How to Induce Explosive Debt Dynamics

by Menzie Chinn,

The debt ceiling and implications of:

“We Republicans need to be willing to tolerate a temporary, partial government shutdown.”

and more recently

“I think it is possible that we would shut down the government to make sure President Obama understands that we’re serious.”

It is one of the oddities of current discourse in macroeconomic policy that there are concerns about the short term sustainability of the Federal government’s abilities to finance debt, despite the fact that the on the run ten year TIPS (maturing 1/15/2022) is -0.760% (1/11/13). I do believe there is reason for concern in the long term. However, should we encounter difficulties in raising the debt ceiling, the timetable could be greatly accelerated. To see this, consider the following expression:

(1)     it10y = (it + Etit+1 + … + Etit+9)/10 + lpt10y + rpt10y

The long term nominal interest rate is the average of expected short term interest rates (the expectations hypothesis of the term structure), plus the term premium associated with liquidity effects (lp), and a premium associated with sovereign risk (rp). In general, for US Treasuries, we suppress the last term. But it is the last term that is relevant in the current debate; if the failure to resolve the debt ceiling manifests itself in a higher rp, then debt dynamics can become more problematic. To see this, consider this expression.

(2)     dt-dt-1 = [(rt-gt)/(1+gt)]× dt-1 – pt

Where d is the debt to GDP ratio, r is the real (inflation adjusted) interest rate, g is the growth rate of real GDP, and p is the primary (noninterest) surplus to GDP ratio. In words, the debt to GDP ratio rises when the real interest rate-growth rate gap is sufficiently positive, or the primary deficit is sufficiently large. Notice that if the risk premium associated with a debt ceiling impasse rises, then ceteris paribus, the real interest rate rises, thereby accelerating the pace of debt accumulation. Moreover, if the debt crisis raises uncertainty that depresses growth, then that too accelerates the pace of debt accumulation (measured as a share of GDP). The logical conclusion: in ostensibly seeking to avert a debt crisis, certain parties could increase the likelihood of a debt crisis, or at least move it forward in time.

Let’s now examine the various channels by which a refusal to raise the debt ceiling could trigger an acceleration toward a debt crisis.

How Likely Are We to Default, and When?

The Bipartisan Policy Center has compiled data to make a guesstimate of the date at which the Treasury’s extraordinary measures will be exhausted. The resulting calculations are depicted below:

Figure 1: Bipartisan Policy Center, Debt Limit Analysis (January 2013).

Now what happens once extraordinary measures are exhausted? Then expenditures will be limited to daily revenue and cash on hand. Once those funds are exhausted, the Treasury according to BPC will default on financial obligations. The call for partial government shutdown, say paying our soldiers but not vendors, does not appear feasible.

First, the legal precedents for such an option –- that is to prioritize spending –- are unknown. [1], pp.8-9 Second, it is unclear whether it is technically feasible to do so, given Treasury’s computer systems (it is conceivable that debt payments could be prioritized over nondebt payments [2]). Hence, a partial shutdown, sparing essential services such as defense (or air traffic control, for those of us who fly a lot), would seem to face some obstacles. Here is BPC’s tabulation of the situation that would confront the USG on 2/15/2013:

Figure 2: Bipartisan Policy Center, Debt Limit Analysis (January 2013).

The astute observer will note that on this particular day, the $30 billion outflow for interest payments alone exceeds the $9 billion in revenue. Hence, regardless of the technical feasibility of prioritizing payments (it is unclear how the policymaker quoted above would like payments to be prioritized), on a day like this day, then, the USG would seem to be technically in default on US debt (not paying other creditors would also be default, just not a debt default). This is the logical implication of not raising the debt ceiling.

What Would Be the Domestic Financial Implications?

Since the US government has never defaulted, it is unclear what would happen. We have experience from the last time the intransigence drove us to the brink of default. From the Government Accountability Office’s (GAO’s) assessment of the 2011 event:

“Our results indicate that the 2011 debt limit event period led to a premium (which is represented by a decrease in the yield spread) ranging from 11 to 33 basis points on Treasury securities with maturities of 2 or more years. For 3-month and 6-month Treasury bills and cash management bills, which typically had a maturity of 56 days, the debt limit event period led to a 1 basis point decline in Treasury yields relative to private security yields during the period (which is represented by an increase in the yield spread), while there was no change in yields on 1-month and 1-year Treasury bills relative to private security yields. Overall, Treasury yields increased relative to comparable-maturity private securities during the 2011 debt limit event period.”

The GAO’s assessment of the public finance implications of this event are as follows:

On the basis of our analysis, we estimated that delays in raising the debt limit in 2011 led to an increase in Treasury’s borrowing costs of about $1.3 billion in fiscal year 2011. We derived this estimate by multiplying the amount of Treasury securities issued at each maturity during the event period by regression-based estimates of the relevant yield spread change attributable to the debt limit event and weighting the result by the portion of fiscal year 2011 during which the security was outstanding. Many of the Treasury securities issued during the 2011 debt limit event will remain outstanding for years to come. Accordingly, the multiyear increase in borrowing costs arising from the event is greater than the additional borrowing costs during fiscal year 2011 alone.

Macro Implications for the US

GAO indicates that interest rates would rise in the wake of an event where we come close to defaulting. What happens if the USG stops paying vendors is unknown, although it would seem reasonable to conclude the impact on yields would be noticeably higher than that experienced in 2011 (so rp rises in equation (1)). Ceteris paribus, r in equation (2) rises, accelerating the rate of debt-to-GDP increases.

To the extent that rates are higher, it would not be surprising that other asset prices (including stock prices) decline precipitously. Uncertainty itself would depress stock prices as well, if it led to higher equity risk premia (I think that a highly uncontroversial conjecture). The WSJ documents the movements in the SP500 and VIX in 2011.

Figure 3: Source: Paletta and Hilsenrath, “Ugly Choices Loom over Debt Ceiling,” WSJ, 14 January 2013.

Further, many commentators have argued that uncertainty depresses output, as discussed previouslyBaker, Bloom and Davis (2013) have attempted to measure economic policy uncertainty and the impacts on economic output. Figure 4 depicts the evolution of the Baker-Bloom-Davis index:

Figure 4: Economic Policy Uncertainty index (new version). NBER defined recession dates shaded gray. Source: Baker, Bloom and Davis, at Policy Uncertainty, accessed 1/12/2013.

Notice the spike at 11/08; that spike is associated with going to, but not over, the precipice. Baker, Bloom and Davis (2013) estimate that a 112 point innovation in the index leads to a 2.3% reduction in GDP after 4 quarters (peak response). The jump in uncertainty associated with the previous debt ceiling impasse was about 120 basis points (bps) (which is not quite the same as an “innovation”, but I’m pretty certain the difference is inessential in this calculation). Suppose for instance we have a default which pushes the index up 240 bps, so a back of the envelope calculation, assuming a similar elevation of uncertainty, suggests a 4.6% reduction in GDP after 4 quarters.

(I find it one of those supreme ironies that those who argue uncertainty impedes growth are often the ones that hold positions that would increase uncertainty.)

Now, consider equation (2). Notice that g enters into the calculations. The slower GDP growth, the faster the rate debt-to-GDP rises. Hence, a debt ceiling crisis which is not resolved, or is acrimoniously resolved with only minimal impact on the deficit, might very well exacerbate debt issues.

Isn’t Default Inevitable? Why Not Get It over With?

The oddity of the ongoing debate is the pervasive misconception that the US is fast becoming Greece. In addition to the obvious fact that the US borrows in dollars (i.e., in its own currency), and conducts its own monetary policy, the amount of fiscal restraint (spending cuts and tax revenue increases) is not sufficiently large so as to resort to desperate measures now. As the Center for Budget and Policy Priorities has observed, $1.4 trillion is sufficient to stabilize the debt-to-GDP ratio over the next ten years ($1.2 trillion reduction in the primary budget deficit).

Figure 5: Source: CBPP.

The question policymakers need to ask themselves is whether it is necessary to chance a self-inflicted wound in order to try to balance the budget via spending cuts.

See also Klein and CFR for discussion.

Global Financial Implications

Note that the previously cited financial impacts are based upon cases where we have come close to, but not actually, defaulted. If one were to contemplate what would actually happen, my answer would be “who knows?”.

That is because a default on US debt has ramifications above and beyond those that would occur for a default on debt issued by other sovereigns. Treasurys by far account for the largest single market for sovereign debt (in a convertible currency). A large number of other asset prices are linked to Treasurys either directly or indirectly (e.g., as benchmarks in derivative prices). As Steven Englander from Citi has noted (via Wiesenthal/BI):

So it is possible that we will get a technical default for a few days, but more likely that Congress will give in, vote the debt ceiling up temporarily, and let the automatic sequesters kick in. Mounting risk of a technical default was USD positive in 2011 because it led to cutting of long-risk positions and the USD/Treasury market remained safe havens. However, it also occurred in an environment of slowing EM growth and intensifying euro zone sovereign risk pressure, so the USD support came from external forces as well. Given that investors are now somewhat long risk again, the position cutting is again likely to be USD positive, however, unattractive US assets were. As was the case in 2011, it is very unlikely that the Treasury will not pay its bills, although even a technical default could have very unforeseen consequences, given the multiple functions that Treasuries play in global financial markets. The more likely scenario of sequester plus grudging debt ceiling rise is USD negative. It will put more pressure on the Fed to keep pumping liquidity into the US economy without giving any reassurance to investors that long-term fiscal issues are close to resolution. [emphasis added – MDC]

My worry is that the consequences of throwing sand (or bolders) into the wheels of global finance are insufficiently comprehended in by some of our policymakers in key veto positions within the legislative branch. Those who have faith that actors in the financial system can mitigate any damage arising from technical default by means of careful positioning and hedging should remember back to September 2008, when it was believed the financial institutions were prepared for the bankruptcy of a relatively small investment bank.

Read More Articles by Menzie Chinn

The Choice: Math vs. Anti-Math in Fiscal Policy

The October Employment Situation: Continued Improvement

[iframe src=”” width=”600″ height=”530″ frameborder=”0″ scrolling=”no”]