The case for GDP bonds

August 3, 2009

Around the world, governments are struggling to drum up buyers for the mountain of bonds they need to sell. And that’s especially true for big deficit, low savings countries like Britain and the United States.

The returns they are offering on conventional government bonds are low and there’s the risk of inflation eating away at their value. Perhaps it is time for a different approach.

Rather than hiring investment banks to bamboozle the public into subscribing for unwanted conventional bonds, or cramming them down the throats of the banks as part of expanded reserve requirements, governments should consider issuing bonds linked to gross domestic product.

GDP bonds are not a new idea but few governments have issued them. Those that have don’t make for a distinguished roll call, including as they do Bosnia Herzegovina, Bulgaria and Argentina, which swapped conventional bondholders into GDP bonds after its currency collapse in 2001.

Instead of paying a fixed return, or one purely adjusted for inflation, the interest payments on these bonds are linked to a country’s nominal GDP. Imagine a country that normally grew its GDP at 5 percent a year. To the extent it grew at 6 percent, the coupon would be reset upwards by one percentage point. If it undershot by one percentage point, the interest rate would be similarly reduced.

Most investors have three objectives: long-term growth; inflation protection and low price volatility. Conventional bonds offer a fixed return but no inflation protection. Equities offer growth and the possibility of a higher return but at the cost of volatility. Index-linked bonds offer inflation protection but no growth. GDP bonds meet all three requirements.

Moreover, they are an ideal investment for those saving for retirement. The objective of a pension fund (whether or not the benefits it offers are defined) is to meet liabilities that grow in line with predicted growth in earnings. Since GDP is by definition equivalent to gross domestic income, GDP is a good proxy for this liability.

For the issuer, GDP bonds also have appeal. As Willem Buiter has recently observed, they would give government debt some of the characteristics of an equity security. Their servicing costs would rise and fall in line with the state’s own ability to pay. This, Buiter observes, “would reduce the expansion of the public debt through the intrinsic debt dynamics that comes out of the product of the interest rate and the outstanding stock of debt.”

Of course there are problems. The biggest is that GDP figures can be gerrymandered and are subject to retrospective revision. There are no easy answers. There would need to be agreement on the methodology for, and timing of, any revisions. Put and call mechanisms could be inserted to protect against substantial changes.

GDP bonds might check stock market bubbles. Their existence, former fund manager turned economist Paul Woolley has argued, should act as a reality check for equity investors and offer a substitute asset, should equity valuations get wildly out of whack. Another argument in their favour is that they are almost ideal buy and hold investments and as such should require very little active management — keeping management costs to a minimum.

This may be why few investment bankers, however many governments employ to sell their debt, would advocate issuing GDP bonds. That must be an argument in their favour.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

“breaking windows fallacy”

not only are gdp figures subject to retrospective revision.. it is also the most manipulated, subjective and archaic statisitc ever devised.

besides many loose assumptions associated with the gdp figures, the methods of calcultion are also flawed.

Here I will outline why “GDP” figure is a failure. I suggest all the readers and the authors to follow up on these points.

GDP is a failure – why??
problems measuring living standards
failure to account for deficits
failure to account for savings and debt
failure to adjust for net output
failure to account for resource depletion
failure to adjust gdp for real inflation
failure to adjust for annual population growth

You could say the cash for clunkers program raises gdp by 2-3 billion… the question is..

does subsidizing consumers to destroy perfectly good running used cars to take on even more debt to finance a new car really equal economic growth? how about hiring millions of federal workers to dig holes in the texas desert for no apparent reason.. does that raise GDP?

tisk tisk.. gdp bonds lol.. thats the biggest scam ive ever heard since the federal reserve

Posted by dvictr | Report as abusive

Investment banks that give bad advice don’t get mandates by governments or debt agencies in the first place. If the fees for GDP-linked bonds are attractive then investment bankers would offer them as an alternative.

The problem is that governments want to issue debt as cheaply as possible. Otherwise when fixed interest rates were low they would try to issue as long as possible to lock-in low interest rates. This reduces their refinancing risk.

However, as investors know that governments control fiscal policy, and the ability to issue debt, while central banks control monetary policy, and therefore the keys to future inflation or currency depreciation, they are not likely to buy long-dated fixed coupon bonds that lack inflation fighting credentials.

Nor are they likely to buy GDP-linked deals at a discount to where they could buy comparable short-dated bonds. The imbedded equity call in faster GDP growth comes at the expense of writing a GDP put for the investor. Ergo, if growth turns out to be lower than expected, the GDP-linked bond holder takes a forced haircut. As a straight bond holder there is no imbedded put option. Just the risk of inflation, over-supply and/or currency depreciation.

When will journalists finally learn that there is no free lunch in finance? If governments covet a triple-A rating and low funding costs then the surest way to achieve that aim is to run balanced budgets, pay down existing debt and keep real interest rates inflation neutral. Then lock-in low interest rates with long-dated debt issuance to reduce any re-financing risk. Its not sexy, but it works, and following Argentina is not innovative, but simply dumb.

Posted by MrBill, Eurasia | Report as abusive

I am not sure dvictr fully understands how a GDP bond would work. Essentially the GDP bit would provide an equity component, but at its base the issuing state would still be borrowing in the markets at a rate that reflected investors’ assessment of its creditworthiness (so deficits etc). So if the US could borrow at say 4 per cent, the GDP kicker would only be activated to the extent GDP exceded or undershot an agreed “normal” rate of growth. So if the “normal” rate were 3% and GDP then grew at 4%, the interest rate could increase by a commensurate amount (i.e. from 4 to 5%). Moreover, because the bond would be based on nominal GDP, it would be impossible to resort to inflation to erode it and cheat savers. Inflation would simply inflate nominal GDP, forcing the borrower government to pay more.

As for MrBill, I bow to his expert knowledge that GDP bonds would not offer enough in the way of fees to interest an investment banker. It is good to see the myth of disinterested advice blown away with such gusto.

I would however take issue with the idea that I am suggesting that there is any “free lunch” here. Nothing could be further from the truth. No one wins big with a GDP bond. Rather, there is a reasonable allocation of the risk.

Posted by Jonathan Ford | Report as abusive

i regret the last segment of my commment; “tisk tisk gdp bond” and any unintentional negative connotation the tone of my reply may have alluded too.

gdp is a bad indicator period, way to subjective and prone to moral hazard to be used as a benchmark in any transparent/efficient financial instrument. the revision problem is already insurmountable.

the real issue is defict spending. i agree with mrbill where he recommends “balanced budgets, pay down existing debt”

Posted by dvictr | Report as abusive

[...] The case for GDP bonds. (Jonathan Ford) [...]

[...] rate of increase of GDP is vastly easier than measuring GDP itself. Bonds like the ones pushed by Jonathan Ford are at least linked to the GDP growth rate; Shiller’s bonds, by contrast, are linked to [...]

[...] are signs that Argentina might be heavily affected by outside factors. One of the measures are the GDP bonds, known as cupones PBI in Spanish. GDP bonds pay interests according to the country’s nominal GDP. [...]

[...] Jonathan Ford: The Case for GDP Bonds [...]

[...] was a bad idea when Jonathan Ford proposed it in August 2009, it was a bad idea when Shiller wrote about it in December of that year, [...]