Yesterday, I pondered our debt situation, and asked:

And the bad signals aren't just to the federal debt market-the flight to quality is ultimately going to push things like mortgage rates down too. Would the people urging the government to take on as much debt as possible also urge our homeowners to once again leverage themselves as far as the banks will allow?

This caused Karl Smith to become shrill:


Ok, so yes you should leverage your house to the maximum the bank will allow. You should always leverage to the maxim your counterparty will allow.

If you don't get this then you don't get the concept of "Other People's Money" which is fundamentally superior to your own money, because it belongs to someone else. If you loose it , they are screwed. They could try to screw you in return but it is always harder to re-screw than to screw.

If there is screwing to be done you want to be the first one to screw, not the last. This is applicable in many contexts, but especially in finance.

This does not necessarily mean that you should consume all of the resources you extracted from your counterparty. You may want to stash them somewhere safe but under your control not theirs. Again, its always better if you have more and everyone else has less. If you don't get this, you are really missing how the world works.

Meanwhile, at Marginal Revolution, commenter Alkali said:

There are a lot of random superlatives here ("as much debt as possible," "as far as the banks will allow"), but if I were in the market to buy a house, and mortgage rates dropped but house prices didn't, I would certainly consider buying a bigger and more expensive house. It would be crazy not to. The problems with homeowner leverage in the recent housing bubble did not arise because people borrowed at interest rates that were too high.

This stretches the question farther than I meant it.

To start with, people seem to be comparing US debt to an actual mortgage, and US borrowing to prepayment.  But US treasury debt is not structured like the kind of mortgage that people like Alkali seem to be imagining: a 30-year fixed self-amortizing loan.   If you want to compare treasury debt to a mortgage, think about it as an ARM, or a 7-year balloon mortgage.  In other words, when borrowing, you can't just think about interest rates now.  You have to think about what interest rates are going to be in the future--unless you're planning to pay off all that debt (i.e. run a sizeable surplus) in the fairly near term.

I don't think I need remind you that we are not planning to run a sizeable surplus in the near term.  Or for that matter, the far term.

Then there are all the people who treat the problem as if spending (consuming the resources) was exogenous to the borrowing.  Karl, for example, suggests that it is economically rational, if morally reprehensible, to borrow as much as possible to stick others with the loss. But of course, the US government does not have any very good place to stash extra resources.  It spends them.  So this is not a question about who to leave holding the bag for a given level of spending--though as a side note, given that much of the money will be borrowed from Americans, the net gain from stiffing creditors is smaller than it would be for a household, unless that household was raiding the kid's college fund.

Would we be better off borrowing as much as possible, spending as much as possible, and then defaulting when it became too onerous to repay.  No, for reasons I've outlined elsewhere:  suddenly balancing a massive budget deficit involves extremely ugly pain, much of it endured by the most vulnerable members of society who structured their lives around government programs which have suddenly vanished.

A related argument is that we should borrow the money and invest it.  But again, the US government does not have a lot of opportunities for revenue-enhancing investment--it's far from clear that the best candidate, infrastructure spending, will yield enough revenue to cover its capital costs, and the spending opportunities go rapidly downhill from there in terms of cash flow.  Yes, there may be benefits to the citizenry, but we cannot pay our bondholders in shorter commutes, lower crime, and enhanced carbon efficiency.  Maybe these things filter through to GDP, but GDP growth is surprisingly consistent over long periods; getting a tenth of a percent over the long term would be a big achievement.  At that rate, it takes a long time to generate enough revenue to cover our debts.

Then there are the people who advocate not spending the money at all, but refinancing our existing debt.  This doesn't make any sense to me. As I mentioned, the bond marked does not work like your home loan.  You have the option to repay your home in full whenever you feel like it, and many people do this by borrowing the money elsewhere and paying off the loan.

But bonds are securities, traded in an open market.  And the rule of securities is that when bond yields go down, bond prices go up.

To see how this works, take a simple example: a 30-year bond worth $1000 that pays 5% a year.  The government raises $1,000 and mails you a check for $50 every year (the coupon); at the end of 30 years, it gives you $1,000 back.

Now say that yields fall to 4%.  Unless the bond is callable, the government is going to have to go out onto the open market and buy back that bond.  How much is that bond selling for?  $1250--i.e, the sum for which $50 a year is a 4% yield.

This might be a good deal if the government is using extra cash to buy the bond.  But if it's simply raising the money in the open market, it's going to have to borrow $1250 at 4% in order to buy back the bond worth $1250. In other words, there's no way to save money on your securitized debt by refinancing, unless you have a cheaper source of capital.  We don't.

(Yes, yes, this is grossly oversimplified and excludes all sorts of things, including transaction costs, but it suffices to make the point.)

I am not actually arguing against stimulus, incidentally.  I am arguing against people who reason backwards from low interest rates to the conclusion that our current level of borrowing cannot possibly be dangerous.  Borrowing to buy a bigger house in 2007 was dangerous behavior, even though people with few better investment options were flooding our markets with capital--dangerous to the borrower, and dangerous to the economy.  The banks should not have let them do it simply because they didn't have any better place to invest the money.  And the borrowers should not have assumed that simply because the banks were willing, it was therefore a good idea to go ahead and borrow.

We are not supposed to worry about the bond vigilantes, we are told, until we actually see them on the horizon.  My worry is that the bond vigilantes may not politely mosey along, giving us ample time to get the hell out of Dodge.  My worry is that when they appear on the horizon, they will be galloping at full speed and shooting at us, and we will not have time to rectify the situation before they shoot us.

Now, I am not 100% worried about this, the way I am about, say, growing old and dying.  Maybe they won't come.  Maybe our current level of borrowing is not dangerous.  But I would like that argument made based upon conservative macroeconomic projections, and some realistic assessment of both the American and Chinese political economy, not the fact that in the immediate aftermath of a nuclear disaster in Japan, frightened people are seeking to park their cash in US government assets.

I think you can still make an argument for stimulus (and I'll save my rather complicated opinions on that question for another day).  You can maybe even make an argument for slightly bigger stimulus on the grounds that the cost of capital has fallen, lowering your carrying costs.  But when people express worries about the debt burden this will impose, I don't think it's a very good rejoinder to point out that interest rates are low--any more than I would be comforted if an overstretched homeowner told me that it must be okay, because hey, he just got an increase in his home equity line of credit.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.