Wed, 09 Mar 2005

Creating growth-linked bonds

Robert J. Shiller, Project Syndicate

A year ago, at the Summit of the Americas, 34 western hemisphere heads of state agreed to promote the creation of government-issued growth-linked bonds whose payout is tied to gross domestic product (GDP). But progress has mostly stalled. Only one major proposal related to such bonds, from Argentina, is on the table. A unique opportunity to strengthen the world's financial infrastructure and increase economic efficiency for many years to come could be slipping away.

I have argued for growth-linked bonds since my 1993 book Macro Markets. GDP is the most comprehensive measure we have of an economy's success. The simplest form of growth-linked bonds would be a long-term government security that pays a regular dividend proportional to the GDP of the issuing country.

Suppose that the Argentine government issued perpetual bonds that paid an annual dividend equal to one ten-billionth of Argentine GDP, payable in pesos. Because Argentina's annual GDP now runs at about 500 billion pesos, one of these bonds today would pay a dividend of 50 pesos (about US$17 or 13 euro) a year. The dividend would rise or fall as the success of Argentina's economy is revealed through time.

The market for GDP-linked bonds would arrive at a price that makes them attractive to investors, reflecting expectations and uncertainties about the issuing country's future. Until there is a market for such bonds, we cannot know what the price will be. But we can expect that the market for long-term GDP-linked bonds from countries like Argentina, where the future of the economy is uncertain, would be volatile, as investors adjust their expectations of future GDP growth up and down in response to new information.

What will happen to Argentina in the next 25 years?

Argentina's long-term GDP growth has been disappointing. In fact, real GDP per capita declined 15 percent over the 25-year period from 1965 to 1990 -- a period that saw some Asian economies quintuple in size.

But the 8 percent real GDP growth recorded in 2004 might encourage some to expect a surge in economic performance, as has occurred elsewhere in the world.

Could there be another decline in Argentina? Or a huge growth breakthrough? Nobody knows.

The economic costs implied by this uncertainty could be reduced if there were a market for growth risk. Indeed, Argentina's economy would be better off today if Argentina had borrowed in terms linked to its GDP decades ago rather than at an interest rate denominated in dollars. Its foreign debt would have declined in line with its GDP, thus sheltering the economy from default and economic disaster. To be sure, investors would have then lost on their bet on Argentina's growth, but they would still have been protected against inflation, even if their bonds had not been denominated in dollars (because Argentina's nominal GDP would have moved up with inflation).

Can we engineer new debt contracts for countries like Argentina?

The stumbling block has been potential investors' fear that accounting fiascos in emerging countries would render the bonds unsafe. If you don't trust the numbers, you can't trust the debt.

Clearly, more work is needed to improve the quality of the numbers. In the meantime, we should not give up on trying to create these bonds. Instead, advanced countries should issue them first.

True, advanced economies are relatively more stable, which means that the bonds would have a less distinctive risk- management advantage. But the demonstration effect would be immediate. Once any major country develops a market for GDP- linked debt, the concept will be established, making it much easier for other countries to join in.

The spread of inflation-indexed bonds serves as a historical precedent. Finland was the first to issue national inflation- indexed bonds, in 1946, in response to massive wartime price growth. Israel and Iceland were next, in 1955, followed by Brazil, Chile, Columbia, Argentina, the UK, Australia, Mexico, Canada, Sweden, New Zealand, the U.S., France, Japan, and Italy. It took a long time, but the contagion of a sound financial idea has been unmistakable. GDP-linked bonds would also allow hedging the risk of inflation, plus respond to GDP growth.

Some will object that there is little point in creating GDP- linked bonds in advanced countries, because there is little uncertainty about GDP growth there. However, even in the relative calm of the post-war era, long-term real GDP growth in advanced countries has been rather variable.

For example, U.S. real per capita GDP grew by a factor of 1.87 -- that is, nearly doubled in size -- from 1961 to 1986, but by a factor of only 1.58 from 1978 to 2003. Such differences in 25- year growth rates are important: if U.S. GDP grows by a factor of 1.87 over the next 25 years, annual GDP will be $3.6 trillion ($10,000 per person) higher than if it grows by a factor of only 1.58.

Much of today's debates about the future of old-age pensions hinges on this uncertainty. If the economy grows rapidly, there will be no problem; if it does not, a pension crisis looms. Creating a marketplace where such uncertainties are traded and hedged would be a fundamental step toward managing the risks involved.

If personal pension accounts or provident funds are invested in GDP-linked bonds, the payments that retirees receive in 25 years will reflect the growth rate of the economy -- and that of the tax base -- to that date, which all makes good sense. Sweden's pension system recently created a link between national income growth and benefits, but the reforms did not include creating GDP-linked bonds, a natural adjunct to such a scheme.

At a time when many advanced countries run government deficits, GDP-linked bonds would improve risk management and bring down financing costs. They might also spur developing countries to do what it takes to join that market.

The writer is Professor of Economics at Yale University, and author of Irrational Exuberance and The New Financial Order: Risk in the 21st Century.