By Alasdair Macleod
The issuance of debt is normally subject to a contract that it will be repaid at the end of its term, along with the coupon interest. The exception is undated bonds, when only the interest contract must be fulfilled. In practice, governments and many corporations roll over debt into new bond obligations at the end of their terms, but at least bondholders have the opportunity to be repaid their capital. Therefore, the credibility of government debt is based on the assumption the issuer can afford to continue to roll it over rather than repay it.
However, the rolling over of old debts and the continual addition of new ones will almost certainly become a problem for governments everywhere. It is less of a problem when the debt is put to productive use, but that is rarely, if ever, the case with government finances. To judge whether the rolling over of debt is sustainable and at what cost, we need to rely on other metrics. The traditional method is to compare outstanding debt with GDP, and by using this approach two economists (Carmen Reinhart and Ken Rogoff) came up with a rule of thumb, that once a government's debt-to-GDP ratio exceeded approximately 90%, economic growth becomes progressively impaired.1
The Reinhart-Rogoff paper was empirically based, and loosely impresses upon us that the current situation for the US and other nations with higher debt-to-GDP ratios is unsustainable. Key to this reasoning is that rising debt levels divert savings from financing economic growth, and therefore, a government's ability to service it from rising taxes is undermined. At the Rubicon level of 90% and over, median growth rates in the countries sampled fell by 1%, and their average growth rates by "considerably more." It is entirely logical that a government forced to tax its private sector excessively in order to pay debt interest will restrict economic potential overall.
This analysis was published in the wake of the Lehman crisis, when an unbudgeted acceleration in the rate of increase of government debt everywhere was a pressing concern. The signals from financial markets today indicate that we could be on the verge of a new credit crisis, in which case tax revenues will again fall below existing estimates, and welfare costs rise above them. Therefore, government debt will increase unexpectedly, as was the case that caused the Reinhart-Rogoff paper to be published in 2010.
To look at the increase of government debt between 2007 and 2009 - as Reinhart-Rogoff did - was not, as it turned out, a long enough time frame to fully reflect the consequences of the Lehman crisis on government debt. The increase recorded over 2007-09 was 32%, yet economists and others were still talking of austerity until only recently. The whole period between the Lehman crisis and the election of President Trump is perhaps a better time frame, and we see that US government debt between 2007 and 2016 increased by an astonishing 217%.
It turns out that the Reinhart-Rogoff report severely understated the problem by reporting early. Their 90% debt-to-GDP Rubicon has been left behind anyway, with government debt-to-GDP ratios around the world in excess of 100% becoming common. In the case of the US, total Federal debt, including intragovernmental holdings, is currently over 105% and rising. The Congressional Budget Office is forecasting substantial budget deficits out to 2028, adding an estimated further $4.776 trillion in deficits between fiscal 2019-23 or $9.446 trillion between fiscal 2019-28.2
This assumes there is no credit crisis, so for those of us who know there will be one during the next ten years, these numbers are far too optimistic. Accordingly, we should look at two possible outcomes: first, a best case where price inflation continues to be successfully managed with a target rate of two per cent, and a second base case incorporating an estimate of the effects of the next credit cycle on government finances.
Best- and Base-Case Outcomes
Our best-case outcome of controlled price inflation is essentially that forecast by the Congressional Budget Office. Working from the CBO's own figures, by 2023 we can estimate accumulated debt including intragovernmental holdings will be $26.3 trillion3, including our estimated interest cost totalling $1.3 trillion4.
That is our best case. Now let us assume the more likely outcome, our base case, which is where the effects of a credit cycle play a part. This will lead to a fall in federal government receipts and an increase in total expenditures. Taking the last two cycles (2000-07 and 2007-18), these led to increases in government debt of 59% and 239% respectively. Therefore, it is clear that borrowing has already been accelerating rapidly for a considerable time, due in large measure to the destabilising effect of increasingly violent credit cycles. If the next credit cycle only matches the effects on government finances of the 2007-18 credit cycle, government debt including intragovernmental holdings can be expected to rise to $51.4 trillion by 2028. This compares with the CBO's implied forecast of only $34 trillion of government debt over the same time frame and makes no allowance for the cyclical effect on interest rates. More on interest rates later.
Because the underlying trend is for successive credit cycles to worsen, the $51.4 trillion figure for federal government debt becomes a base figure from which to work. But there are still considerable uncertainties, particularly over the form it will take.
The character of the next credit cycle is unlikely to replicate the last one, which was a sudden financial and systemic shock. Today, the US banking system is better capitalized, and off-balance sheet securitization has been brought largely under control. There are, however, uncertainties concerning the eurzone banking system. There are also risks in global derivatives markets and the potential knock-on effects of counter-party failures on the US banks. Furthermore, there can be little doubt that the sudden systemic shock of Lehman afforded a degree of protection for the purchasing power of the dollar, and therefore, of the other mainstream currencies, despite the unprecedented monetary expansion.
However, it would be complacent to expect an outcome of relatively low price inflation to be simply repeated at a time when government finances are even more dramatically spiraling out of control. Last time the threat was systemic to the banks, but next time the inflationary consequences of government finances is likely to be the dominant problem.
The explosion in the quantity of government debt that our analysis implies has many economic consequences. In the context of our rough analysis, we should comment on the point made in the original Reinhart-Rogoff paper, which is that the reduction in GDP potential that results from an increase in the ratio of government debt-to-GDP is likely to be significant. The growth in Federal debt that replicates the post-Lehman experience will leave the US government with a debt-to-GDP ratio of over 170%. The CBO assumes GDP will increase by 48% by 2028 to $29.803 trillion, whereas our cyclical case is for debt to rise to $51.4 trillion. While both these figures should be taken as purely indicative, clearly, US government debt will increase at a faster pace than the growth in GDP and will strangle economic activity.
If the purchasing power of the dollar declines more rapidly than implied by the CBO's assumed 2% price inflation target, interest payable on Federal debt will in turn be sharply higher than expected, compounding the debt problem. The federal government will face a potentially terminal debt trap from which there can be no escape.
1 Growth in a Time of Debt Carmen Reinhart and Kenneth Rogoff, NBER Working Paper No. 15639, 2010. See here.
2 See Table 2 here.
3 The CBO estimates of debt and debt interest are based on debt in public hands only. They exclude $5.8 trillion of debt held in intragovernmental accounts, which are included in debt ceiling calculations and my estimates.
4 This includes estimated interest on intragovernmental debt, which is not in the CBO's figures, and assumes an increase in average nominal funding costs of 3.25% by 2022, which is a real rate of 1.25% at 2% price inflation. This further assumes the deflator accurately captures the rate of price inflation, which it certainly does not.
Disclosure: No positions.
Editor's Note: The summary bullets for this article were chosen by Seeking Alpha editors.