When bonds lose their bid

By Felix Salmon
March 2, 2011
Bill Gross, who's worried about what's going to asset prices -- both bonds and stocks -- when QE2 comes to its scheduled end on June 30. He has two main points:

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

A couple of big names are out with cautious bond market views this week. For the big picture, turn to Bill Gross, who’s worried about what’s going to asset prices — both bonds and stocks — when QE2 comes to its scheduled end on June 30. He has two main points:

  • For the duration of QE2, the Fed has been buying 70% of all new Treasury-bond issuance, and foreigners have been buying the other 30%. When the Fed stops buying, who will step in to replace it? After all, with a $1.5 trillion budget deficit, there’s a lot of new supply coming.
  • Treasury yields are about 150bp too low. The yield on the 10-year Treasury is typically the same as the GDP growth rate; it’s now 150bp below that. Real 5-year Treasury yields are normally about 1.5%; they’re currently negative. And, of course, the Fed funds rate is artificially low. All of this implies that yields will rise. When that happens, it’s reasonable to assume that discount rates and credit spreads will rise along with them, driving all asset prices lower.

This need not happen immediately upon QE2′s demise, but it might: Gross foresees “immediate uncertainty and fear” come June 30, and strongly implies that he’s not going to be venturing far out the curve unless and until rates rise significantly. For the time being, he’s derisking: “PIMCO’s not sticking around,” he tells us.

Meanwhile, on the state level, David Nowakowski and Prajakta Bhide of Roubini Global Economics have a big report out called “States of Despair,” in which they estimate that muni defaults could reach $100 billion over the next five years. They do stress, however, that the recovery value on those defaults is likely to be very high — roughly 80 cents on the dollar — and that “state and local debt problems are not systemic in nature, and will not infect the financial system, though they will dampen economic recovery.”

Indeed, the report is actually bullish on the muni market: if recoveries really are 80%, then you’d break even buying the MCDX index even if the five-year default rate hits 34%. In other words, even if there are $100 billion of bond defaults over the next five years, you’d still make good money buying munis at these levels.

There’s lots of very good analysis in the report, which I’ll come back to later today. But my main takeaway is that this is clearly a market requiring good analysis — it’s not something that individual investors should buy blithely, as they have in the past, on the assumption that muni bonds are not only tax-free but also risk-free. As such, the muni market has a similar problem to that facing the Treasury market: when the biggest buyer of the bonds goes away, who will replace them? So far, the answer in both markets is very unclear.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/

Gross’s first point shows the same poor understanding of government deficits as those who engaged in the “widow maker” trade by shorting Japan’s government bonds. As a matter of accounting, government deficit spending adds to the financial savings of the combined private (domestic) and foreign sectors! The addition to savings, on aggregate, is exactly enough to buy an equivalent amount of newly issued treasury debt. The portfolio shifts occurring under QE are largely irrelevant.

Too see this, try “government spends”, “government issues debt”, and “government spends (consolidated)” in this visualizer:

http://econviz.com/balance-sheet-visuali zer.html

Gross’s second point only tells half the story… if the historical relationship between treasury yields and GDP growth is out of alignment, then yields may rise *or* current GDP growth may fall. There’s no way to be sure at this point which is more probable, nor the timing.

Posted by hbl0 | Report as abusive

@hbl0 –

Somebody is wrong, either you or Bill Gross. By your logic, all government debt is safe so long it is covered by a bid (until suddenly it is not!…)

Yes, government debt is matched by savings, but there is a major time mismatch. I save presently to finance government spending but the government must pay it back in the future. (And please no argument about government not having to pay it back and just rolling it over… of course every bondholder expects to be able to get their money back at some time in the future). Even if the government never pays back its debt in full, at some point it must reach rough equilibrium where what is going out roughly matches what is coming in.

You fail to distinguish between financial savings and real savings. Real savings examples:
(1) I have $10,000 to spend on a roof and my roof is at the end of its life. I invest in a new roof and, if the roofer’s claims are true, I will have something that will deliver value for the next 30 years, a roof over my head, paid in full.
(2) I get spend $200,000 on a degree and education in medicine and will be able to practice profitably for the rest of my career.
(3) I buy shares in an oil driller whose skill, equipment and reserves should be able to deliver value for many years to come.

What does the government’s “financial savings” consist of? I am buying a claim on future revenues of the government, essentially taxes. Who pays those taxes? I do.

On an individual level, bonds do represent savings but on a collective level they do not, because they are balanced by a future tax burden.

Just because Japanese bonds have held up so far is no indication that they will do so in the future. Japanese bonds have ridden a roller coaster up as an aging population collectively saves for retirement all at once. They will then ride the same roller coaster down as the same population retires with few people behind them. What will help them after their bonds crash is that many people won’t actually be able to retire.

Posted by DanHess | Report as abusive


None of your points seem to address the point I was responding to, namely “When the Fed stops buying, who will step in to replace it? After all, with a $1.5 trillion budget deficit, there’s a lot of new supply coming.” In fact, you agree that “government debt is matched by savings”.

I disagree that there is a time mismatch problem, because the bond ISSUANCE will continue happening in PARALLEL with the deficit spending after QE ends… thus no problem with insufficient buyers of treasuries.

I think may be suggesting that potential buyers will be scared that the debt won’t be repaid, potentially leading to a buyers’ strike. I disagree because:

(A) the aggregate debt (versus individual bonds) never has to be repaid, nor is it likely to be. The private sector and foreign sector together choose the size of the government debt via their savings preferences (effected mostly via the automatic stabilizers).

(B) the central bank is always available as buyer of last resort (even if via the primary dealers as an intermediate step) and the market knows it and thus is happy to bid instead and hold an income-yielding asset.

“Real” investment matters too but is independent of these financial savings flows, since the money you spend always goes somewhere (to the roofer, to the university, to the driller, etc).

Posted by hbl0 | Report as abusive


No, I don’t agree that “government debt is matched by savings”. In fact, it emphatically has not been, as you concede later in your comment. Quantitative easing has filled the gap.

It may take much higher interest rates to get a bid on bonds in the coming years. This increase in interest rates would be very contractionary.

“The private sector and foreign sector together choose the size of the government debt via their savings preferences” — really? Score one for originality. If you want proof that is not true, look at China. Enormous savings, little debt market. Many would want Chinese bonds if they were available.

My personal debt is not driven by the credit card cos willingness to lend to me. I can use my credit line or not. America uses all of its credit line and more.

- Dan

Posted by DanHess | Report as abusive

I’m pretty sure this is related to what Krugman once quoted someone as calling “the doctrine of immaculate transfer”. If demand decreases, the markets will adjust such that the amount bought is still equal to the amount sold; okay, very good. It’s an accounting identity, and it’s necessarily true. That doesn’t, however, mean that the prices will stay the same. If a major lender goes away, something has to happen to bring about an increase in lending from someone else, a decrease in borrowing, or some combination of the two. You seem satisfied to stop there, but we’re worried about the something.

Posted by dWj | Report as abusive

Er, that was largely addressed to hbl0. Sorry I didn’t make that explicit.

Posted by dWj | Report as abusive

Dan Hess,

You said “Yes, government debt is matched by savings, but…” in your second 3:53pm sentence. That agrees with what is explicitly shown in the the visualizer I linked to. I’m not sure how you can argue with that. Gross and others have expressed a concern many times about who will be left to buy treasuries when the Fed stops. The point is that the deficit continually adds new financial savings in an adequate supply to ensure the bid for treasuries remains roughly unchanged.

The yield that treasury investors demand can of course change, but it is independent of the above (it’s based on things like expected future short term interest rates, inflation, economic growth). A lot has changed since the gold standard ended.

Your other points are worth discussing in more detail, even though they are not directly related to my original reply and I mostly disagree with your positions, but I’m out of time, sorry!

Posted by hbl0 | Report as abusive

@hbl0 –

I am afraid you are still missing the time mismatch. If the government has sold a bunch of 20 year bonds to me, that ‘financial savings’ is not available to me at any time in the near future to be recycled back into further bond purchases by me. That is the whole point. The government doesn’t have the money, so they are borrowing mine. My new bonds are an asset but they are not deployable savings. They have already been deployed.

If I as a holder of 20 year bonds tried to use that savings to buy more bonds, I would have to sell my own bonds first to raise the cash, but that only adds to the bond supply problem. If I must sell bonds to buy other bonds, I am not helping things at all.

I could wait 20 years for those bonds to mature but there is a good chance that inflation will have reduced their value by then. And that won’t in any case help the government that needs to raise more money in 2011 and 2012.

Posted by DanHess | Report as abusive

DanHess seems to have the high ground here. If inflation rises unchecked by a FED bent solely on reducing unemployment than the “real” bid for bonds will vanish. China will buy gold, copper, energy, or whatever instead of American bonds. When the real buyers disappear then the FED will be the only buyer left. They will be left buying new issues and the rolling bonds as well. That’s a big chunk of change.

The average maturity of the debt is something like 4 years… and the 30′s skew the average to the right. so in the next 4 years we’ll issue ballpark 5 trillion in NEW debt and roll at least that much. 10 trillion in bonds to issue in 4 years.

The dollar is TOAST.

Posted by y2kurtus | Report as abusive

Some data on the debt maturity here:
http://www.treasury.gov/resource-center/ data-chart-center/quarterly-refunding/Do cuments/Monday%20chart%20template%20Feb% 202010_Final.pdf

A little over half the marketable debt will mature within three years. (Roughly $7T?) Plus new issuance. Call it $10T in *three* years.

Average maturity is now over 60 months and trending higher (new issuance averaging around 80 months). Better take advantage of these low interest rates while you can…

Posted by TFF | Report as abusive

Dan Hess,

Let me continue your example from your 10:54pm comment. You have bought a government bond, so yes you alone cannot buy new bonds on top of what you have. But, imagine next the government deficit spends an additional $50,000 into the economy by hiring your neighbors (yes, government really can spend to some degree before issuing bonds, since treasury always has a positive account balance at the federal reserve). Your neighbors spend some of that money and save some, but all the money is held by someone. Whatever mix of people ultimately saves the money has that extra “buying power” (as a result of the deficit spending itself) to buy newly issued government bonds and thus earn a higher rate of interest on their savings.

Whether the price (and thus yield) demanded by those buyers changes is largely unrelated, since at the macro level there is no shortage of newly-saved-and-ready-to-invest money available to buy the new bonds. If buyers demand higher yields it will be for other reasons. Markets price assets (including treasuries, stocks, etc) according to anticipated risks and return, even though the market may not always get it right.


I don’t think this is like Krugman’s “doctrine of immaculate transfer” scenario. I started a more detailed answer but realized I can’t cover such a complex topic adequately in the time (and comment space) I have. I guess I need to be done on this comment thread, but thanks all for at least reading my comments!

Posted by hbl0 | Report as abusive

@hbl0 –

You say “Whatever mix of people ultimately saves the money has that extra “buying power” (as a result of the deficit spending itself) to buy newly issued government bonds” but there is a gaping hole. The money that the government spends does not have to go back into treasury bonds at all, especially when there are all manner of other dollar-denominated assets available, many of which are much more attractive and do not risk destruction by inflation.

Posted by DanHess | Report as abusive

Dan Hess,

You are correct that far less than 100% of savings arising from deficit spending will directly generate a bid for treasuries. But I think maybe our difference of opinion on this comes down to how much power you believe arbitrageurs have in the marketplace.

If you believe an overpoweringly dominant percentage of investors are “rule based” buyers (e.g., buy 60% bonds and 40% stocks, no matter what the current valuation of each is) such that this type of investor dominates all asset price movements, then I think your concern about effects on treasury bond prices are valid.

But if you believe there are enough arbitrageurs out there to move markets (e.g., selling their stocks or corporate bonds to buy treasuries any time treasury yields start getting more attractive on a relative basis) then I think my assessment is accurate, i.e., deficit spending creating the savings needed on aggregate to support an equal issuance of new treasuries without meaningful impact on yields.

I can’t point to research on which is more accurate (though I’d love pointers if it’s out there), but the post-gold-standard historical evidence I see strongly suggests the latter scenario is more representative (e.g., Japan since 1990 with its huge deficits and low JGB yields).

Posted by hbl0 | Report as abusive

An interesting discussion, DanHess and hbl0.

* If arbitrageurs dominated the market, then wouldn’t it be practically impossible for Fed actions to move the bid? I suspect that pension funds (and most individual investors) are percentage-based, and that most arbitrageurs operate *within* an asset class rather than *between* asset classes.

* I’ve read suggestions that massive public borrowing tends to depress corporate borrowing. Surely this makes sense? The arbitrage behavior will kick in earlier between similar asset classes (corporate bonds vs. Treasuries) than between dissimilar asset classes. Moreover, if there is more cash sloshing around the economy then it is more likely that corporations will be able to meet their needs without borrowing.

* Surely risk perception plays an important role in pricing? And debt loads are definitely a factor in risk perception? This is admittedly contrary to the example of Japan, however the countries are quite different in some ways — in particular, the majority of Japanese debt is held domestically. (This macro-analysis fails to capture that distinction.)

* Part of the reason why this is such a mess is that the currency itself is only valuable insofar as it can purchase valuable goods. At some point you have to step back and consider whether an action ultimately creates or destroys real value created in the economy — and that is a hugely difficult question to answer. Everything else is ultimately circular.

Consider the “water cycle” as an analogy. Evaporation, rain, groundwater, rivers, lakes, glaciers, oceans… Certain small segments of the water cycle create value (which can be increased or decreased through our choices), yet no matter how we alter the system the whole remains a closed infinite cycle.

Posted by TFF | Report as abusive


Arbitrageurs might include investment banks, hedge funds, sophisticated individuals, day traders, etc — I don’t know exactly who but it seems likely that collectively they have enough power to move asset prices even across asset classes.

It’s not just Japan… look at US treasury yields versus deficit spending since the Great Recession started in December 2007. Even during times of no QE, rates didn’t even begin to soar, they mostly went lower.

I don’t think the Fed currently (the way they run thinks like QE now) controls long term interest rates (and many people agree). They might move the bid in the very short term, but why wouldn’t there be price-sensitive market participants out there countering much of the effect?

If you think that rule-driven investors drive asset prices, and you also recognize that deficit spending adds to private sector savings (see my visualizer link above), then logically you should predict that broad asset prices (stocks, corporate bonds, etc) should actually RALLY when the government deficit spends and people invest those savings using their allocation rules… at the same time as government bond prices would be falling. I think most suggestions that public borrowing depresses corporate borrowing are politically motivated. But of course there are cases in theory where it could be true if public borrowing truly is “excessive”.

Bottom line – I think it’s important to distinguish between:

1. Is there a “shortage” of savings available to “fund” government deficit spending such that either the money simply isn’t there, or finding it would drive rates higher? That was the focus of Felix’s post I responded to, and my answer is “no”. Government deficit spending precedes government “borrowing”.

2. Does the government’s deficit spending alter the fundamentals such that treasury yields should be driven higher? (By expectations of higher fed funds rate, inflation, economic growth, etc). Yes, too much deficit spending absolutely COULD have that effect. But in most advanced economies today (like the US) the deficit spending is not excessive, it mostly results from the automatic stabilizers and it barely fills the recessionary drop off in private aggregate demand. But while important, this second issue isn’t what the post was about.

Posted by hbl0 | Report as abusive

“logically you should predict that broad asset prices (stocks, corporate bonds, etc) should actually RALLY when the government deficit spends”

Isn’t that pretty much the conventional understanding? Fiscal stimulus = asset rallies?

There are other factors in play, of course, which is why asset classes don’t all move in lockstep. But I thought that the reasoning behind QE and QE2 was the need to inflate the stock and real estate markets to prevent massive defaults. No?

Posted by TFF | Report as abusive


That paragraph of mine you just quoted from wasn’t very clearly written, but I was responding to the notion of whether government borrowing would depress corporate borrowing. (I believe the answer is usually no.)

Yes, fiscal stimulus can support asset rallies but that’s because of the support it gives for aggregate demand and thus for corporate earnings. i.e., I don’t believe the rallies are dependent on “rule-driven investors driving asset prices” as they invest saved proceeds of the new deficit spending.

Also QE is not fiscal stimulus, it’s monetary stimulus, which is very different. Whether (and how sustainably) it inflates asset prices is controversial but that’s a different topic entirely.

Posted by bjornar | Report as abusive

Oops that last comment was me.

Posted by hbl0 | Report as abusive

Thanks for the thoughts, hb10. Much to ponder on.

Posted by TFF | Report as abusive