The debt trap
Long-time observers of the Japanese economy are beginning to worry publicly that the country is slowly falling into a government debt trap. The last time I recall this sort of talk was in the late Eighties, when commentators worried that pre-Euro Italy was in the same bad shape.
What is a government debt trap? The general idea is that the government debt situation spirals out of control as the cost of servicing government borrowings rises dramatically, through some combination of high interest rates and the volume of government debt outstanding.
The easiest way to explain the phenomenon is to give an illustration. I’m going to keep it very simple. Governments have become very crafty at keeping items like the cost of a war, or of senior citizen benefits, or explicit or implicit guarantees for failing “government-sponsored” enterprises. But I’m going to ignore that.
I’m also not going to have the country in my illustration grow over the years. Naturally, if GDP–and therefore tax receipts–grow, but government expenditures remain flat (fat chance!), then the situation I’m about to describe may gradually improve. If, on the other hand, government expenditures grow at the same rate as tax receipts, the situation becomes worse.
Also, in theory at least, governments run deficits in downturns and surpluses in upturns, so that over a business cycle they are breakeven. I’m going to ignore this reality as well. (This is kind of scary. It’s like I’m turning into an academic! I hope it wears off soon after this post.) Take the “years” I’m talking about as being an average of what happens over a business cycle. Bear with me through the initial part. I’ll inject a dose of realism at the end.
Let’s make the following assumptions about a country:
annual GDP 100
tax receipts 30
govt expenditures 35
govt debt 0
interest rate on govt debt 5%.
This situation is relative benign. The government runs a primary deficit (tax receipts minus government expenditures, before debt service) of 5. Interest expense on that debt is .25, so the overall deficit is 5.25. Interest expense makes little difference to the overall deficit. The deficit is also structural, not cyclical; that is to say, it won’t go away as the business cycle develops.
Let’s look at something more interesting.
Everything is the same as in case 1, except that government debt is now 100.(this might arguably be the UK or US in a few years).
The primary deficit remains 5, but now we also owe bondholders another 5 in interest expense. So the government has to borrow 10 to cover expenses.
In year 2, we have the primary deficit of 5, but government debt is now 110, so interest expense is 5.5.
If this same situation persists through year 10, then government debt is over 200 and interest expense alone is 10.
At this point, in order to get the budget into primary balance, the government would have to cut expenditures by about 15%. To get into overall balance, expenditures would have to go down by 40%. What are the chances of that happening?
The same as case 1, except interest rates are 10% and outstanding government debt is 150. (This is, more or less, the Italy of the late 1980s.)
In year 1, the primary deficit is 5 and interest expense is 15, so the government has to borrow 20 to cover expenses.
In year 2, the primary deficit is 5, interest expense is 17 and total debt rises to 216.
If we follow this progression out to year 5, debt rises to 270 and interest expense is 24.
In year 7, interest expense exceeds tax receipts!!!
Who’s going to lend to this country, even in year 1? Well, there are the people who lent to Nakheel in Dubai, but they’ve got enough trouble as it is. And, PT Barnum’s beliefs to the contrary, would there be enough of them in any event?
The same as case 1, except debt is 200 and interest rates are 1.5%. (This is more or less today’s Japan)
In year 1, the primary deficit is 5 and interest expense is 3. Total government debt rises to 208.
In year 5, the primary deficit is 5, interest expense is 3.6 and total debt is just under 250.
In year 10, the primary deficit is 5, interest expense is 4.2 and total debt is just under 300.
In this case, the overall government debt is gigantic, equalling 10 years’ tax receipts. How will the government ever be able to pay this money back? Note, however ,that the government can achieve overall budget balance relatively easily, by cutting expenditures/raising taxes by a little over 10%. The biggest danger is that bondholders decide not to roll over existing debt at current interest rates. If interest rates–and therefore interest expense– go up, or if the government is unable to pay current maturities from the proceeds of new debt sales, big trouble arises very quickly.
(A little) realism injection
1. If we can figure out, at least in general directional terms, the possibility of severe future trouble in just a few minutes, so too can any potential government bond investor. So the government in question begins to have difficulty in rolling over existing debt long before the year 10 situations emerge.
2. Legislators, or their aides, understand the developing problems as well. But their desire to bring government largesse to their constituents–and thus assure their reelection–typically overwhelms and desire to cut expenditures or raise taxes. In the “benign” case, they stick their collective heads in the sand. In the worse case, they deliberately foster inflationary policies, on the thought that this lowers the real value of what the government needs to pay back. This throws gasoline on the fire.
3. Off balance sheet liabilities make the debt situation worse.
4. All the bond issuance by itself is inflationary, as is the currency weakness that ensues as foreigners sell their holdings an repatriate the money, or simply hedge the local currency exposure they have.
5. It makes a difference whether the government is bought by locals or foreigners. In the first case, the current generation is borrowing from its sons and daughters, saddling them with debt that finances current consumption. Foreign buyers can come and go much more quickly, so the need to finance a portion of the government’s debt externally usually makes the situation more volatile. One exception: my observation (and, remember, I’m a stock guy) is that in past debt crises in the US, domestic investors have been much quicker to withhold their funds than foreigners.
6. Case 2b is the more “normal” development path of debt running wild. Case 2c, Japan, is unusual in that Japanese savers continue to commit money to the bond market despite very low nominal interest rates. Observers have attributed this to a self-reinforcing circle of economic weakness, as follows:
a. Government policy expresses the social desire to preserve a traditional way of life. So Tokyo’s actions tend to preserve the status quo. In particular, highly inefficient, money-losing companies are not allowed to fail or to be bought by more competent management. This action damages the prospects for healthier firms and results in low economic growth and possible deflation.
b. To stimulate growth, the government engages in (basically worthless) public works projects, which require public borrowing. Because the economy flirts with deflation, nominal yields are low.
c. Citizens recognize the situation and save heavily, rather than consume, in order to have funds for retirement. Since most publicly-traded companies have poor profit prospects, the equity market is seen as too risky. Bank deposit rates are effectively zero. This leaves government bonds, even with a 1.5% coupon, the best alternative.
The big risk to Japan, therefore, is that one day domestic savers find another vehicle for their savings and want their money back. This is unlike the “normal” case, where rising nominal interest rates are typically the country’s undoing.