that hit the US in 1907 to the San Francisco earthquake of the previous year.

The Panic of 1907 started with a sharp drop in share prices (the “Rich Man’s Panic”), but quickly reached beyond the financial sector. In the second half of the year U.S. industrial production fell by 30%, while Gross National Product (GNP), a wider measure of economic activity, declined by 6.7% in real terms. This in turn hit lenders. In October, New York’s Knickerbocker Trust Company collapsed.

A series of bank runs followed, until Federal aid and private sector intervention personally led by financier J.P. Morgan managed to restore an uneasy calm. Liquidity remained tight, however, pushing up interest rates. And higher interest rates attracted gold from other money centers, notably London, squeezing liquidity abroad. As a result, U.K. GNP fell 0.9% below its trend level. So the crisis took on an international dimension as well.

Willis Re Analytics Review: 2013 Issue 1

Download our latest ReView for more articles by Willis Re Analytics

International gold flows also played a crucial role at the beginning of the Panic. In 1906, the Bank of England and other European central banks repeatedly raised interest rates to stem consistent outflows of gold to the U.S. In addition, drastic restrictions on the discounting of American finance bills were imposed in London, Berlin and Paris.

All this had a dramatic impact on liquidity available to New York banks, thus setting the stage for the Panic. But what caused the initial gold outflows from Europe to the U.S.?

Odell and Weidenmeier show that most of this gold was needed to pay claims for the earthquake that hit San Francisco in April 1906. The earthquake caused insured losses of $235 million, close to 1% of U.S. GNP in that year. A significant share of these claims were paid by U.K. insurers, who had a strong presence in California at that time, but German and French companies also suffered major losses.

All these insurers had invested part of their reserves in American securities, but were reluctant to liquidate them because they thought U.S. financial markets were too thin to absorb that volume of sales – hence the decision to pay claims with domestic funds and bank borrowings. Since California policyholders preferred to be paid in gold, this meant actually shipping bullion across the Atlantic, which lowered money supply in the insurers’ home countries.

Could it happen again? Not exactly, but don’t forget about it

The story told by Odell and Weidenmeier could not repeat itself today for several reasons. First, gold bullion is no longer used as a means of payment. Second, nowhere is money supply limited by the amount of gold held by central banks, whose mandate includes adjusting liquidity on a daily basis to maintain orderly conditions in the money market.

Third, the role of foreign insurers in the U.S. is quite limited today as insurance cover is mostly provided by domestic carriers. Last but not the least, U.S. financial markets are the most liquid and efficient in the world and could easily absorb the volume of sales required after an event such as the 1906 earthquake.

Over the last two years, the average trading volume on the U.S. bond market has been $8-9 billion per day, i.e. 40 times the total insured losses caused by the San Francisco earthquake. A fairer comparison uses the size of the economy as measured by Gross Domestic Product (GDP) as a benchmark; the results are equally dramatic.

The average daily trading volume of U.S. bonds is now close to 6% of the country’s GDP; insured losses from the San Francisco earthquake were less than 1% of U.S. GNP in 1906 (GNP is not the same as GDP, but they were probably not very different at that time).

But what if a bigger earthquake struck California or New Madrid tomorrow?

For the sake of argument, let’s consider a catastrophic event causing insured losses equal to 2% of the U.S. GDP, more than twice the 1906 San Francisco earthquake. Such a mega-catastrophe would still generate “only” $300 billion of claims, less than half the amount of bonds traded on the U.S. fixed income market on a typical day. No wonder President Clinton’s adviser James Carville wished he could reincarnate into the bond market (“You can intimidate everybody”).

Would the result be different if a giant catastrophe event (a hurricane, say, or a tsunami) happened in another part of the world? Odell and Weidenmeier showed how a localized disaster damaged the international economy. Today this might perhaps happen through the reinsurance sector.

A significant part of catastrophe exposures in peak zones is usually transferred abroad to specialized reinsurers, many of them European. But again, a quick look at the numbers demonstrates that the risk is remote. Total assets held by reinsurers included in the IAIS’ Global Insurance Report amounted to $772 billion at end-2011.

The amount outstanding on the global bond market at the same date was close to $100 trillion. As long as each reinsurer is prudently managed and can cover its reserves with investment assets of good quality, even mega-catastrophe events appear manageable when it comes to their impact on the financial sector.

So where do our back-of-the-envelope calculations leave us? We can draw three conclusions. First, the world has become a safer place with the demise of gold as the foundation of the international monetary system. Second, liquid and efficient financial markets are crucial for (re)insurance. Third, (re)insurers need assets of good quality to invest in.

Most government bonds used to be treated as risk-free, but the Eurozone crisis on the one hand and American politicians’ inability to sort out budget and deficit issues on the other have shaken investors’ trust.

This, in turn, has raised questions on the creditworthiness of many other bond issuers. At the same time, insured property continues to increase at a steady pace – especially in regions exposed to catastrophe perils, such as Asian coastal areas. Surely the next 1907 isn’t slowly approaching?

About Giorgio Brida

Giorgio works in Willis Re’s Strategic Capital and Result Management (SCRM) team in London, providing analytical …
Categories: Analytics, Financial Services, Reinsurance | Tags: , , ,

Leave a Reply

Your email address will not be published. Required fields are marked *