Never reason from a price change, part 423
Here’s Macroeconomics Advisers:
In principle, FOMC communications can be very powerful. If the FOMC could encourage the market to shift out its expectation of the time of the first rate hike by six months, the impact on the ten-year Treasury yield would be comparable to that of $760 billion of QE! Our analysis suggests that a six-month shift in the expected time of the first rate hike would have a significant impact on the yield curve.
But recall that the larger the monetary stimulus, the more quickly interest rates are likely to rise. This creates a problem. It very much matters whether the Fed’s decision to hold down rates is seen as an expansionary move, reflecting greater than anticipated monetary stimulus, or a contractionary move—the Fed throwing in the towel and adopting the Japanese monetary model of ultra-slow NGDP growth. Never reason from a price change.
It might be argued that the Fed has only a single policy, and hence there is only one way that they can affect the path of short term rates—buying bonds. But in fact the Fed has many tools, the most important of which are not current purchases of bonds in the open market. If the Fed promise to lower rates is tied to communication hinting at much more aggressive stimulus in the future, it is likely to be expansionary. If the low rate promise is accompanied by a disappointing refusal to commit to QE3, level targeting, or any other robust stimulus, it will may interpreted as a contractionary move (as in Japan.)
In the first half hour after the recent Fed promise to hold rates near zero, the move was interpreted as being highly contractionary. Later in the day markets (presumably) focused on other aspects of the statement, hinting at more expansionary policy to come. When they did so stock prices rose sharply, but so did interest rates!
New Keynesian models incorporate all these expectations channels, but I rarely see mainstream economists, even elite economists, talk about Fed policy moves using this sort of sophisticated model. Are there any recent analyses I am missing? I’d like to link to other economists who use the expectations channel of NK models (and financial markets reactions) in their analysis. Any links would be appreciated.
PS. Interested readers might want to look at Andy Harless’s different interpretation of Tuesday’s event in the comment section of the previous post. Both of our interpretations are consistent with NK models, although in my eyes his interpretation implies a greater degree of market inefficiency than mine. Other’s may disagree.
HT: Alex Tabarrok
Tags:
12. August 2011 at 05:53
“When they did so stock prices rose sharply, but so did interest rates!” Surprise, surprise!
12. August 2011 at 06:41
There are two ways the policy is expansionary. The Fed reduced the risk that they would raise short term interest rates in the face of a depressed economy in the next few years. There are people advocating that the Fed raise rates despite a depressed economy. It is a possibility, I suppose. The Fed communicated that it won’t do it.
The other approach involves an assumption that the Fed is lying. They committed to keeping interest rates low, and while they said it was because they expect a depressed economy, they are lying about that. They will keep interest rates low in the face of rapid real growth and/or higher inflation expectations. In other words,they will allow a boom to develop. This is the orthodox new Keynesian solution to the zero bound.
There are two ways it is contractionary.
The Fed has communicated to everyone that its cracker jack forecasters believe the economy will be depressed through 2013. This is evidence for households and firms to reduce their expectations for growth. I was expected a strong recovery next year, but the Fed guys are telling me I am wrong.
And, of course, there is the part you emphasize. The Fed is saying that they won’t take do what it takes to cause a recovery.
12. August 2011 at 06:48
If the market expects a particular forward path for the policy rate prior to the central bank committing to implement a lower path (i.e., committing to depart from the zero lower bound later than the market was expecting), doesn’t this raise inflation expectations, lowering real interest rates, thereby stimulating aggregate demand?
Is your claim that the market is interpreting the Fed to be holding its (implicit) inflation target fixed, the consequence being that lowering the expected forward path for the Federal Funds Rate must reflect expectations of real economic weakness rather than a commitment to temporarily higher inflation?
The reaction of the TIPS spread in the hours following the FOMC statement seems to confirm this story. Inflation expectations fell sharply in the immediate wake of the statement. While they began to rise shortly thereafter, they never recovered (on that day) to where they were prior to the statement. Such behavior doesn’t look like the market responding to a temporarily higher inflation target.
12. August 2011 at 07:57
No doubt I am biased – because I want to see both bernanke and Obama FIGHT the deflationists. This sickening passiveness on the part of both of them just drives me crazy.
I have come to believe that Obama cannot fight. He does not know how. Whenever he tries it just comes off as self-pity and whining. And, of course, support for the wonderful public-sector unions.
But informed market participants can read Bernanke in a different way. Bernanke knows that we (informed participants) know what he has said in the past, about Japan. And he knows that we have all been wondering why he has not acted in a Rooseveltian way up to now. And he knows that we participants can now realize that he has not been willing to act in this bold way because he has wanted to keep consensus on the Fed board – and has therefore not been doing what he would like to do but has rather been splitting the difference with the deflationists. And with the recent decision he has now announced that the deflationsists can just go get lost. He will accept their dissents and just go ahead with the Board majority that he knows he has. (Those two empty seats would be real helpful. Obama wont get off his ass to fill them, but Bernanke does have a firm majority without the – we hope.)
Of course it would have been better for Bernanke to announce what he wants – to state his aims. To say that it is his intention to keep short rates at zero until either unemployment falls to 5% or NGDP gets back up to where he wants it to be. But that was a step too far for him. For one thing the deflationists would likely have tried to impeach him – and he doesnt want that.
But I do think that his willingness to openly and directly ignore and fight the deflationists is a big step forward. He needs now to keep on going.
He knows there is plenty of cash in the banks and on the balance sheets of corporations. He is directly saying to them – SPEND IT.
I sure do hope I am right.
12. August 2011 at 08:03
Hi Scott.
To the extent that I may be a proxy for at least a certain segment of the market, this is what I was expecting/thinking:
1) The core catalyst for the fall in the markets beginning last Thursday was the debt ceiling deal and its implications for growth. I took it as It demonstrating that there was very little likelihood that those factors currently impeding restructuring and investment (and thus the recovery) – the nature and extent of government intervention – will be reduced or eliminated. People dump everything that isn’t money or money-equivalent. Margin calls kick in, etc.
2) To the extent that the S&P downgrade added any new information, it was at most a confirming signal that low growth is to be expected and, perhaps, an element of “gee, if even S&P is downgrading the US fiscal situation, it must really (really) be bad”.
3) The dramatically worsening situation in Europe is/was elevating money demand and dramatically so late last week and early this week.
4) In light of the severity of the situation, I (and I think a good chunk of the market) was expecting something more from the Fed in the way of accommodating the spike in money demand than “we promise not to tighten what may be an already too tight monetary policy”, whether that might have been QE3 or more of a short term accommodation, I wasn’t sure. However, given the criticism of the Fed from so many quarters, it struck me as likely that the Fed wouldn’t want to move on QE3 until the market and others were literally been begging it to.
5) The expectation in 4) was probably putting something of a floor on the market. The market fell initially because that expectation had been disappointed.
6) Then, based on commentary that emerged (some of written a day or two before the Fed announcement), I came to the conclusion that perhaps it had been unrealistic to have expected QE3 to have been announced so soon after the end of QE2, mostly for political/bureaucratic reasons, and that perhaps the announcement should be viewed as a signal of likely future willingness to take more dramatic action, once a sufficient amount of political or bureaucratic cover had emerged. Also, to the extent that the fall in stock prices was due to (European) liquidity concerns which to some extent were likely to be temporary (at least until they return on a more permanent basis), and given the “lumpiness” generally of policy formation and implementation, there is an argument on the Fed’s part to be made for waiting for the those aspects of the storm that are expected to be temporary to have passed so that the more permanent scale of the problem could be assessed.
12. August 2011 at 08:13
[…] A thought experiment inspired by this Scott Sumner post: […]
12. August 2011 at 08:15
I view monetary policy as a sort of retributive justice. When the economy offends by running too slowly, the Fed punishes it (specifically punishing savers who buy financial assets) with low (often negative) real interest rates. When the economy offends by running too fast, the Fed punishes it (specifically punishing those who own real assets and those who have over-spent and have nothing left with which to buy financial assets) with high real interest rates. Usually, the mere threat of punishment is enough to get the economy to behave itself: hence the relative mildness of interest rate fluctuations during the Great Moderation. But sometimes it doesn’t quite work, and the Fed has to implement the punishment (as, for example, in the early 1980’s or recently).
My interpretation of the Fed statement is that the Fed is saying, “OK, economy, we see that you’re determined to behave badly. We can’t stop you (at least not without doing something too radical to contemplate), but don’t think for a second that you’re going to get away with it. Here’s the whip, and we intend to keep on using it.” This has two implications. The first is that interest rates will be lower than the average agent previously expected, because anyone who doubted the Fed’s intention to punish has now been (one hopes) convinced. The second is that maybe, just maybe, the economy will react to the threat by behaving a little better than previously expected. If that happens, the punishment will be reduced, and interest rates will rise.
I think the market’s reaction to the Fed’s announcement is consistent with sorting out in its collective mind the relative prevalence of these two effects. You have a precipitous drop in interest rates immediately following the announcement — an indication that the punishment-doubters have been quelled. Then for the rest of the day you have the markets going back and forth as to how large they think the “deterrence” effect will be. When they think it will be strong, stocks go up and bonds go down. When they think it will be weak, bonds go up and stocks go down. At the end of the day, the conclusion was that the deterrence effect would be strong enough to be important (so the stock market went up) but nowhere near strong enough to outweigh the effect of the punishment itself (so the bond market stayed up, though not quite at its peak).
12. August 2011 at 08:45
[…] this subsequent post, Scott reminds us never to reason from a price change, and explains that a drop in long-term […]
12. August 2011 at 08:46
@Andy
” I view monetary policy as a sort of retributive justice. When the economy offends by running too slowly, the Fed punishes it (specifically punishing savers who buy financial assets) with low (often negative) real interest rates.”
It’s an odd form of retribution (and of “justice” for that matter) given that it’s generally monetary policy error (i.e., when it isn’t fiscally induced) which causes the economy “to run too slowly” to begin with.
12. August 2011 at 08:58
Here is the optimistic take – Bernanke was being bold and he will do more if necessary. As I said I do hope this is right.
http://www.marketwatch.com/story/bernanke-is-bold-now-its-obamas-turn-2011-08-12
One reason to think Bernanke might have been unwilling to state his aims directly is because he knows that a wave of deflationist criticism would in itself have been deflationary. Imagine having to listen to Richard Shelby and Rand Paul yell and scream and call for Congressional hearings.
12. August 2011 at 09:37
Maybe this explains why Bernanke did not go straight to QE3 – because he thought QE3 would just create straight panic.
http://ftalphaville.ft.com/blog/2011/08/12/652166/the-feds-secret-qe-equivalent/
The banks already have a ton of cash. To give them more, and to extract quality collateral while doing that – would just have lead to voluntary capital destruction, as happened in the first depression. Better to change expectations on the cash they already have rather than just give them more cash to just sit on.
But then it would seem all the more important for him to directly say what his real aim is. One hopes he will do that.
12. August 2011 at 10:35
Desperate and very frightened cash:
http://www.ft.com/intl/cms/s/0/d5733486-c42d-11e0-ad9a-00144feabdc0.html#axzz1UMJJrCPS
12. August 2011 at 11:02
JimP – read the second article. Couldn’t really follow it. I’ve not heard interest on reserves described as anti-deflationary before. Cutting through the arm waving, I read it as saying that QE3, by reducing the supply of treasuries (referred to non-ironically in the article as “quality collateral”), would increase the demand for money (base money) by banks (and others?) because they would never dream of loaning it for purposes of real investment. I’m not sure why that would represent a huge change from the current situation. I’ve been viewing the demand for treasuries as the big-investor (above FDIC limits) equivalent of holding money. So effectively, from the perspective of real investment and recovery, I don’t understand what would change.
However, I have for some time been of the view that the demand for money is elevated because fiscal policy and regime uncertainty (a la Robert Higgs) have, by polluting market signals, impeding market adjustments and raising government-related risk, undermined the business case for immediate productive investment and encouraged entrepreneurs to await the time when they can have more confidence in market data and their own expectations.
According to this logic, fiscal policy may not just be potentially wasteful and ineffective, but actually a negative force which undermines, and potentially neutralizes, expansionary monetary policy, particularly where some restructuring in the economy is required, by raising the demand for money. The question is whether, in such a case, anything short of scarily expansionary monetary policy can satisfy/overcome the high demand for money.
12. August 2011 at 11:32
Just as there are two possible paths after the fed sets a low interest rate, there are two possible results of the fed setting a higher rate. They can say “We are going to buy treasuries (and create inflation), until the rate rises to 2%”. Or, “We are going to sell treasuries (and restrain inflation), until the rate rises to 2%”. They just need to be clear about whether they are going to get to the target by buying, or selling.
Et voila, the zero bound is destroyed as there’s never any necessity for a negative rate!
12. August 2011 at 11:37
david
You said:
“The question is whether, in such a case, anything short of scarily expansionary monetary policy can satisfy/overcome the high demand for money”
I don’t think Bernanke wants to scare people. They are scared enough.
I do think (hope) he wanted to change expectations about the cash already horded here and there. There is plenty of cash around – just spend it.
But if that is what he wanted to do I do agree that he might have said so. I don’t think the deflationists could impeach him even if the wanted to.
12. August 2011 at 11:48
:
If it’s true that a 0% rate can indicate tight or loose money, the same would be true about any rate, no? Therefore all the fed needs to do is add one piece of information to their target rate announcements: whether they will be buying or selling treasuries until their target is achieved.
It’s not as ideal as level targeting, but at least you don’t have to shift the entire paradigm of the profession. It’s more in the realm of the possible. A conservative institution like the Fed is only going to approach the ideal with baby steps and this would seem to be a useful one.
12. August 2011 at 12:02
Hi JimP.
I wasn’t suggesting that Bernanke should try to scare people. In fact, if he did, it may well cause a flight from the currency.
Rather, I was wondering aloud whether, in our current situation, the monetary disequilibrium (in this case, excess demand for money) can be fully addressed by monetary policy and whether the problem would persist until the cause of the elevated demand for money (fiscal policy/regime uncertainty) was removed.
12. August 2011 at 12:08
JJ, The Fed doesn’t know if it will be buying or selling. The OMO’s are done to hit the target.
12. August 2011 at 16:46
david
More on the deflationary risks of QE3 – as perhaps perceived by Bernanke
http://ftalphaville.ft.com/blog/2011/08/11/650656/when-a-government-bond-becomes-a-giffen-good/
12. August 2011 at 19:28
@david stinson:
“It’s an odd form of retribution (and of “justice” for that matter) given that it’s generally monetary policy error (i.e., when it isn’t fiscally induced) which causes the economy “to run too slowly” to begin with.”
I think the analogy holds. If crime is rampant, it’s because the criminal justice system has failed. If the economy is running at the wrong speed, it’s because monetary policy has failed. A “monetary policy error” consists of failing to specify or credibly threaten the appropriate punishments for macroeconomic misbehavior. If the criminal justice system isn’t good at catching criminals and figuring out and proving who committed a crime, then crime becomes more common. Similarly, if the Fed isn’t good at diagnosing a misbehaving economy, the economy runs badly. (I’m no expert in sociology, so I won’t claim that my statements about crime are necessarily accurate, but I think the analogy makes sense.)
13. August 2011 at 07:41
Bill, Good analysis.
Liquidity trap, The Fed has many tools. If an announcement calls for a lower policy rate path than expected, but is otherwise much more contractionary than expected regarding other tools (QE3, etc), the net effect may be contractionary. In that case the fall in market interest rates on T-securities reflects slower expected growth. The fact that stock prices and bond yields moved in tandem, hour by hour, suggests that markets were focusing on something other than the liquidity effect of lower fed funds targets.
JimP, Yep. he’s not a fighter.
David, You said;
“The core catalyst for the fall in the markets beginning last Thursday was the debt ceiling deal and its implications for growth.”
Stocks rose on news of the debt ceiling.
You said;
“To the extent that the S&P downgrade added any new information, it was at most a confirming signal that low growth is to be expected and, perhaps, an element of “gee, if even S&P is downgrading the US fiscal situation, it must really (really) be bad”.”
Demand for T-bills rose sharply after the downgrade.
I agree with all of the rest of your comments.
Andy, Interesting analysis, but I see things differently. I don’t think the economy misbehaved in the 1970s, I think the Fed causes 11% NGDP growth from 1972-81. I think they corrected their own mistake. And I think the Fed caused the slow NGDP growth in recent years with tight money. And I think that slow growth caused lower rates. And I think interest rates are falling because NGDP growth expectations are falling because Fed policy is getting tighter in the only sense that matters–NGDP growth expectations relative to target. Not because money is getting easier.
jimP, I hope you are right.
jj, But if they just target rates, prices become indeterminate–exploding to zero or infinity.
13. August 2011 at 09:08
How is the Fed supposed to reduce current aggregate demand without raising real interest rates? Given that it is keeping the Federal Funds Rate near zero presently, the Fed would have to lower inflation expectations to reduce current aggregate demand. That would require it to raise the forward path for the FFR relative to expectations. But that’s precisely the opposite of what it promised to do!
If what you mean is that while it lowered real interest rates, it did so less than was necessary given a falling a natural rate of interest, then I concur. But it still did more than the market was expecting it to do in this regard, which should be expansionary from the market’s perspective.
As for QE, my view is that QE2 worked to the extent that it credibly committed the Fed to pursuing a lower path for the FFR than the market was expecting. The Fed was betting that longer-term nominal risk-free interest rates (which reflect the expected forward path of short-term nominal risk-free interest rates) would decline, but since they control that path, this bet functioned mostly as a commitment device. If the market was disappointed by QE3’s absence, on this view, then they should have been expecting a path for the FFR lower than what the Fed announced, but I don’t think that squares with the market’s reaction to the announcement.
14. August 2011 at 07:57
Liquidity Trap, Your analysis assumes interest rates are the Fed’s only policy lever. Not so. If they tighten one of their other policy levers enough, interest rates can fall at all maturities, and monetary policy can still tighten.
Suppose the Fed did less QE3 than expected. That could lower NGDP growth expectations, and hence lower interest rates. Then assume the Fed did more than expected on short rates. Short rates can fall due to the liquidity effect, while long rates are falling due to the income and inflation effects.
From a Wicksellian perspective: The Fed does something to reduce the Wicksellian equilibrium rate by 1%. The simultaneously cut rates by 1/2%. Money is tighter.
14. August 2011 at 10:45
This is where you’re losing me, I think: With the FFR near zero, the only way for the Fed to cut (real) rates by 1/2% is for them to raise inflation expectations by 1/2%. How, then, is the Fed supposed to simultaneously slow NGDP growth expectations, while it is raising inflation expectations? If it can only do one or the other, then it can only lower real rates, or lower the natural rate, but not both. I guess I’m not seeing how it could do both in one go.
15. August 2011 at 10:59
Liquidity trap, I was talking about the cut in the two year bond yield, which represented a cut in future short term interest rates (say 12 to 24 months out.) Obviously they can’t cut nominal rates below zero–I agree with that.
15. August 2011 at 12:05
Scott, we both know how the fed operates so I assume you misunderstood me when you said targeting rates would lead to an indeterminate price level. The fed sets the “federal funds target rate” now and prices are determinate. The rate is only a target, and the OMO are the actual actions taken to achieve the target. Obviously the rate target moves to achieve the dual mandate of inflation and employment.
If you would be so kind to address my real point, though: my interpretation of what you’ve been saying is that to achieve today’s 0.25% target they can either A) Buy, inducing tight money and low rates, or B) Sell, inducing loose money and low rates. Would not the same should be true for any rate, not just near-zero? Therefore, to escape the zero-bound without resorting to unconventional policy, the fed need only specify that they will ONLY be buying treasuries until the target rate reaches X%. As TheNumeraire mentioned, today the Fed doesn’t know if it will be buying or selling to achieve the target. This indeterminacy is the root of the zero-bound problem, and is also completely unnecessary.
To repeat myself: this proposal isn’t as ideal as level targeting, but at least you don’t have to shift the entire paradigm of the profession. It’s more in the realm of the possible. A conservative institution like the Fed is only going to approach the ideal with baby steps and this would seem to be a useful one.
16. August 2011 at 07:32
jj, I’m not sure it’s true at all rates. The near zero rate is different, as it is consistent to many very different levels of the base. On the other hand you can only get a 4% fed funds target with a very specific level of the base, unless I am missing something.
But yes, given the near zero level of the ffr, it would be better to increase the monetary base, rather than reduce the monetary base. Of course all this is complicated by IOR, which in practice is another tool being used to reduce NGDP.
16. August 2011 at 08:09
My hazy understanding of IOR is that they serve a valid technical purpose, something about the banks transferring all of their reserves out during the day and then back in at the last minute end of day. Of course you’re correct that in practice, today, they are reducing NGDP. However this should be addressed with the proper tools (NGDP futures targeting natch) and IOR should be left in place.
Nonetheles, removing IOR may be more politically feasible.
16. August 2011 at 19:49
jj, I have no problem with that.
7. May 2012 at 07:44
[…] they represent, rather than an outside interference dictated by a uniform entity. To paraphrase Scott Sumner, ‘never reason from an action of the state’ – that law or program has been […]
29. July 2012 at 20:02
[…] these two very different cases is an instance of the Sumner analytical principle–never reason from a price change. The 1873-1895 goldzone deflation was a shifting mixture of both elements. The 1929-32 deflation […]