The debt-inflation myth, debunked by UBS | FT Alphaville

The debt-inflation myth, debunked by UBS

Here’s an interesting counterpoint to the theory that governments are attempting to inflate their way out of their financial crisis-related debt dilemmas.

It comes courtesy of UBS economist Paul Donovan, who argues:

While most investors today acknowledge that deflation is likely to be a feature for the OECD economies during the second half of 2009, inflation pessimists cling resolutely to the belief that inflation will inevitably return. “Fiscal deficits are rising dramatically” goes the argument. “Governments will have to create inflation to reduce debt:GDP ratios, as they have done in the past.”

The problem with the idea of governments inflating their way out of a debt burden is that it does not work. Absent episodes of hyper-inflation, it is a strategy that has never worked. Government debt: GDP burdens tend to be positively correlated with inflation. Market mythology has created the idea that inflation will help reduce government debt ratios. The facts do not support the myth. OECD government debt rises as inflation rises. Meaningful reductions in government debt will require a low inflation future.

To wit — this chart, which purportedly shows year-on-year levels of inflation on the x-axis and change in government debt (as a per cent of GDP) over one year on the y-axis. The two axes cross each other at the 5 per cent inflation level because that’s deemed, by UBS, to constitute a genuine inflation shock.

UBS chart of inflation vs change in government debt (as % of GDP)

The point then, is that there aren’t many instances in the lower right-hand quadrant, which would coincide with relatively high levels of inflation and a decline in government debt as a proportion of GDP. The picture is much the same, according to UBS, over a two-year period too.

But why? Here’s Donovan again:

The fundamental obstacle to governments eroding their debt through inflation is the duration of the government debt portfolio. If all outstanding debt had ten years before it matured, then governments could inflate their way out of the debt burden. Inflation would ravage bond holders, and governments (with no need to roll over existing debt for a decade) could create inflation with impunity, secure in the knowledge that existing bond holders could do nothing to punish them. In the real world, of course, governments roll over their debt on a very frequent basis. As a result, governments are vulnerable to higher debt service costs if market interest rates change. If markets move to price in the consequence of higher inflation by raising nominal interest rates, then the debt service cost will rise and increase the debt. Thus a period of high inflation will tend to raise both the numerator and the denominator of the debt:GDP ratio.

As an example, the US can expect to roll over almost 45 per cent of its debt in the next 12 months and some 55 per cent over the course of the next two years. So according to UBS, if there is an inflation surge in the next 12 months, the US government would expect that to be reflected in higher borrowing costs — thus negating any ‘sympathetic’ inflation-impact on its national debt. There’s also the issue of TIPS, or index-linked (inflation-linked) securities.

Add to that the notion that real yields tend to also rise in times of uncertain inflation — by as much as 100 to 150bp, according to UBS’s analysis, as investors demand a premium for the uncertainty — which further adds to government borrowing costs.

Thus, according to UBS, the problem governments face is that high inflation is likely to generate higher nominal and higher real interest rates. This means the rate of increase in debt servicing costs will probably exceed the rate of increase in nominal GDP, as a result of the higher inflation, and voila — you have very little government benefit associated with stronger inflation, according to the bank.

Donovan’s conclusion:

The idea that governments can readily inflate their way out of their debt problems is a misnomer — arising, perhaps, from confusion between the fate of the individual bondholder and the response of the collective market. An individual holder of a long duration bond will lose out as a result of inflation. However, modern governments can not rely on markets to remain collectively indifferent to inflation. Inflation will raise the nominal cost of borrowing (of course) but through the inflation uncertainty risk premium it will also add to the real cost of borrowing.

The higher debt service cost becomes a problem for a government that is pursuing an inflation strategy because government debt does have to be rolled over. Unless a government is willing to pursue hyper-inflation as a strategy, raising inflation will not reduce the government debt burden. Indeed, history indicates that the reverse result will be achieved.

So caveat hyperinflation. Not sure that will appease the uber-inflation hawks, really.

Related links:
Sticky inflation, redux – FT Alphaville
Inflationistas, Deflationistas and Goldilockeans – FT Alphaville
JPM says mild inflation good for stocks too – FT Alphaville