June 13, 2005
Worries about the yield curve
The yield curve has inverted eight times in the last half century, and in six of those episodes, the U.S. ended up in recession. Some analysts are concerned that we may see a ninth inversion before the end of this year.
One of the economic indicators that many of us follow closely is the difference in yields between assets of different maturities. Below is a graph of the difference between the yield on 10-year U.S. Treasury bonds and the average federal funds rate (an overnight one-day interest rate) going back to 1954. Usually the longer-term maturity pays a higher yield; historically this spread has averaged around 0.9% (a little below 100 basis points). Having gone through an unusual three-year episode in which the spread was well above this average, we're now right about back to the historical average.
But most analysts are predicting that the Fed will raise the fed funds rate up to 3.25 at its next meeting, and fed funds options are currently reflecting a belief by the market that there's as much as a 15% chance that the Fed will leap up to a 4.0% target for fed funds by October, according to David Altig's latest calculations over at macroblog. If 10-year yields hold on to a value below 4.0%, that would leave us in the historically unusual situation of an inverted yield curve, in which the fed funds rate is actually above the 10-year rate. This possibility has raised concerns by many analysts, such as Bloomberg's Caroline Baum.
So what's the worry in an inverted yield curve? The above graph indicates with shaded areas the times in which the U.S. economy went into a recession as determined by the National Bureau of Economic Research. The yield curve only inverted 8 times in the last half century, and on six of those occasions, the U.S. went into a recession. There have been many dozens of academic investigations of the statistical correlation between the yield spread and economic activity, which generally confirm what you'd expect from the above graph: the lower long-term yields are relative to short-term yields, the more pessimistic your forecast of GDP growth becomes. James Stock and Mark Watson have a nice survey of this literature.
There are several theories for what may explain this correlation. The most straightforward has to do with monetary policy. The Fed controls the fed funds rate directly through its control over bank reserves and the nation's liquidity, but influences the long rate more indirectly through a combination of the intermediate-term liquidity prospects and longer-term inflation. When the Fed is tightening, as it has been for the last year, short rates go up relative to the long rates. And if the Fed tightens too much, the economy slips into recession.
So why would the Fed keep doing this to us? Conducting monetary policy is complicated by the fact that there is a significant time delay between the date at which the Fed takes an action and the date at which the economy responds. Doing the right thing therefore calls not just for an assessment of where the economy is at the moment but also a prediction about where it's going to be 6 months or a year down the road.
It's awfully easy to fall behind in that game. A common historical pattern is that the Fed gets concerned when the economy starts to stumble, so begins to lower interest rates aggressively. The economy doesn't seem to be responding, so the Fed gets even more aggressive, and continues to pump money into the economy because they don't recognize that a recovery is under way. Later on when the economy starts to boom, the inflationary consequences of that expansion catch up with us, so the Fed starts to ratchet rates up in an effort to bring inflation back under control. The policy delays between monetary expansion and inflation are even longer than those between a Fed contraction and recession. So the Fed sees that inflation continues to climb despite its efforts, and figures that it needs to tighten still further. Once the recession begins, the Fed belatedly figures out that it went too far, and the whole cycle is repeated again on the way back down.
It sure seemed to me that the first part of that common historical pattern was repeated this go-round, with the Fed overstimulating the economy in bringing interest rates all the way down to 1% in 2003. Now it appears that they're worried about the inflationary consequences of that overexpansion. If we'd held the rate at 2% through 2002, there might never have been a need to raise it above that now.
As I noted here, a rapid rise in oil prices is another indicator that has also been historically associated with economic recessions. If we do see the 100-basis point increase in the short rate over the summer that some market participants seem to think is a possibility, a prudent person should be quite concerned.
I hope that the market guess of a 4% fed funds rate is wrong, and think it probably will be. The Fed couldn't be that stupid, could it?
Posted by econbrowser at June 13, 2005 12:24 PMdigg this | reddit
Listed below are links to weblogs that reference Worries about the yield curve:
Tracked on July 2, 2005 10:56 AM
» Fed Governor Olson: Fed Unlikely to Tighten Rates Too Much from Economist's View
Federal Reserve governor Olson gave a speech today on rural development. After the speech, he said he wasn't worried about over tightening: Olson Isn't Worried the Fed Will Raise Rates Too Much, Bloomberg: Federal Reserve Governor Mark Olson said he [Read More]
Tracked on December 5, 2005 09:59 PM
But don't you think the Fed is looking at this chart now and has looked at it in the past? A more interesting scenario will be if the recession starts while the curve is relatively steep.
Posted by: michael at June 13, 2005 04:09 PM
Paul Volcker had an elegant way to describe the Fed's actions on April 20: They belatedly started rising interest rates. Now they face oil price induced inflation that will not go away as the far out oil futures show and this was indirectly confirmed today by OPEC voices.
Core looks good, but I am still looking for the convincing answer to the question whether eating indexes in the cold dark without commuting to work truly reflects the development of the cost of life of the 4 lower quintiles of the population.
Food and energy are the most unflexible in consumer spending for the majority of people. Hedonic changes won't make them eat less. An often overlooked fact is that the US are expected to become a net food importer this year for the first time. Haven't checked prices of softs lately.
The rapid rise in oil prices will also show up in European export prices soon as the old continent got a double whammy this year from higher oil prices to be paid in a 13 percent stronger dollar.
So does the Fed have any other choice than to rise rates to stay at least on a par with the CPI in order to remain credible or will they lean to their role as a protector of employment by staying below "neutral"? After all the current upswing is the one with the smallest employment gains in the private sector ever recorded in a period of economic expansion.
Posted by: Toni at June 14, 2005 02:41 PM
Although financial markets are notoriously anticipatory, high levels of uncertainty can cause very mixed and confused and volatile market signals, such that forecasting based on almost raw noise is hardly an advisable practice.
Although we all "know" that short rates "will" go to 3.25, 3.50, 3.75, 4.00, 4.25, or maybe even 4.50, there simply isn't enough certainty to effectively and reliably trade off of, other than the use of a dart.
Key point: Until short rates actually arrive at a given level and had some time for effects to filter out into the real economy, we don't know what those real effects will actually be relative to some modelled effects. And since we know neither the final short rate nor the state of the economy at that point, forecasting the equilibrium state of the economy months after short rates stabilize is quite a fool's errand.
Only as we get to that short rate equilibrium point will will finally be able to say anything meaningful about long rates relative to some perceived causal relationship between actual short and actual long rates.
Why should we believe that there should be any hard and fast relationship between wildly-varying perceived future short rates and current long rates?
Surely inflationary expectations will evolve as we travel from today to the point of short-rate equilibrium; won't those final inflationary expectations be a far more important driver of long rates at that point than the combination of current inflationary expectations, current short rates, expected inflation, expected short rates?
What is it that we think we know and why do we think we know it?
And, why do so many people feel that the markets are being "stupid"?
One problem with the markets: You can't tell the difference between a position that is a hedge for a primary position placed elsewhere and an outright bet on a given outcome. Some people may have real financial positions that will work at short rates of 3.50, and simply "bet" on 3.75 or 4.00 simply as an insurance hedge to protect their real investment exposure.
-- Jack Krupansky
Posted by: Jack Krupansky at June 14, 2005 03:22 PM
Wonderful blog, I have learnt a lot. I have been worried about the growth implication of high oil prices for a while. People always say "this time it is different", which I rarely buy. But your piece about how previous oil shocks came with significant production (thus were really "shocks") which is lacking this time makes me much less worried (although it also makes me worry more about inflation).
Now I want to suggest that "this time is different" on yield curves :-) My view is that yield curve, in particular, long-term rates have become much less informative in the past year or so. The reason is the foreign central banks' purchase. According to the latest Treasury Bulletin, for the year ended in March, net foreign purchases of Treasury debt roughly equals to the total issuance of Treasury securities that are available to the public (~$350B) for the same period (table OFS-2), and data from other sources suggest that purchases by foreign central banks can account for almost all of the foreign net purchase and most of the purchases are on longer term securities.
I think it is safe to assume that at least some of the foreign central banks don't care much about the narrow investment returns on their treasury holdings (in fact, some BOJ guy said this explicitly not long ago) and I think the probability is pretty high that the smart money is not fighting foreign central banks (these CBs can certainly continue to throw their money at the treasuries longer than a mutual fund can stay solvent), so the information content of the yield curve maybe particularly low this time around ...
Posted by: pat at June 15, 2005 12:31 AM
Note that the yield curve for municipal tax-free bonds (which foreign investors don't buy because they can't capture full tax benefits) is upward sloped in a pretty normal way. This supports the view that the flat treasury curve is a result of international investors parking money in treasuries.
Posted by: jult52 at June 15, 2005 06:40 AM
This last remark is important. Evidence is strong that foreign buying has kept long term rates low, presumably by Asians trying to keep the dollar from declining. Recent weakness of the euro might lead them to ease up on such purchases, which could push those long term rates up. What that might lead to is an end to the housing bubble.
Posted by: Barkley Rosser at June 16, 2005 10:08 AM
Prof. Hamilton: Great blog! If you believe that the economy was overstimulated until late 2004 when the funds rate reached 2%, and you think that there are long and variable lags, why would it be stupid for the Fed to take the funds rate to 4% -- which amounts to about a 2% real rate?
Posted by: econ 704 student at June 23, 2005 08:31 AM
My first point was that if the Fed had not overstimulated the economy in 2002 and 2003, we might not be in the position we are today of needing to raise rates. Because I believe that the Fed tends to overstimulate the economy at the end of a business cycle downturn and beginning of the upturn, and overcontracts toward the end of a business cycle expansion, that makes me an inflation hawk before the end of the downward moves in interest rates and an inflation dove before the end of the upward moves. But I don't see anything internally inconsistent about wanting to see a steadier hand from the Federal Reserve over the cycle.
But, if you want to take 2002-2003 as water under the bridge and ask, so what should we do now from this point, my concern is that the Fed needs to be careful not to raise rates too quickly. In particular, I was reacting to the market's assessment that the fed funds rate might be up to 4% by the early fall and that we would actually see an inversion of the yield curve. If the rate hikes were to come that strongly and quickly, it definitely would be a mistake in policy from my perspective. Historically, inversion of the yield curve has been one of the key warning signs that the Fed has gone too far.
I'm relieved that this week's fed funds options reported over at macroblog have all but abandoned that 4% funds rate possibility by October. As I said in my original post, I would be surprised and concerned if the Fed were to be that aggressive, and suspected that last week's fed funds options might have been calling it wrong. This week the market's moved more into my camp.
Posted by: JDH at June 23, 2005 09:20 AM