Developing countries have started to use GDP-linked enhancements to their coupon payments.
GDP-linked debt could have provided solid returns in the U.S. context.
The recent positive trends in Greek government debt may be partly due to the GDP enhancement post-2015.
Confidence does not reign supreme on the ability of the Argentine government to honor its debt, whether restructured or holdover.
The Argentine bond market fell today, after news that the government plans to allow investors in government bond to swap their debt for dollar-denominated bonds with payments hitherto made through the central bank, instead of Bank of New York's Mellon Corp., the current trustee.
One of the live experiments conducted by the Argentine government during the 2005 debt restructuring was the issuance of GDP index-linked debt, essentially a bet by investors on future Argentine growth. When these bonds were originally issued, they had the following payout formula. A GDP-linked warrant is attached to every post-2005 restructured Argentine bonds. The following conditions must be met simultaneously in every period between 2005 and 2036:
- Real GDP growth must be higher than a base level.
- Real GDP growth versus the previous year must be greater than the growth implied by base GDP (after 2015, assumed to be flat growth rate of 3%)
- A payment cap which will not exceed $.48 per unit of currency of the warrant.
If these three conditions are simultaneously met, the government will pay 5% of the difference between actual GDP growth and the base growth. So far, so good. However, recent developments have put paid to the notion that adding on these warrants have helped. Here's a graph of recent movements in the price of 30-year government bonds which are GDP index-linked. Clearly markets have not expressed great confidence in these bonds recently.
It was not always thus. When the Argentine economy was in the growth phase in 2009-2010, the returns on the GDP warrants which are attached to the government bonds was a juicy 24%. In 2009, the warrant holders were rewarded with a $3.11 payment, when the economy grew in 2008 by 6.8%.
What are the might-have-beens for U.S. trills?
Argentina issued a GDP warrant attached to a long-term government bond. But let us recall what a trill is: a bond, possibly perpetual, whose coupon is indexed to the GDP level, not inflation, as are TIPS. The idea has been advanced by Robert Shiller as a stabilizing influence on the budget and as an interesting portfolio diversification strategy. These bonds would also be of interest to pension funds whose benefits are linked to wages as they accrue, and to inflation.
Trills would also be useful in financial planning, since younger investors would be less likely to invest in trills, as their income would be more tied up with wage income, and hence, economy-wide risks. Older people would be more attracted to the diversification potential offered for trills, since their income is less and less influenced by wage income. An investor in trills would benefit from a growing economy, just as his tax liabilities might be increasing. It is curious why this idea has never really taken off.
Suppose trills had been issued way back in 1947. The bond coupons would pay one-trillionth of the past quarter's GDP level (constant dollar). The coupons would have been paid in perpetuity. Under this hypothesis, here is the graph for the price of such a security. Using these calculations, it is worth noting that trills would have returned on average 6.7% over the entire post-war period, 7.9% since the 1960s to the present, and 10% since the 1990s!
Source: Bureau of Economic Analysis
The Greek surprise?
The Argentine government issued detachable GDP-linked debt warrants. The Greek government, for its part, tried to take part in the trill experiment by issuing a GDP-linked feature to their 2012 debt swap, which provided investors with a basket of long-term bonds, coupons gradually increasing with maturity for maturities between 2023-2042. The interesting feature of these bonds was that the coupons were enhanced by a "GDP-link", with a coupon growth formula which depends on a) the past year's GDP growth, b) the past year's inflation rate, c) a reference GDP growth rate and d) a cap on the GDP enhancement. The GDP enhancement feature is scheduled to kick in for the coupons after 2015. Overall, the GDP enhancement to these bonds is minor; nevertheless, it is interesting to examine the recent performance of long-term Greek bonds.
Source: Bank of Greece
After the original swap in 2012, the strong initial returns have tapered; nevertheless, it is worth noting that the average return over the period was 66%. According to Trading Economics, GDP growth in Greece, currently at -.2%, is anticipated to become positive in 2015 at .15%, and significantly higher at 2.75% by 2020.
GDP-linked bonds seem like a fail-safe way in betting on a country's upside, with little downside risk if they are designed in the appropriate way. The advantages for governments that issue sovereign debt are obvious. The question is whether the very governments that would benefit from these GDP-linked enhancements are the kind to enter into successive default crises such that the GDP enhancement becomes irrelevant. It seems that Argentina and Greece provide opposing answers to this question.