August 08, 2005
Is there a danger that the Fed will fall behind the inflation curve?
Both Macroblog and Capital Spectator raise the prospect today that rising long-term yields might mean that the Fed waited too long before trying to stamp out the cinders of an incipient inflation fire. I would suggest instead that the increase in long-term yields over the last few months is the natural development that we expect to see in a situation of Fed tightening and has little to do with inflationary expectations.
The long-term interest rate is driven by two factors-- expectations of where future short rates are headed, and investors' preferences for short-term versus long-term holdings. Mathematically, the 10-year rate can be written as the sum of the discounted present value of the market's expectation of future 3-month rates over the next 10 years plus a term premium. That decomposition is actually just the definition of what we mean by the term premium, but it's a definition that can be a useful way of thinking about what accounts for changes in long-term interest rates.
When the Fed is raising short-term rates rapidly as in the present environment, there are several reasons why longer term yields could change. If the Fed is expected to continue tightening, near-term short rates are headed up, and through the expectations component (the first term in that decomposition of the 10-year rate), this would be a factor driving the long rate up. Short rates at more distant horizons might be expected to change in either direction-- down, if the market believes that the Fed will be successful in slowing the economy and bringing down inflation, or up, if the market discovers that there really was a stronger economic boom and a bigger inflation threat than it had previously recognized. As far as the second term in the decomposition is concerned, the effect of rapidly rising short rates is usually to drive the term premium down.
|Episode||Fed funds change||10-year change|
|1958:05 - 1959:07||2.84||1.48|
|1967:05 - 1968:07||2.08||0.65|
|1972:01 - 1973:03||3.59||0.76|
|1983:06 - 1984:08||2.66||1.87|
|1988:01 - 1989:03||3.02||0.69|
|1993:11 - 1995:01||2.51||2.06|
|2004:06 - 2005:08||2.47||-0.41|
As a theoretical matter, then, it would seem that the long-term yield could either rise or fall as a result of rapid Fed tightening. The historical experience, however, is that it almost always seems to go up. The table on the right looks at the seven episodes over the last half century when the Fed drove the fed funds rate up by more than 200 basis points within the space of a year (not including the hectic 1979-82 period when there were all kinds of wild swings in interest rates). The table summarizes the change in the fed funds rate and the 10-year Treasury yield over the 14-month period following the first funds rate hikes. Up until the current episode, in every single one of these episodes, the long rate rose along with, but not as much as, the short rate.
Given these numbers, it might seem surprising that some analysts are thinking of rising long-term yields as a source of concern in the current episode, since, in contrast to every previous episode, long yields still stand significantly below their values of a year ago. I commented earlier on this anomalous behavior of long yields in the current episode, coming to the conclusion that the fall in yields over the last year was largely due to a decline in the term premium.
|Instrument|| Change over |
last two months
The reason that some analysts are nevertheless currently talking about rising long-term yields is that, over the last two months, the trend in long yields has reversed, and long-term rates have come back up significantly. However, even if we only focus on the last two months, the rise in long rates is less than the rise in short rates. It's also interesting to note that inflation-indexed Treasuries have gone up together with the 10-year nominal yield, suggesting pretty strongly that it's not inflation fears that have contributed to the increases in the long rate, at least up to this point. Instead it looks much more like the market has come to expect ongoing rate hikes, driving the 10-year rate up through expectations of a higher fed funds rate over the next few years.
Granted, the concerns that others have raised have to do not so much with the hikes in long rates that we've seen so far, but rather with a climb in long-term yields that some traders are expecting may be yet to come. But Dave Altig is quick to point out that nobody has a very compelling track record for making such predictions.
Capital Spectator's concerns hinge on the possibility of further big oil price increases. I suggested last week that concerns about stability in Saudi Arabia have been responsible for the latest moves up in oil prices, a factor of course increasing significantly just today. Oil prices have gone up so far based on the still relatively small probability of a major disruption. If there is major civil upheaval in Saudi Arabia, then yes, we'd have some more inflation in the U.S. But it strikes me that would be the least of our problems.
Posted by James Hamilton at August 8, 2005 10:22 PMdigg this | reddit
Another issue that might be relevant is expected changes in exchange rates. Theoretically, these should be the critical driver of interest rate differentials across currencies. If US bond yields have risen more than foreign yields, this could be taken to mean that the dollar is expected to decline more rapidly than previously anticipated. When China announced that that the Yuan would be pegged against a basket of currencies rather than the dollar alone, perhaps there was a decline in the long-term prospects for the dollar relative to the currencies in which other bonds are issued.
Posted by: knzn at August 9, 2005 05:30 AM
Jobwatch, http://jobwatch.org, has an interesting chart on IT employment pre and post recession. Think there was a bubble in IT employment? Think again. All there was was linear growth at a good clip from 1994. It never hit bottom until 2004 and has only edged up slightly since then. IT employment is still off 25%.
Eventually we won't create any jobs after recessions and the worker shortage touted to appear once the boomers begin retiring will fail to materialize.
Posted by: Lord at August 9, 2005 10:36 AM
Watching the change in the ten-year rate over the past few months it seems there were two events that, either directly or indirectly, may have had an upward effect on long-term rates. The first was the July 20th announcement that China was re-evaluating its currency. The second was the announcement that the U.S. Treasury department would reissue the 30-year bond.
A quick look at the change in the ten-year rate during the month of July shows an almost 50 basis point increase. In fact, the current ten-year rate of about 4.4% is the highest since April 2005. Could these two events, as well as tightening of monetary policy, have played a role in the recent long-term rate increase?
Thanks for your continuing comments on economic activity.
Posted by: Jason Buol at August 9, 2005 10:44 AM
Jason, on July 20 the nominal 10-year constant-maturity rate was 4.17. It's currently quoted at 4.43, a 26 basis point increase. On July 20, the 10-year TIPS was 1.88, currently quoted at 2.04, a 16 basis point increase. So, even over the period to which you call particular attention, I think rising real rates have been the most important part of the story rather than fears of inflation coming from the dollar depreciation. And the main reason real rates have risen over that 3-week period, in my view, is that the market has come to believe that the Fed is going to keep pushing the short rate up and up.
Posted by: JDH at August 9, 2005 11:11 AM
Anticipated dollar depreciation could be an issue even apart from inflation fears. The severity of the trade deficit suggests that a real depreciation is necessary, and the Yuan announcement could have signaled an acceleration in that process. Particularly since the immediate currency effect of the revaluation was relatively small, the anticipation of future dollar depreciation (apart from any possible effect on the US inflation rate) means that US bonds became less attractive relative to foreign bonds, and therefore US yields needed to rise.
Posted by: knzn at August 9, 2005 12:01 PM
Another possible reason why the long rate is acting the way it is, is that the market may be saying that the Fed is wrong,that they're overdoing it now and will have to aggressively lower the fed funds rate later.
Not too long ago Bill Gross of Pimco made that argument and said the Fed would need to be lowering rates in the near future.
For those of a monetarist persuasion, a look at M2 supports this argument. For roughly the last year (until the middle of June) the growth rate of M2 decelerated steadily and actually became negative for awhile (it has since turned up). This would indicate a slowdown coming.
It's a theory anyway.
Posted by: Jim Miller at August 10, 2005 05:03 AM
Justin Lahart observed in the WSJ Ahead of the Tape column yesterday that riskier fixed income instruments like junk bonds still have relatively low yields, not just treasuries. Yes, there's an argument that treasury yields remain low because of recession expectations. If that were true, though, you'd think that yields of riskier issues would be pushing up. Not so in this weird market.
Posted by: Mark Sullivan at August 10, 2005 06:03 AM
The ratcheting up of long rates is currently like the difference in cost push inflation and demand pull. By forcing up the cost of short rates, the Fed is lowering the inflation premium in the long bond but stong economic growth is indicated. Buyers find it psychologiclly tough to extend maturities while the Fed is pushing so hard on its string.
Posted by: Jack Miller at August 10, 2005 10:47 AM
1. I don't like this: "up, if the market discovers that there really was a stronger economic boom and a bigger inflation threat than it had previously recognized. " Why are you mishmashing the two different things. And the way this is written it almost sounds like you are equating the two or saying that increased production causes inflation (which makes no sense).
2. Based on the TIPS rising, either people have an increased feeling of risk (in US government, in funny business in the TIPS adjustment, etc.) which is bad. Or the market thinks that more attractive "projects" (investments) exist. So that competition for capital is heating up. Which is good. If equities are heading down while TIPS is up, that would argue against the second case, no?
Posted by: TCO at August 11, 2005 02:24 PM
Why is it that one cannot determine empirically that substantial infusions of capital from Chinese banks, Indian banks, etc. are largely responsible for the curious delay in the rise of long-term rates?
KNZN noted that U.S. long-term bonds may be less attractive than foreign bonds. In addition, if short term rates are higher than long-term rates, then wouldn't rational investors who are indifferent between buying long-term bonds or buying short-term instruments prefer the short-term instruments?
If the demand for long-term bonds drops, then the U.S. government would have to offer higher yields to fund debt.
Isn't it also possible that some institutional buyers may have some sort of preference for long-term instruments, even if they are offered at a slightly lower rate?
Posted by: Ed Sasaki at August 11, 2005 03:50 PM
1. Capital is fungible. If the chinese are idiots, then why don't the hedge funds jump in and take the opposite positions?
2. The rationale for buying a bond that has lower rate but longer term is that you expect a dramatic reduction in inflation. for instance, if inflation is high.
Posted by: Anonymous at August 11, 2005 07:09 PM
I'm certainly not an economist, and venture into this discussion with some trepidation...but is it not at least possible that the bond market's message is being distorted by the current less than optimal functioning of the futures markets?
Recent reports below:
Fed's fight squeezes banks, spooks markets
Short-bond shortage isn't over
Posted by: Greg Newton at August 11, 2005 07:54 PM
I read that bank squeezing article. that article said nothing about "problems with the futures market". What's the "problem", tickertape get stuck?
Posted by: tco at August 11, 2005 08:05 PM
I'm not an economist
I'm a trader and I'm long gold.
(these are my biases)
It seems to me that a significant contributor to the connudrum is the scarcity of long-dated Treasuries. The absence of new long bonds and the shift of issuance towards the short end of the curve has created a scarcity. Many pension funds, endowments, and other institutions are required to match their term balance sheet. Hence, these large players are forced to buy up (relatively) scarce 10 and 30 year bonds regardless of the yield.
This effect may be exacerbated by (at least) two other trends:
Firstly, the emergence of China, India, the 5 Tigers, etc, as strong export-based economies. This (relatively) recent phenomenon has left the Pac Rim flush with dollars. These coutries must buy dollar-denominated assets or risk appreciating their currency and thereby slowing their economic development.
What could be a better buy than the highest yielding "risk free" asset, aka, long dated treasuries?
Secondly, the structure of US (and worldwide) demographics. If you are 65 years old today, you were born in 1940. The years 1929-1945 was a period of low birth rates in most countries due to the global depression and WWII. I speculate that the final 6 1/2 years (comprising the war) were the trough of this period of few births; i.e., the calm before the baby boomer storm.
What does this mean?
I interprete it to mean that there are relatively few expected retirements over the next 5 years to be followed by an unprecedented period of very numerous retirements as the baby boomers stop working.
If you are a pension fund manager and you are bound to match the term structure of your assets to your liabilities, this means that anything shorter than the 5 year note is not much good to you. You must buy long-dated bonds to match your expected payout term.
Ok, I'm sure that I have said little (if anything) new to the professional economist.
But what I conclude from the above is that there exists a worldwide situation of both enhanced demand for, and reduced supply of, long-dated dollar-denominated bonds. In Econ 101 parlance:
the supply curve has shifted left and the demand curve has shifted right AT THE SAME TIME.
This sets a very high price for long-dated treasuries. I believe that this situation will be unwound (and I remind you that I am positioned to benefit financially if this occurs FWIW).
The Pac Rim exporters cannot keep their currencies artificially weak forever. Continuing expectations of dollar weakness must affect their investing calculus at some point, no? There are other dollar-denominated stores of value out there (I might suggest Gold, lol).
Additionally, the mandated institutional preference for long-dated maturities will unwind as the baby-boomer's retirement nears. Indeed, this preference will probably be reversed as pension funds find themselves in need of cash on hand.
Therefore, I see major factors contributing to the connundrum as historically unique and temporary.
Posted by: My 2 Cents at August 12, 2005 02:53 AM
I think there is another explanation for the failure of the US bond yield to follow the fund rate higher earlier on, and its mini-reversal in the last few weeks.
Your decomposition of long rates, as the sum of expected short rates and the RP, is of course correct. However, you place a "restriction" in assuming that (expected) short rates may vary just in response to business cycle stimuli. Whilst those normally drive short rates, there is a stickier component, that is the "natural" real rate of interest (this enters both long and short rates). I think the last two years' surge in (Asian, central banks)demand for US bonds caused a near-structural decline in that natural rate... Much like if the intercept in the bonds' demand function had been increased. As we know, this (Asian) demand was truly "exogenus", due to the need to sustain the USD against Asian currencies, yuan first. This may have been responsible for a non-fundamentals-related addition to demand for bonds that has unusually compressed the yield. Of course, it is likely that bonds' demand has abated (for ever?) after the revaluation of the yuan, and more recent signs the Chinese may revalue/liberalize further. Were that the case, the earlier lid on the real (short) interest rate is being removed and we should expect a medium-term increase in US bond yields, perhaps catching up on what earlier changes in the Fed fund rate would have normally implied for long yields.
Posted by: giorgio radaelli at August 12, 2005 05:29 AM
TCO, I did not mean to suggest (and do not believe) that higher real output necessarily means higher inflation. I was listing higher real output and higher inflation as two separate forces that could lead to an increase in higher nominal interest rates (maybe I should have used "and/or" in place of "and" in the sentence to which you object). Higher real output can mean higher real interest rates when the boom means a greater demand for investment and marginal return to capital.
Posted by: JDH at August 12, 2005 10:33 AM
Thanks JDH. I do see these two things confounded a lot in the press and in discussion. I figured you wer probably "listing" but wanted to make it clear. For all I knew, you might school me and show me why a boom should lead to inflation. I don't know how many times I read that the Fed is considering stepping in because they think the economy is growing too fast. As long as the CPI is ok, why jerk with looking at growth rates (and yeah...booms could be indication that they mishandled the money supply previously, but that whol approach seems likely to lead to overdriving.)
Posted by: TCO at August 12, 2005 04:52 PM