Risk, Cost, and Obligations

The responses to yesterday's post about America's fiscal future reveal what I think are a lot of misconceptions about the way financial symptoms work:

Cost doesn't equal risk  One of my commenters complained that I was using costs and risk interchangeably, and that they are not the same thing.  Indeed, they aren't always.  But new obligations always add more risk to your balance sheet, even if you think you're also adding the revenue to cover them, because if the revenue doesn't materialize, you're in a dreadful hole.  And the greater the existing deficit between revenues and obligations, the more risk any new obligations add.  If our deficit were in rough balance, the new obligations would add only trivial risk.  But we're running a roughly $700 billion structural deficit over the next ten years, which is greater than or equal to 6% of GDP. 

The US won't actually go bankrupt, because we can always inflate away the value of our debt  This won't work.  We have an independent central bank, and any concerted effort to appoint a Fed chair willing to inflate away our debt would be made difficult by both the dirth of pro-hyperinflation economists/bankers, and the fact that markets would freak out and jam up our interest rates sky high.  Moreover, inflating your debt only works if your structural deficit is small relative to the outstanding amount of debt.  As things now stand, by 2019 our structural deficit is scheduled to be about 6% of GDP, while our debt will be around 65% of GDP.  That may not sound like much, but over a few years, rising interest rates could rapidly confound any attempt at inflating away our debt.

The US won't actually go bankrupt, because sovereign nations don't go bankrupt  This is sort of true.  But sovereign nations can and do become insolvent when the size of their obligations gets too out-of-proportion to the size of the revenues they take in. That's when it becomes hard to borrow money, and states have to start making bankruptcy-like choices--picking which of their obligations they're going to renege upon.  With our entitlement system, that probably means cutting benefits too people who did long-term planning around them.

The US won't actually go bankrupt, because the government will act long before then  This is quite possibly true.  But what does "acting" look like?  It looks like the things we're frightened of doing now:  making deep cuts to entitlements, including the new entitlement we're so busy enacting.  And it probably won't be done until things are pretty dire, which means it will be done hamfistedly.  California is what a sovereign almost-bankruptcy looks like.  New York is the slightly more respectable version.  Both are ugly, and getting uglier.

Everything's okay because the TIPS spread looks fine  The spread between standard and inflation-indexed treasuries is a good measure of our fiscal health if you think the biggest risk is inflation.  If you don't--for the reasons I outline above--then it's not very helpful.  Long term US debt isn't particularly cheap when you factor in things like our low rates of inflation and projected economic growth. 

Everything's okay because Treasury yields are low  This is a decent argument, but only decent.  Short term treasury yields are low; long term treasury yields are not, as I just pointed out, all that fabulous.  We've been borrowing a lot of money at the short end of the curve, and we may get a nasty surprise when we try to convert that into longer maturities.  Moreover, since the financial crisis, we've been in an odd spot:  demand for treasury debt is high, because demand for riskier securities--equity, emerging market debt, and so forth--has fallen abnormally.  (The same thing happened in the wake of the Great Crash).  Over the next five years, treasury yields may stay low, in which case I'm wrong.  Or they may start to rise, in which case we have problems, because we've got to roll over a whole bunch of short term debt.