The Fed did not have enough room to cut interest rates before hitting the zero lower bound when the recession hit. Raising the target inflation rate, which would increase average interest rates and give the Fed more space for rate cuts, is something the Fed ought to seriously consider.He's not quite as blunt as I might like, but he's saying that, if a central bank wants to hit a higher inflation target, it has to set nominal interest rates higher, on average. So, in the course of transitioning to a higher inflation target, the central bank must, at some time, have to raise nominal interest rates in order to produce higher inflation. But then, it must be true that, if the central bank has an inflation target of x%, and inflation is persistently y%, where y < x, then the central bank must raise its nominal interest rate target.
What's happening in monetary policy and macroeconomics.
Friday, August 26, 2016
We are All Neo-Fisherites
I was looking at this piece by Mark Thoma on increasing the inflation target. In some discussions of this issue, people seem to have a hard time getting to the core of the argument, but Mark does not. He has a good discussion of the Fisher effect, and his concluding paragraph is:
Posted by Stephen Williamson at 8:16 AM
Subscribe to: Post Comments (Atom)
Millennials won't buy at higher prices. If the Fed tries to set a higher inflation target without allowing some way for credit to enter the society, all that will happen is banks will get more IOR. So, main street may pay more for goods, while banks get richer. I understand the argument to avoid negative rates. But the Fed has screwed everything up by weakening the middle class by mispricing its real estate, and then stealing its real estate by allowing the commercial paper market and helocs to crater. What a Fed!!!ReplyDelete
You seem to be suggesting that if one accepts the notion of the Fisher equation, one is also accepting the NeoFisherite proposition. I'm not sure that's correct.
The debate does not concern whether the Fisher equation holds. I think the debate is more about the direction of causality. The tradition (monetarist) view is that an increase in inflation expectations (brought about say, by helicopter money) causes the nominal interest rate to rise (as bond holders seek to protect themselves against the higher future inflation). The NeoFisherian proposition is that the direction of causality is somehow reversed. That a persistent increase in the nominal interest rate will *eventually* lead to higher inflation. The part the people find confusing is that they don't understand the *mechanism* that causes inflation to rise. Simply stating that the Fisher equation is not a sufficient explanation for most people. What is the mechanism that you think is at work here? It would be good for you to lay it out explicitly for your readers!
David read my Regional Economist piece:Delete
Then, he answered his own question (edited somewhat - hopefully I got this right):
One way to think about this, based on our conversations on the subject, is to think how an increase in the nominal interest rate affects the opportunity cost of holding money. Obviously, the opportunity cost rises. When it does, people are compelled to economize on their money balances. Of course, collectively, people cannot get rid of their money balances and so what you get is a sort of "hot potato" effect -- the price of goods that people are willing to sell for money starts to go up, that is, the purchasing power of money has to fall, because people are trying to economize on their money holdings.
And, if you combine that intuition with the fact that, for an interest rate target the Fed must be accommodating money demand, then you're done. That is, while the demand for real money balances goes down, the *growth rate* in the supply of *nominal* money balances must go up -- the Fed is accommodating the nominal demand for money that comes along with a more rapidly rising price level.
It is possible that Williamson is getting to the wrong conclusion. But it is unfair to imply that he has given no thought to the direction of causation here. The guy has been utterly painstaking, although damned if I can follow it. My bad.Delete
I am often struck by how intelligent people can assume other intelligent people have not considered the most obvious things.
Fun times. Peace out.
A question about this sentence: "collectively, people cannot get rid of their money balances." Why not? The relevant instrument here is the IROR. If the Fed raises the IROR, the non-banking sector will get rid of their money by depositing it at the banks which will in turn deposit it at the Fed. So, given the existence of a central bank with a perfectly elastic demand for money balances at the IROR, I think that, in fact, people can collectively get rid of their money balances. I am not saying that the result does not hold, but either I am missing something or that mechanism does not seem to work.Delete
Just to add to my previous comment, it seems to me that the mechanism is the faster increase in Fed liabilities associated with a higher IROR. E.g., initially there may be a decline in the price level as spending goes down but over time, as deposits grow faster due to the higher interest rates, so will nominal spending and hence prices. Of course, that is pure intuitive hypothesizing (which is what I guess we are discussing here), not the result of some model.Delete
Hopefully I made it clear that what appears after the colon in my comment are David's words, not mine. The more I think about this, the less I like it. I'm not a big fan of "intuition" in the form of an anecdote about results in a model, where the anecdote has nothing to do with the model. In economics, typically the model is all we have to put some structure on what we're observing. The model makes some predictions, which will have something to do with the internal mechanics of the specific model. We can provide some words to describe what is going on, and typically those words will make reference to bits and pieces of the model that the reader will be familiar with in other contexts, so that the reader can understand what is going on. David's narrative isn't tied to a specific model, so I don't know what to make of it. Indeed, for NK models, it wouldn't make any sense, as there are no "money balances" in such models.Delete
Independent narratives are just that - narratives not tied to an explicit theoretical framework, spelled out in all its glory. We construct models for a reason - to ensure consistency and honesty.
"the price of goods that people are willing to sell for money starts to go up, that is, the purchasing power of money has to fall, because people are trying to economize on their money holdings..." Or instead of bidding up the price of goods, people bid up the price of assets, especially safe assets. Demand for goods fall, et voila, you're in a liquidity trap. This interpretation fits the facts without the messy causation problem that can only be solved with an iterative process that assumes a level of numeracy that is sorely lacking in the general population.ReplyDelete
But then again, for some, I suppose unrealistic assumptions are a small price to pay for ideological consistency.
Your comment illustrates the problem with what David is looking for, which is some "intuition" that will somehow convince people who are not on board with the idea. Any "intuition" that people have comes from thinking about some model that is familiar to them. In some cases, that seems to be some undergraduate IS/LM/Phillips curve model with fixed inflation expectations. In that type of model, tighter monetary policy makes the interest rate go up, spending goes down and, via a Phillips curve effect, inflation goes down. If that's your "intuition," then nothing will convince you that higher nominal interest rates make inflation go up. Checking our ideas for internal consistency involves constructing rigorous models, and trying to understand what these models have to say about the real world, and whether or not the models fit the facts as we know them. That's the level at which the convincing gets done, not in telling "intuitive" stories. I have no idea why you think the underlying assumptions behind the idea are "unrealistic," but the theory is what macroeconomists have worked on for some 40+ years - it's boilerplate.Delete
40 years ago, Prof, we didn't have maddening bond hoarding. We didn't have banks and counterparties fearful that a little rise in interest rates would force collateral to decline in price, causing the need to hoard even more bonds, pushing the long yields down even further.ReplyDelete
As far as NeoFisherism is concerned, the bond hoarding is what creates the conundrum, and Greenspan said this very thing in 2007:
"To be sure, the benefits of derivatives, both to individual institutions and to the financial system and the economy as a whole, could be diminished, and financial instability could result, if the risks associated with their use are not managed effectively. Of particular importance is the management of counterparty credit risks. Risk transfer through derivatives is effective only if the parties to whom risk is transferred can perform their contractual obligations. These parties include both derivatives dealers that act as intermediaries in these markets and hedge funds and other nonbank financial entities that increasingly are the ultimate bearers of risk."
Greenspan created, IMO, too big to fail, and put the risk squarely on the counterparties and off the TBTF banks. He wanted the S&L crisis to never happen to another bank. But what he failed to understand was that risk doesn't go away, it just goes to the counterparties and the clearinghouses that demand more and more collateral for derivatives.
So, I will bet you a dollar, Prof, that any efforts to raise short rates will continue the conundrum, which is really not a conundrum at all, but is exactly what Greenspan engineered to happen.
I honestly believe the Federal Reserve Bank knows long yields cannot go up, but it is in on this tantrum behavior.
The only thing I don't quite understand is if there is a real shortage of the treasury bonds as collateral, or if there are plenty of other bonds, like corporate bonds, which can be used with a haircut. I am thinking that Greenspan's investors don't want to pay that haircut premium, which is why Greenspan is now engaged in tantrums continually, to get weak hands to get rid of their long bonds.
I am sorry if I seem cynical, but how can anyone seriously not be cynical about this system.
It might be boilerplate in the abstract world of the theoretical orthodox economist. In the real world it is riven with rust and the rivets are popping everywhere.ReplyDelete
The real world is of course where the rubber hits the road. In this case, boilerplate works really well, haven't you noticed? We all live in the same real world. And if we know economic theory, so much the better.Delete
OK, OK. If you must have it this way, then we can all agree that if the Fed raises its target to 10%, it ought also to shift its dots up by around 8% too. And if the Fed were credible then bond yields would surely rise in tandem -- voila! -- rising interest rates and higher expected inflation, which should soon lead to actual inflation. I'm not sure that the Fed Funds needs or should be raised at t=0, until policymakers know for sure that expectations are being translated into price dynamics, but eventually they'll go up too.ReplyDelete
That said, I'm also not sure that this thought experiment validates neo-Fisherianism, which states the proposition that even without such a shift in target, an immediate 800bps hike in Fed Funds would, after perhaps a brief and cleansing depression, result in a happy inflation in what's left of the economy.
We should also watch carefully the evidence from Sweden, where -0.5% NIRP has been so far successfully reflating the economy, Japan, where years of zero rates and mild deflation coexisted, Brazil, where rates of 20% were followed by drops in inflation and conversely lower rates produced credit booms and price rises, and many other countries which provide empirical data to compare to ivory tower theories.
Well, you're not exactly doing objective empirical work here, Mr. Real-World-Not-In-Ivory-Tower. Sweden, after four years of sustained, approximately zero inflation, has had some since the beginning of this year, to the point where year-over-year CPI inflation is at 1.1%. I wouldn't claim success for that project yet, as what they're shooting for is sustained 2% inflation. Not sure what you're trying to say about Japan - that's 20 years of low nominal rates and roughly zero average inflation, followed by massive unconventional intervention producing little or nothing. Don't forget the Euro area (three years of inflation at about zero).Delete
Also, you need to understand that I'm not telling anyone they should raise nominal interest rates, and I recognize that doing so can have negative consequences for real economic activity (though I think you're exaggerating the likely effects). But, in economics you rarely get something for nothing. If you actually want higher inflation, you can't get it without higher nominal interest rates. If you insist on low nominal interest rates, you're insisting on low average inflation forever. Which maybe isn't so bad.Delete
I'm still trying to figure parts of this out (or at least understand it better). This isn't an attempt at a gotcha question, I just want to know the NF answer to this questionReplyDelete
Suppose in late 2008 the Fed announced that its new inflation target was 3% instead of the old 2%. If everything else it did "mechanically", that is the bailouts, QE, was identical what does the nominal interest rate look like now?
Well, during a financial crisis, I think we could agree that inflation targets go out the window. But, suppose that in 2012, when the Fed issued its first official statement about the 2% inflation target, that it had specified 3% instead. What then?Delete
1. My guess is that the course of policy would have been exactly the same. The Fed would just be further south of its specified inflation target.
2. What I think you're asking is: What would the path for the fed funds rate have had to look like so that the Fed was achieving 3% inflation today? There are potentially some subtleties, to do with the low real interest rate, and what accounts for that (it's possible that this means that increasing the nominal rate can permanently increase the real rate), but suppose we ignore those things. Then, assuming that 4 years is a sufficiently long time (I'll go with that) that the Fisher effect is one-for-one, the Fed would have had have started nominal interest rate hikes long ago. Given that the pce inflation rate is currently 0.8%, to hit 3% inflation now, the Fed would have had to be up to a fed funds rate range of 2.5%-2.75% for some time now.
#1 was what I was hoping to get an answer on, but #2 is helpful as well. Can I interpret this to meanthatexpectations plays a much lessor role in MP than in Other monetarist's positions (SS has said that the Fed can get higher inflation just by announcing a higher target, an I believe that Nick Rowe has agreed with that position).Delete
If just saying it made it so, that would make policy really easy, don't you think? A specific inflation target has to come with a policy rule that will achieve the target, coupled with actions as dictated by the policy rule. If a central bank is not achieving its target, on average (you have to allow for uncertainty) then:Delete
(i) it's got the wrong rule.
(ii) it's not following through on the actions dictated by the rule.
So, we observe many central banks missing inflation targets (usually 2%) on the low side. Some people think the problem is (i), but that central banks aren't doing enough unconventional policy - they want more or different QE, or they want negative rates, or more negative rates. Some people think it's (ii) - if the central banks did what they said they would do, everything would be OK. My schtick is that it's (i), but they've got the sign wrong on the nominal interest rate.
Do you have a post on the mechanics? Not the equation, but how real world transactions turn the higherfed rate into higher inflation?Delete
Wouldn't 1940s-style financial repression do the trick? Don't touch the nominal rate, but shove a negative two percent real rate down people's throats.Delete
Well, negative real rates could in some sense be "financial repression." If it's a global safe asset shortage, there are ways to solve that. But, indeed, you can have low nominal interest rates and satisfy 2% (or 4%) inflation targets if the real rate is low enough - due to sufficient repression.Delete
Baconbacon: Before we get to that, what do you think causes inflation?Delete
Seems to me, the only time we may have been in a liquidity trap, this is how we got out of it.Delete
Of course, there's a caveat: it's hard the measure the collateral damages from financial repression. It came at a time when the population was young and growing, trade barriers were dropping fast, and both our countries' trade partners were war-ravaged, thus fast growing, Europe and Japan. Government back then could enact a lot of wrong policies without paying much of a price.
So instead of "raise rates to raise inflation", a Neo Fisherian slogan could be "kick banks to raise inflation"
"Baconbacon: Before we get to that, what do you think causes inflation?"Delete
Lots of things? For a short answer I would modify Friedman's quote to "inflation is often a monetary phenomenon", but I don't have a good answer that I have confidence in.
baconbacon: Well, in this case we're talking about the "monetary phenomenon," as the level of nominal interest rates is driven by policy. A higher nominal interest rate has to be supported by the appropriate monetary policy. I'll just describe the Fisher effect, as this makes it simpler - just suppose real rates of return are constant. If the central bank announces a higher nominal interest rate today on nominal assets (leaving aside which nominal assets we're talking about - just think of these as assets for which the payoffs are defined in terms of dollars), this implies that, given the anticipated rate of inflation, the real rate of return people would anticipate on nominal assets makes holding those assets very attractive relative to any assets for which the payoffs are actually defined in real terms. So that asset markets clear, the central bank has to support its nominal interest rate announcement by increasing the future money supply through a future open market purchase of government debt - the money supply goes up between today and tomorrow. That's important, and maybe that's why people are confused about this. The higher money stock in the future will increase prices in the usual quantity-theoretic fashion. With a stock of excess reserves outstanding when the nominal interest rate increase is announced, things work in excactly the same way, except that no open market operation is required. In the future, the nominal stock of currency will be higher - this causes the inflation, through the usual means. But in real terms the quantity of currency is lower (if the nominal interest rate is higher permanently) and the quantity of reserves is higher.Delete
Belanger: I didn't think you were serious about the financial repression. Surely this can't be a good policy. It can make inflation go up, but it seems economic welfare goes down.Delete
It cannot. My point was that this would be possible in a 1950s economy when things were so good that governments could do things that lower welfare without visible effect. I would not recommend it today (I thought my caveat made that clear, my fault). Last line was a joke: what a Neo Fisherian slogan could have been in the 50s. I thought it was good, again my fault.Delete
I realize Fed officials should never joke about stuff like that. I should have been careful.
Now I get the joke. Actually, Fed officials need humor as much as anyone else.Delete
Several questions/issues come to mind (only thinking of them now, so forgive any gross errors I make, and I appreciate the responses and the time)ReplyDelete
"So that asset markets clear, the central bank has to support its nominal interest rate announcement by increasing the future money supply through a future open market purchase of government debt - the money supply goes up between today and tomorrow"
1. Many economists claim that at or near the ZLB government bonds are money (or are money like, or function like money in some settings), CB purchases of new government debt are (under this assumption) swapping out money for moneyish assets, and so has limited to no effect on the quantity of money. Back to my question about announcements and expectations, if you are near enough to the ZLB and announce an interest rate increase but cannot effect the MS because of the above, and you don't get a boost just from announcing how do you achieve "lift off"? If you accept that premise I guess you end up in a situation where you are close to Cochrane's (and others) FTPL, where the Treasure controls monetary policy through its issue of new debt, and the CB is effecting only the distribution and not quantity. If you reject that premise, why?
2. During a period of increasing interest rates I thought it was traditionally done through the sale of government bonds, which reintroduced them to circulation, increasing the supply and forcing the government to offer a higher rate to attract investors. Purchases of bonds should drive rates down in a vacuum, right?
"With a stock of excess reserves outstanding when the nominal interest rate increase is announced, things work in excactly the same way, except that no open market operation is required. In the future, the nominal stock of currency will be higher"
Excess reserves implies that there are more willing lenders than willing borrowers at current rates, increasing the interest rate should decrease the demand to borrow while increasing the desire to lend. An interest rate hike (in the short term) should drive excess reserves up, not convert them to currency.
1. At or near the ZLB doesn't matter. Note, in fact, that as long as there are excess reserves outstanding, then given the interest rate on reserves, there's a liquidity trap - swaps of reserves for t bills don't matter. But changing the interest rate on reserves does.Delete
2. When I told the story, I left out the liquidity effect, which is what you're getting at, I think. Part of the short run response can be that, if the nominal interest rate goes up, there is intertemporal substitution - people want to consume more in the future relative to today. In some theories that happens through a reduction in current consumption, and in others it's an increase in future consumption. In any case, that's a short-run increase in the real interest rate. As long as that's smaller than the increase in the nominal rate, inflation has to go up. But, if you just think of level effects, if there were a one-time open market sale of government bonds, the liquidity effect tells you that the price of bonds falls and the nominal interest rate goes up. But that's different.
"Excess reserves implies that there are more willing lenders than willing borrowers at current rates..."
"Excess reserves" means reserves in excess of reserve requirements, willingly held.