Monday, May 18, 2015

Thinking About the Unthinkable Bond Bear Market

All of the federal, state, and municipal governments are planning to borrow to cover their operational and pension shortfalls. They think it will be no big deal thanks to low interest rates.

After all, who cares about promising to pay $X next year if what that really means is that you are going to amortize $X over the next thirty years.

The steady state federal budget deficit is probably a good $1T now (wait untill a bear market takes away capital gains), but social security should at least double that in the next few years [pdf].

Then, state and local governments have been averaging $300 billion a year in borrowing, but that could easily grow to be about a trillion every year, given that they will need to borrow for the pension benefits.

I could easily see $3T a year of sustained new government debt issuance. Plus, the Obama-era federal debt is almost all (~75%) notes, which is about $8.3T and about 1/8th being rolled every year makes ~$1T. There's also $1.4 trillion in bills which will of course need to be rolled every year [pdf].

This adds up to like six trillion a year and the thing is it would be more in bad years when government deficits are worse AND pension assets do not return what they were assumed to, so the plans have to borrow more. [And it does not even count roll over of existing state and local debt, which totals $3 trillion, until I figure out the maturity profile of that debt.]

The point is, the total federal debt held by the public was only $6T as recently as when Bush left office, and now the governments are going to need to borrow/roll that much every year.

I could never really figure out what would cause the bond bull market to end, but this is enough to do it in my mind.

The other scary thing for bonds is the effect that interest rate increases have on this system, because it contains so many feedback loops. The pension assets become worth a lot less, the interest expenditures rise (and they are borrowing to pay the interest since there is no debt service), so the credit quality (such as it is) deteriorates.

Also, corporate pensions have the same problem and an interesting feedback loop of their own. To the extent that their pension funds lose money, earnings will take a hit. As we know from Grantham, when earnings fall multiples fall too.

It is very, very nonlinear, because once bonds lose momentum, who will want to own them? Professional asset management and retail investor sentiment are both all about momentum. And every credit - government or corporate - looks much worse with rising interest expense. I think we will come to realize that a lot of stuff in the economy (junk bonds, private equity) was part of a virtuous interest rate cycle.

Maybe this is why the low on the 10 year yield is almost three years in the rear view mirror now.

For the counterargument that the Fed will just buy bonds to "keep rates low", you have to face the fact that QE invariably caused rates to rise, and you could (and we did) make money buying bonds every time the Fed stopped buying them. As I kept trying to explain, the QE bond purchases may have been respectably large in relation to the flow of debt issuance, but they were puny in relation to the stock of $60T of dollar denominated debt. It freaked creditors out about inflation more than it helped.


Anonymous said...

Agree that this is the endgame, but I think it will be years before we get there. Right or wrong, Treasuries are still considered a safe haven, and this creates a lot of countercyclical demand for USgov debt.

Most of the world has bigger, more immediate problems than the U.S., so when the next crisis begins, I would expect the countercyclical demand to appear again like it has in past crises-- at least for a while. Ten or fifteen years from now? All bets are off.

High Plateau Drifter said...

For the past year or so I have been trying to figure out what metric will tell me when the financial system is about to come unglued. Ultimately, the stock markets around the world are being driven higher by increasing debt, and it is increasing debt which is driving corporate stock buy backs ever higher. Wages are being replaced by welfare paid out of government borrowing. Thus, welfare is steadily increasing as wages fall, while total consumption consisting of welfare plus wages is greater than the cost of wages alone to corporate America, thus producing outsized profits so long as debt and welfare outlays keep increasing.

Imagine what would happen if the sovereign borrowing were to stop and corporations were taxed in amounts sufficient to fund this increasing welfare burden. The steady state federal budget deficit is probably a good $1T now, but social security and rising medicare outlays should at least double that in the next few years.

And the borrowing binge is not limited to the federal government. State/local governments have been borrowing on average about $300 billion a year, but that could easily grow to be a trillion every year.

The question is why are these governmental entities so comfortable continuing this borrowing binge and when and how will they be forced to stop. First, with short term interest rates at zero percent, every dollar of past debt is depreciating at about 1.5% if you use official figures (about 6% if you use shadow stats) so there is no question that borrowing by governments at all levels pays. But on a deeper level it appears that every governmental participant in this drama is convinced that the debt will never need to be paid back – that it can be rolled over at negative real rates forever despite its increase along an exponential curve. Zero interest rates are telling them that lenders – and especially central banks – do not expect to stop printing money to purchase debt and understand that they will never ever be repaid, or otherwise retire the debt. So why worrry.

As CP says, we could easily see $3T a year of sustained new government debt issuance. Then, the Obama-era federal debt is almost all (~75%) notes, which is about $8T and about 1/8th being rolled every year makes $1T. There's also about $1.5 trillion in bills which will of course need to be rolled every year.

The short answer is that increasing debt financed welfare consumption means that an ever increasing share of that central bank printed money used to buy debt will hit main street. In response, non-central bank debt holders will sell and drive interest rates up. In fact we have already seen this as the bottom in 10 year treasury rates was achieved in July of 2012. Since Feb. 2, 2015, bonds have fallen in price.

Beginning approximately now, the central banks are faced with the hobson's choice of letting rates rise, or purchasing an ever larger share of governmental debt so as to make central bank monetization of non-repayable debt obvious along with the obvious prospect of future inflation.

I think we are at the beginning of a bottoming in rates with an upward bias that could last as long as several years, in which stocks are likely to print a mirror image top.

AllanF said...

I'm loathe to under-estimate how long the music can play. If there's one constant it's that events such as these take longer to unfold than anyone paying attention ever expects. But at the same time, once they get rolling they snowball incredibly fast, faster than even those that predicted them expected. So, precipitating event in 10-15 years. I don't think you can put a date on it. If 10, why not 15? If 15, why not 20? Seriously, what is unique in 10 yrs vs. 15 yrs vs. next 23 months?

But when the bets do come off, look out. It's going to be everyman for himself. It seems like it's always some unexpected shock, that at the time seems innocuous enough (and maybe ironically that's why it is able to take root), but it forces the marginal buyer out of the market and once that happens there is a general awakening there are no more greater fools left to buy.

Hasn't Japan's current account gone negative? Isn't China's as well? Saudis? Anyone!?! That's my expectation. It will seem no big thing, some country isn't buying US debt like they used to -- strawberry pickers aren't buying spec houses to flip, next tier of start-up .COM's aren't buying SUN servers and CIsco switches like last year -- but then it will dawn on everybody at once, if <insert country> isn't buying US debt, then who is?

I love that Hemingway quote when asked how he went bankrupt. "Two ways. Gradually, then suddenly."

CP said...

Notice that the 10 year bond yield bounced higher off the 200 day moving average:

CP said...

Same with 30 year yield, clear rising rate trend now:

Anonymous said...

CP makes a big list of bold claims...which need to be taken 1 by 1.

#1 - The steady state federal budget deficit is probably a good $1T now (wait untill a bear market takes away capital gains), but social security should at least double that in the next few years.

Looking at the real numbers...
Social Security 2014 Cost (Billions)
$679.5 (retirement)
$143.4 (disability)
$822.9 (total)

SS 2014 Income (billions)
$747.3 (SS payroll taxes + taxes on SS benefits)

$75.6 Billion in the Red in 2014.

From 2014 to 2024 the number of social security beneficiaries will rise from 48 million to 64 million (33% increase). Then you have to factor in the working population increase over the same period.

CBO estimates cost of benefits in 2024 to be $1,263 Billion (includes inflation).

So let's say payroll increases 2% annually (inflation). Then SS income will be $910 billion.

For an in-the-red annual amount of $353 billion in 2024. (inflation-adjusted is lower)

It's a problem, but not a trillion dollar problem. they probably save $100 billion if they just move the retirement age to 70 and stop paying out benefits to people making more than $200,000 in retirement.

Conclusion, first claim, no big deal. CP needs to take more time to look at real numbers instead of making hyseterical claims.

CP said...

One issue with the Pew report, I think, is that it accepts the assumptions built into these funds, e.g., that investment returns close to 8 percent can be obtained every year, that public employees who retire to the golf course at age 50 will drop dead on the actuarially predicted schedule, etc. Generally the economists who have analyzed public pension funds using returns that are actually available in the bond market, for example, have concluded that the gap is closer to 3X the stated gap, in which case states need to scratch up roughly an additional $3 trillion for pensions plus whatever additional they’ll need to pay for promised retiree health care (a worker who retires at age 50, for example, needs 15 years of free health insurance to bridge the gap until Medicare eligibility).

CP said...

This summer it will be four years since the low in 10 year bond yields!

CP said...

With the break above the 2013 high in yields, it seems clear now that we are in a bond bear market:$TNX/technical-chart?plot=BAR&volume=0&data=MO&density=X&pricesOn=1&asPctChange=0&logscale=0&sym=$TNX&grid=1&height=500&studyheight=100