There used to be a debate over interest rate control vs. base control of monetary policy. We always knew (at least some of us always knew) that interest rate control couldn't work in theory, but it seemed to work in practice, so eventually even the die-hard defenders of base control quietened down, or were ignored, and we all became "horizontalists" in practice. Modern central banks conducted monetary policy by setting a short term interest rate, so the LM curve, at least in the short-run, was horizontal.
It is time to revisit that debate. We need to understand why interest rate control cannot work in theory, why it seemed to work in practice, and why it now no longer works in practice.
I am following the same theme as my earlier post on whether quantitative easing is trying to raise or lower interest rates, and as many of Scott Sumner's posts, like this one, and especially this comment.
Why can't interest rate control work in theory?
It has long been known that for every equilibrium of a monetary economy there exists another equilibrium (or a whole set of equilbria) in which all real variables are the same as in the first equilibrium and all nominal variables are different in the same proportion (multiplied by some scalar k). Don Patinkin's "Money, Interest and Prices" formalised this insight, but it has been known at least since David Hume's essay "On money". This insight need not rest on any particular theory of the economy, and could even be re-formulated to apply to disequilibria. It rests only on the idea that real variables, not nominal variables, are what matter. All behavioural functions are homogenous of degree zero in real variables (including the real stock of money). Monetary units should matter no more than whether we measure in metres or centimetres.
If you take this homogeneity insight, and add the assumption that the supply of money is exogenous, you get the Quantity Theory of Money (a change in the supply of money will cause an equi-proportionate change in all nominal variables), and the Neutrality of Money (a change in the supply of money will affect no real variable).
Post-Keynesian horizontalists (and we are all horizontalists now, unfortunately, because that's the underlying problem) reject the Quantity Theory because they reject the assumption that the supply of money is exogenous. But that misses the point. A revised Quantity Theory can be re-formulated taking any nominal variable as exogenous — the price of gold, or nominal GDP futures, for example. The homogeneity insight does not depend on any definition of money supply being exogenous.
One implication of the homogeneity insight is that interest rates stay the same when all nominal variables change. This is true for both real and nominal interest rates, provided the whole time-path of present and future nominal variables changes by some scalar k. (We think of the nominal interest rate as being a nominal variable, but it isn't really, because it has the units 1/time, rather than $; instead it represents the rate of change of a $ nominal variable, just like the rate of inflation.)
If we think of interest rate control as the central bank setting the time-path of the rate of interest, the price level is indeterminate. There is nothing to stop the price level (and the stock of money, and all nominal variables) jumping onto any of the other equilibrium time-paths.
Monetary policy via interest rate control can't work in theory.
Why did it seem to work in practice?
The standard argument of why interest rate control works in practice is that prices are sticky. The economy can't jump from one nominal time-path to another. It takes time for monetary policy, seen as a gap between the interest rate set by the central bank and the neutral or natural rate of interest, to affect inflation. Provided the central bank could adjust the rate of interest more quickly than prices can adjust, the central bank can keep the price level determinate. If the actual rate is below the natural rate, and inflation rises relative to target, the bank must quickly raise the interest rate in response, and raise it more than the increase in inflation (the Taylor Principle), so that the real rate of interest rises relative to the real natural rate, to bring downward pressure on inflation. Provided the bank could respond faster than inflation, and faster than expected inflation, interest rate control could work in practice.
Interest rate control is like riding a bicycle. You can't keep the steering fixed and expect to stay upright. You need to keep moving the steering, and move it faster than your tendency to fall over, if you want to stay upright. And also like a bicycle, you need to steer left if you want to turn right. If you want higher nominal interest rates you first need to lower interest rates, so that inflation starts to rise, and expected inflation starts to rise, at which point you can raise interest rates, and raise them higher than originally, so that inflation and nominal interest rates eventually settle down at some new higher level.
That's how it was supposed to work in practice. Why did it stop working in practice?
I think the short answer is that something moved more quickly than central banks' response to that something. But what exactly was that something?
The first and obvious candidate for the something that moved too quickly is expected inflation. If the actual rate is above the natural rate, actual inflation will fall, and expected inflation will fall. If expected inflation falls more quickly than the central bank can lower the rate of interest, it will continue to fall, because the gap between the actual rate and the nominal natural rate gets wider and wider. The system is unstable.
With a credible commitment to an inflation target, that's not supposed to happen in practice. Expected inflation is supposed to remain "well-anchored" to the inflation target. But what that means is that the public trusts the central bank to move quickly enough to validate that trust. If the public believes that the central bank is too far "behind the curve" to validate those expectations, it would be irrational to continue to believe that future inflation will stay at target. And as soon as expected inflation comes adrift from the target, the bank must move even more quickly to get ahead of the curve.
There are always lags in monetary policy. It takes time for the bank to realise that the actual rate is above the natural rate, to lower the actual rate, and for the lower actual rate to stop the disinflationary pressures. If people trust that the bank will eventually get ahead of the curve, the curve stays fairly flat, and it is easier for the bank to get ahead of it. But if people lose that trust that the bank will eventually get ahead of the curve, that loss of trust makes the curve steeper, and makes it harder for the bank to get ahead of the curve. The loss of trust can become a self-fulfilling equilibrium.
Is it just a coincidence that the Federal reserve does not have an explicit inflation target, and yet is the most important central bank in the world, and at the "epicentre" of the financial crisis? It is easier for expected inflation in the US to become un-anchored, because there is no explicit anchor in the US.
A second candidate for that something that moved too quickly is the real natural rate itself. If the actual rate is above the natural rate, it is not just inflation that will be affected. Real economic activity will also fall, and will be expected to fall. When people expect future real demand for goods to be lower, investment demand falls (the accelerator), and real consumption demand falls (the multiplier), so the real natural rate also falls.
If the actual nominal rate of interest is above the natural nominal rate, the gap widens for two reasons: expected inflation falls; and the real natural rate falls.
Again, if people trust that the central bank will eventually get ahead of the curve, expectations of future real activity will remain well-anchored, just like expected inflation. But if people lose that trust, the curve gets steeper, and it becomes harder for the bank to get ahead of the curve.
But there have been many times in the past when central banks needed to lower interest rates, because they saw disinflationary pressures, indicating that the actual rate must have been above the natural rate. Why did it not always end in tears?
The financial crisis must have been the final straw that exposed the inherent instability of interest rate control. When the financial system is fragile, and bits of it break, the natural rate moves more quickly than normal, and the financial system also becomes more fragile when the actual rate is above the natural rate. Falling expected inflation, and falling expected real activity, weaken an already fragile financial system, and this lowers the natural rate by more than it would with a robust financial system.
With a strong financial system, maybe the inherent instability of interest rate control is a problem in theory, but not in practice. The curve is flat enough that people's trust that the central bank can eventually get ahead of the curve is a self-fulfilling trust. But with a weakened financial system, maybe the curve is just too steep. People lose trust that the bank can eventually get ahead of the curve, and this loss of trust makes the curve steeper still.
We knew that implementing monetary policy via an interest rate instrument was an unstable control-mechanism in theory, but it seemed to work well enough in practice, so we shrugged our shoulders and kept doing it. But it only worked because we trusted it to work. There were always two equilibria: one with validated trust in which prices were determinate and fairly stable; a second with validated mistrust in which prices were indeterminate. We were lucky to be living in the first equilibrium. The financial crisis eliminated that first equilibrium. It will be hard to go back to it, even when the financial crisis is over.
We need to switch to some other instrument for monetary policy, one which does not cause that inherent instability. The homogeneity insight tells us that any nominal variable (anything with $ in the units) will work in principle, and make the price level determinate. But some will work better than others. I am not advocating a return to base control, or the gold standard. Though each of these would make the price level determinate, the equilibrium real money base, and real price of gold, are too variable. I've written a long enough post, so I am going to leave that second question unanswered.
But whatever nominal variable we choose for central banks to buy and sell and control the price of, we could use immediately for quantitative easing, as well as permanently for the new monetary regime. Zero nominal interest rates do not mean that monetary policy is powerless; they just mean we were holding onto the wrong monetary lever. That damned lever has a very peculiar connection to the price level, and in the long run has no connection at all.
Worthwhile Canadian Initiative 
(Of course I come from variable rate Australia, so that biases my view).
I don’t think I quite rejected it as a policy lever but I thought it wasn’t what we wanted to pin all our hopes on.
From my side though this is largely a theoretical arguement, but an important one. Sumner thinks that it all works through the quantity of money and worrying about interest rates is dangerously misleading. I think this is all wrong, that it mostly works through the REAL rate and that worrying about the quantity of money is dangerously misleading.
Finally though, my story was exactly the textbook story Nick gave, just rephrased. I don’t see how it’s controversial. Now Euler equations don’t specify causality so if you change AD and then that changes the real rate via the Euler equation it may not be that real rate was the transmission mechanism. But all the simplest, most basic models (like IS-LM and the New Keynsian ones) work by changing the real rate and having that cause the change in AD just as Nick himself described. I also think that’s by far the most theortically coherent story and I think we should keep that in the front of our minds.
JKH:
If I am thinking of the same case, what looked like a refusal to engage in argument was really a mutual failure of communication, so that each saw the other as failing to engage. I know your blogging well enough to know that your knowledge and way of thinking is very different from academic economists like me. (On a good day, I can remember what “CDS” stands for!)
To give you an example: I am very relieved that you liked my fine-tuning/gross-tuning description of the settlements balances/OMO question. I was scared I might be totally wrong about something I ought to understand. I find it very hard to follow practitioners’ (like David Longworth’s) detailed descriptions of the workings of the settlements system. I can only understand such things at a very abstract oversimplified level. But that same need to abstract and oversimplify gives me perhaps an advantage in other ways.
“But somewhere I read that the international gold standard sort of failed in the 30’s because the Fed more or less attempted to engage in a type of OMO quantitative easing that was essentially incompatible with maintaining the gold level. Does that sound like a remotely accurate interpretation?” Yes. And more than remotely accurate. It is also why fixed exchange rate systems commonly break down. But there’s also a deeper, underlying reason. Why do central banks follow OMO policies that are incompatible with maintaining a fixed price of gold (or fixed exchange rate)? Because they have ultimate implicit targets (like preventing mass unemployment, or keeping inflation at roughly the right level, or printing enough money to finance the government deficit) that are incompatible with the stated target of maintaining a fixed price of gold. They try to hit both targets at the same time, and the system eventually collapses. In the long run, monetary policy can only have one target. (And not everything can be a target either, as we learned in the 1970’s, when we tried to use monetary policy to target unemployment, a real variable).
Back to your main question. The economy is like a box, with lots of levers sticking out of it. But all the levers are connected up inside the box, with a clockwork mechanism. So if the Bank grabs hold of any one lever, and moves it, all the other levers will move as well. So isn’t it arbitrary which lever the Bank grabs hold of?
It depends on the clockwork. As a kid, I had a pedal powered fire engine. The pedals were on a crank. You alternately pushed the right then left pedal to make it go. But half the time, if you pushed the wrong pedal, it would go backwards. And a few times, when one pedal was at the exact bottom, and the other at the exact top, pushing on either pedal would do nothing. So I pushed the rear wheel round by hand. Interest rate control is like pushing on those pedals. Buying plywood (or whatever) is like turning the rear wheel.
If the bank starts buying plywood, raising the price of plywood, and tells everyone it is going to continue to raise the price of plywood, nominal interest rates will rise.
If interest rates rise because the bank reduces settlement balances, that’s a tightening of monetary policy (the fire engine goes backwards). If interest rates rise because the bank buys plywood, that’s a loosening of monetary policy (the fire engine goes forwards). Exactly the same rise in interest rates in each case, but very different implications for the rest of the economy.
I must work on my metaphors.
I apologise for getting very behind in responding to the many good comments.
Nick,
“Why do central banks follow OMO policies that are incompatible with maintaining a fixed price of gold (or fixed exchange rate)? Because they have ultimate implicit targets (like preventing mass unemployment, or keeping inflation at roughly the right level, or printing enough money to finance the government deficit) that are incompatible with the stated target of maintaining a fixed price of gold. They try to hit both targets at the same time, and the system eventually collapses.”
That makes sense to me, as does the idea of interdependent levers. Let me apply/extend this to the point I think I’m trying to make.
First, there may be an unnecessary complication in the hypothetical case of a central bank using the futures markets rather than the cash markets for its intervention. So let me first use the case of cash plywood instead. In effect, this is more comparable to the case of gold or the case of fiat money where the central bank typically funds the fiscal deficit by the accumulation of treasury bonds as assets. All of these cases can be analysed easily in terms of the effect on the central bank balance sheet and the cash flows that go through it. I think that the root of the communication challenge in relating to academic models is that I tend to focus on balance sheets and balance sheet management. Moreover, Fed speeches have typically been based on an embedded importance of understanding the Fed as a balance sheet operation, or an asset-liability management operation as it’s conventionally known in commercial banking. It’s very interesting to me that the Fed is focusing in recent speeches even more intensively on communicating its policy by explicitly explaining its operations in the context of its balance sheet. I’ve linked earlier in this thread to two such speeches made in just the past several days. Bernanke and Company are now conducting a surge in heavy duty balance sheet explanation. I relate well to this type of presentation. So for me the challenge is to try and translate the academic models/proposals I’m trying to understand to the prototype central bank balance sheet they imply.
So the question becomes, what are the implications of cash plywood intervention for the Fed’s balance sheet?
“If the bank starts buying plywood, raising the price of plywood, and tells everyone it is going to continue to raise the price of plywood, nominal interest rates will rise. If interest rates rise because the bank reduces settlement balances, that’s a tightening of monetary policy (the fire engine goes backwards). If interest rates rise because the bank buys plywood, that’s a loosening of monetary policy (the fire engine goes forwards). Exactly the same rise in interest rates in each case, but very different implications for the rest of the economy.”
I disagree with you here, Nick; I think conventional easing and plywood easing should have comparable yield curve effects. Let me explain.
Let’s talk normal times first, above the zero bound.
Suppose the Fed is using its normal interest rate intervention. How does it normally “ease”? It lowers the target Fed rate. The bond market reacts. If the market thinks the Fed is too easy, it will expect inflation, so the yield curve will steepen. In fact, the default response is close to a steepening simply because the short rate drops.
Now, suppose the Fed uses cash plywood intervention. It “eases” by purchasing plywood.
But here’s the important question – how does the purchase of plywood affect its balance sheet? The corollary question is how does it affect the clearing balances that banks hold with the Fed; and further, how does it affect the overnight rate? And then, how does it affect the yield curve? My contention is that there’s a fairly clear mapping between target Fed rate intervention, and cash plywood intervention.
Here’s how I would see it:
The Fed intervenes in the plywood market. Why? To respond to its concerns about the same type of economic variables that it responds to in the case of conventional interest rate intervention.
Pick a variable. Suppose its nominal GDP. We want to get nominal GDP up. So the Fed “eases” by buying plywood.
What happens to its balance sheet? That depends on what the Fed wants to happen.
What does the Fed want to happen to its balance sheet? And why?
It’s got a bunch of plywood in inventory. It’s paid for the plywood by cutting a cheque. What’s it going to do with the potential increase in bank settlement balances that the cashing of those cheques will produce for its own balance sheet and for the balance sheets and reserves of the banking system?
What good is a huge inventory of plywood in terms of having an effect on nominal GDP? Well, if the Fed accumulates the amount of plywood that China accumulates in dollar reserves, probably quite a bit. But that’s silly. The Fed isn’t after “quantitative tightening” in the plywood market per se in order to produce inflation in the economy at large. It’s using the price of plywood as a signal, not the quantity of plywood.
But the signal is no good unless there’s substance behind it.
What’s the substance?
It’s not quantitative plywood tightening. I claim that’s silly.
So what is it?
I claim it’s the effect the Fed wants for bank excess reserves and therefore for the overnight rate.
This is the same sort of effect that the Fed aims for in conventional easing.
The Fed wants to ease, and it’s signalling to the market that the degree of its easing will be known via its actions in the plywood market.
Now technically, the Fed will probably sterilize almost all of its quantitative plywood tightening. This is because it needs only a miniscule amount of excess reserve easing in order to produce a desired change in the fed funds rate.
The net result will be a net amount of excess reserve easing and fed funds rate decline that could have been achieved by fed funds rate target.
The signal is different – the price of plywood versus the target fed funds rate.
But the NECESSARY intervention is the same – the “effective” (as defined) fed funds rate.
So the yield curve effect, my point of disagreement with you above, should be exactly the same.
The difference is that plywood intervention requires a real economy intervention as a demonstration of an intended knock-on interest rate intervention. Fed funds targeting requires only an announcement of the target rate and interest rate intervention to manage the effective rate according to target. But both amount to the same type of interest rate intervention at the end of the day.
This is also why now we can say that the plywood futures market would be a realistic alternative to the plywood cash market. Either way, the ultimate deployment of fed cash would mostly be sterilized, since the necessary excess reserve adjustment would typically be small relative to the policy outcome in terms of the fed funds rate. Using the cash plywood market as a price signal would probably use up a lot more cash to achieve the desired economic intervention, compared to futures.
Either way, plywood or direct fed funds, the modus operandi can be used in an attempt to influence any economically important variable – inflation, real growth, nominal GDP, broad money supply, etc. E.g. The Fed can try to encourage future inflation (at the margin, or 1st derivative) or nominal GDP. It’s just that plywood intervention gives the additional transparency of specific price targeting and indexing, versus the looser and more eclectic methodology that the Fed uses today.
Zombies eat money, not brains.
The almost 1:1 correspondence between excess reserves and the increase in base money seems to my non-economist brain to be a very telling symptom that something is really broken somewhere. Has there ever been a liquidity trap without zombie financial intermediaries? Banks where a mess during the GD, they where a mess in Japan, and the biggies are a mess today in the US…
Oops. Banks were a mess.
As Koo points out in his balance sheet recession arguement and as Krugman pointed out in his “It’s Baaack” paper, the 1:1 correspondence between excess reserves and and the increase in base money doesn’t necessarily mean something is wrong with the banks. It could just reflect the lack of aggregate demand, the lack of demand for credit or real balances. The banks are keeping excess reserves because nobody else wants the money.
However, to my mind it does show that Sumner is 100% dead wrong in his insistence that money is what matters.
The 1:1 correspondence between excess reserves and the increase in base money results from the fact that the Fed is using excess reserves as its liability instrument of choice in funding its own balance sheet expansion.
The extraordinary level of excess reserves is far more important as a means of funding for the Fed, than it is as statement of inference regarding the condition of the commercial banks.
I agree at least directionally with the connection to Sumner’s view.
JKH, but does that answer the question of why the money isn’t loaned out? Or are you agreeing with me that their’s just no demand for it?
Adam,
That’s a more complicated question. I’m not ducking it, but one of the things to keep in mind that given the fact that the Fed determines the level of excess reserves in the system, there’s no way of determining whether or not the money is being “loaned out” simply by looking at the level of excess reserves per se.
The level of required reserves is about $ 50 billion right now. That’s $ 50 billion supporting a banking system with a balance sheet size of about $ 12 trillion, I think. The excess reserve level has averaged around $ 700 billion or so in the past few months. So if as the accepted wisdom goes according to some, the banking system were to “use up” these excess reserves by lending and generating new deposits (as per the magnificent multiplier), thereby converting excess reserves to required reserves, then the US banking system would have to expand to an aggregate size of about $ 170 trillion in order to do it. This is absurd. But so goes the thinking of those who don’t reflect a lot on the balance sheet proportions of what is actually happening here.
That said, and more simply, I can’t disagree with the idea that credit demand is an important factor, but I think you have to slice and dice a lot of cross currents in the overall profile of current credit flows.
“… the Fed is using excess reserves as its liability instrument of choice in funding its own balance sheet expansion”
From here: http://www.frbsf.org/news/speeches/2009/0104b.html
Yellen says:
“… the Fed, like the Bank of Japan, has increased the quantity of excess reserves in the banking system well above the minimum level required to push overnight interbank lending rates to the vicinity of zero. The creation of such a large volume of excess reserves, in the Fed’s case, results from the enormous expansion in the Fed’s discount window lending, foreign exchange swaps, and asset purchases.”
Ok then … Could someone please explain the mechanics of this? If excess reserves are in fact not measuring excess reserves (i.e money in excess of required reserves the banks are keeping in the proverbial vault), then how can we get a read on , ummm, excess reserves?
Some interesting tidbits of bank data… See for yourself here: http://research.stlouisfed.org/fred2/
Total Consumer Credit Outstanding
Securitized Total Consumer Loans
Total Revolving Credit Outstanding
Bank Credit of All Commercial Banks (% Chg. from Year Ago)
Loan Loss Reserve / Total Loans for all U.S. Banks
Adam P @ 1.45, 1.46, 1.49, and 3.45:
The “Keynes effect” story of the downward slope of the AD is indeed the standard, Keynesian, interest-rate only channel of the transmission mechanism. It assumes away other possible channels.
If the price level fell, and the expected future price level fell by more, then we do have two offsetting effects. The fall in the price level increases aggregate demand, but the fall in expected inflation reduces it. The net effect can go either way. And this is because the demand for money can either decrease (because of falling P) or increase (because of falling inflation), So we can get either an excess demand or an excess supply of money.
Let me try another thought-experiment: According to the standard interest-rate only story, A has an excess supply of money, so goes (only) to the bond market, bids up the price of bonds, and forces down the rate of interest. B sees the lower rate of interest, and so decides to increase consumption and investment. Now, suppose A and B are the same person?
I’m getting way behind. Might just have to start a new post!
Re: excess reserves
There’s no contradiction. The Fed decides on the level of excess reserves it wants for the banking system as a whole, and then creates them. It simply buys an asset (or lends money) and credits the seller/borrower’s bank’s clearing account at the Fed.
The Fed’s been creating excess reserves aggressively through its various credit programs and balance sheet expansion.
The important point is that the banking system itself as a whole doesn’t determine its own excess reserve position; the Fed does. From there, individual banks compete for their required share of reserves. The competition in normal times determines the actual trading level for the fed funds rate.
Now, with the Fed’s extraordinary credit easing, there’s way more than the usual level of excess reserves in the system – by a factor about 300 or so. The Fed has done this deliberately. It’s paying interest to the banks on these reserves to compensate them for the fact that all of that excess is basically useless to them.
There’s a misconception about the role of excess reserves. In normal times, it’s simply a system buffer to have smooth trading in fed funds. Normally, the banking system doesn’t actually need excess reserves on hand to lend. That’s because the Fed has the discretion to create new reserves whenever it wants. It can always supply any new requirements generated by new banking system lending and related new deposit creation. The result is that the ongoing operational requirement for system excess reserves is very small – normally about $ 2 billion – in a $ 12 trillion banking system.
These are not normal times. The Fed has expanded $ 2 billion to over $ 700 billion. The banks don’t need it operationally to lend. But the Fed has created it as the funding offset to its own asset expansion (excess reserves are a balance sheet liability for the Fed; i.e. a source of funding).
JKH @ 12.27: “My contention is that there’s a fairly clear mapping between target Fed rate intervention, and cash plywood intervention.”
But the Hume/Patinkin homogeneity thought experiment tells us this cannot be true in the long run (and so, when we introduce expectations of the future, it cannot be true in the short run either).
Start in one equilibrium. There exists a second equilibrium, in which the price of plywood is doubled, all other nominal variables are doubled, the stock(s) of money are doubled, and yet all rates of interest are the same.
There is no long-run mapping between levels of interest rates and levels of nominal variables.
There is a long run mapping between rates of change of nominal variables and (nominal) interest rates, but the mapping is perverse, because higher rates of inflation map onto higher nominal interest rates.
The question is, how to translate from these long-run mappings (or non-mappings) into short-run mappings, when inflation adjusts slowly, and expectations (of inflation, real output, and the health of banks, etc.) are endogenous.
(Not sure how much this is helping you, but it’s helping me get my head clearer.)
Nick,
I don’t see that your latest though expirement counters my story but that’s ok. It does bring up another point which I’ll lay out. Before I said that the marginal dollar of excess reserves goes to the bond market because people are minimizing their real balances subject to being able to fund desired consumption (plus possibly a buffer against surprises). The formal model in the back of my mind that generates this story has money entering via the cash-in-advance constraint. As you get caught up you might counter with the portfolio choice story in which people are indifferent at the margin to putting an excess dollar towards the bond or goods market. Well, this won’t save the monetary without interest rate channel story.
Why not? If it’s monetary policy we’re talking about then we DON’T mean an increase in total government debt (money and bonds outstanding). That would be FISCAL policy, that’s basically a tax rebate. We mean an open market operation where the fed buys back a government bond for cash, thus the supply of cash has increaed but the supply of bonds has decreased. Before this happened everyone was maximizing their objective functions and so was indifferent between a marginal increase in real balances, bond holdings or consumption. But after the fed intervenes their is an excess supply of real balances in aggregate, however, if interest rates haven’t changed then there is an excess demand for bonds. To restore equilibrium EITHER the interest rate changes or the excess real balance must find it’s way to the bond market somehow. If the interest rate doesn’t change then the extra money must go to the bond market.
How am I so sure that there was an excess demand for bonds? Because before the open market operation everyone was at a point of maximum in their objectives. If nothing else changed but the composition of their holdings then they are no longer at the max and will seek to return their.
Changing the money supply only has real aggregate effects if it changes the real interest rate.
correction in the second paragraph: “But after the fed intervenes there is an excess supply…”
An afterthought, my last comment also explains why QE is about lowering interest rates and not about increasing the money supply. The fact that it does happen to increase the money supply is irrelevant.
Nick,
“But the Hume/Patinkin homogeneity thought experiment tells us this cannot be true in the long run.”
Case 1: The Fed eases by dropping short rates; inflation expectations increase; long rates increase. The yield curve steepens by directionally opposite changes at each end.
Case 2: The Fed eases by purchasing plywood; inflation expectations increase; long rates increase; AND short rates decline because the Fed’s expenditure increases the level of excess bank reserves (unless sterilized), which will have the same effect on the fed funds rate that an announced change has in Case 1. Again, the yield curve steepens by directionally opposite changes at each end.
The mapping I’m referring to is the short rate effect in Case 1 to the short rate effect in Case2.
I’m inferring not a particular mapping between interest rates and nominal variables, but a mapping between two similar effects on short rates in the two different types of easing in my example.
JHK: “it simply buys an asset (or lends money) and credits the seller/borrower’s bank’s clearing account at the Fed” … “it’s paying interest to the banks on these reserves to compensate them for the fact that all of that excess is basically useless to them”
Forgive my ignorance, but if the Fed buys an asset (e.g. a bond of some kind) from a bank and credits the bank’s account, thus creating ‘excess reserves’, why is that money unavailable to the bank? What’s the point of the asset sale then? What’s stopping the bank from using the cash they get for the asset?
JKH @ 3.39: Yes, if you assume that expectations respond in the same way in both cases. But what is anchoring those expectations? In Case 2, expectations of future plywood prices are anchored by the Fed’s planned path for plywood prices (and this anchors expectations of all other prices). In Case 1, since the path of the price level is indeterminate for any given path of nominal interest rates, there is nothing to pin down expectations of future prices (of plywood, or anything else).
Nick,
Yes. That makes sense. The plywood pricing signal is more specific than the more general and uncertain price path inferred in Case 1. I agree that expectations wouldn’t be identical. But they should be the same directionally.
Adam @ 3,17, 19, 22: If it were any good other than money, the medium of exchange, I might agree with you. But money is different.
The people who sell bonds to the Bank in an OMO willingly accept money in return. But this does not mean they plan to hold that extra money. They temporarily accept money because money is the medium of exchange. They plan to spend it on something else (and presumably not bonds, because otherwise they wouldn’t have sold the bonds in the first place). Individually, each one thinks “I will get rid of this extra money by buying X”. And individually, each one can do this. But in aggregate they can’t.
Patrick,
Those are all the right questions, deserving of a thoughtful answer.
Let me get back in a bit.
Of course they don’t necessarily plan to hold the money. Nonetheless, unless people don’t want to return to the optimal point, some money must get back to the bond market or bonds must go up in price to become less desirable. It doesn’t have to be the same dollars but if prices don’t change then optimal holdings don’t change and so somewhere some money gets back to the bond market.
I mean really, we start in equilibrium with no excess supply or demand of bonds. Now you take some bonds away but you don’t change the price of bonds, doesn’t that mean we now must have an excess supply of bonds?
JKH @ 4.03:
There is no guarantee that those expectations paths should be the same directionally.
Take a very simple macro model. The output gap depends on the real interest rate. The real interest rate equals the nominal rate minus the expected inflation rate. The rate of change of the inflation rate depends on the output gap.
In a model like that, very small changes in initial expected inflation could turn out to have very big consequences for long run inflation, for any given time path for the nominal interest rate. If people expect a little bit more inflation initially, the initial real interest rate is lower, so output increases, and inflation increases without limit, and people will rationally expect hyperinflation. And if people expect a little bit less inflation initially, the result is ever-increasing deflation.
By contrast, small changes in the expected price of plywood will only have small changes on everything else.
Yes, there does exist an equilibrium path for Case 2 which coincides with the equilibrium path for case 1. But Case 2 is an unstable path, and anything which slightly moves expectations away from that equilibrium (and currently they are well away from any stable equilibrium) will lead to an ever-increasing divergence.
This is hard stuff. My brain hurts. http://www.youtube.com/watch?v=IIlKiRPSNGA
I think I need to take a break, and come back to this topic later. There has to be another way to think about it. so it’s clearer, both to me and to everyone else.
Patrick,
Reserves are at once an optical illusion and a logical conundrum. The Fed creates them out of thin air, but the banks don’t really need them. What on earth do I mean by this?
I’ll give you an example of how this works, but first let’s look at a system where the reserve requirement is actually zero – Canada.
The following is really simplified but essentially true of Canada:
The objective of banks in Canada like everywhere else is to meet their reserve requirement – which happens to be zero. What does this mean? It means their objective is to avoid an overdraft position with the central bank. Otherwise, they will be charged a penalty rate and face the stigma of discount window borrowing, etc. etc.
So the Royal Bank is sitting with zero reserves and wants to make a new loan. How can it make a new loan without having “the money” to do so?
That’s easy. It credits the borrower’s deposit account. That doesn’t require reserves. What happens if the borrower then moves his money to Bank of Montreal? Well, RB’s Bank of Canada account will be overdrawn, and BMO’s will be in surplus. How does the RB make this up? Well, it can go to BMO and bid for an interbank deposit. And it can cover its position in other ways that will have the same effect in terms of getting net credit at the Bank of Canada. Similarly, all of the rest of the activity that’s going on in the Canadian banking system is pushing deficit and surplus bank positions around and forcing the banks to react in the market in order to attract or lend enough funds to square their positions. And none of this requires the existence of positive reserves for the system as a whole. It’s the differentials against zero and the distribution of those differentials that make the system go round.
Now look at the US system, where there is a positive reserve requirement. Say the US system requirement in total is R. It actually resembles the Canadian system, except that the positions that are being squared consist of {r +} and {r -}, where each r is an individual bank reserve requirement with the Fed, and each r + and r – is an actual reserve position, surplus or deficit to requirement. Conversely, the Canadian system is equivalent to the US system where R = r = 0.
The most relevant thing is the distribution of the + and – positions as differentials against the r requirements. It is these differentials, not the r requirements per se, that are relevant. And that’s why Canada can elect to have R = r = 0. And the fact that Canada can set R = r = 0 is practical evidence that banks don’t require positive reserves to lend.
Going to your example, when the Fed buys an asset from a bank, the bank swaps its asset for a reserve credit. And you’re right. That puts the bank in a better position to be able to lend, other things equal. But the point is it doesn’t need that reserve credit to be able to lend. In the normal course of operations, it can just as easily cover its lending outflow with a deposit inflow, or for that matter sell an asset to somebody other than the Fed. In your example, the Fed has created additional reserves for the system with that particular transaction, which is an extra detail. (It may drain reserves from the system to offset the excess.)
The first moral of the story is that system required reserves R can be anything, including R = 0. Banks don’t require positive reserves to lend. They just need to be able to cover the differential of their shortages against a positive reserve requirement (US), or an overdraft against a zero requirement (Canada).
The second moral of the story is that the central bank sets the excess reserve position for the system as a whole, and in doing so affects the ability of individual banks to meet their requirements, thereby putting pressure on the interbank interest rate according to supply and demand for any excess reserves that are available. This is how central banks control short rates.
Now, look at today’s extraordinary situation in the US. Excess reserves have ballooned. But the story above should illustrate that the banks don’t need them to lend. One can construct a theoretical scenario in which all banks lend the same amount of money on the same day, resulting in new deposits all over the place, so that new assets and liabilities have been created throughout the system, but there is absolutely no change in total excess reserves or the distribution of those excess reserves. Everything nets out because individual banks are basically netting out against each other in clearing their positions at the central bank. A milder version of that scenario is happening all the time in the banking system. And in that scenario, the $ 700 billion that the US banks have now really serves no purpose.
In summary, the effective relevance of reserves is that of an accounting mechanism for the tracking of the flow and distribution of funds and the creation and destruction of asset and liabilities as they clear around the banking system in the normal course. The actual required level of reserves is irrelevant; it is the differential of actual against required that indicates a bank’s effective position and requirement to take further asset-liability action.
That’s long winded, but your questions are not that easy to answer. There are some counterintuitive aspects to the dynamics of the reserve system.
You didn’t last very long, Nick. I’m disappointed.
🙂
Nick,
“But Case 2 is an unstable path, and anything which slightly moves expectations away from that equilibrium (and currently they are well away from any stable equilibrium) will lead to an ever-increasing divergence.”
I think I see where something analogous to plywood targeting is intuitively more stabilizing for expectations than free form fed funds targeting.
Thx for the discussion.
Yes! Greenspan famously pursued a gold-price driven policy ’87-’95. In particular, he influenced the Fed’s rate-setting on the basis of regulating the gold price. One popular press account of this: http://www.smartmoney.com/investing/stocks/good-as-gold-18980/
Adam P.
I guess your arguments are so good you can look after yourself.
But you missed a good come back to this
“If it were any good other than money, the medium of exchange, I might agree with you. But money is different.”
Of course money is NOT only the medium of exchange it is also a store of value. Maybe a BETTER opportunity is expected in the future (yes I’m not a believer in continuous equilibrium, and I think some of these hypothetical arguments are just silly).
In my view the most important theory in economics is the law of second best. We should spend more time finding out how things actually work and less time worrying about how they should work.
That is, I cannot understand why we are having a theoretical about this question, we need to dig out some evidence about empirically actually happens.
And Adam P. I realise that you have started off in that direction now.
Good rule of thumb – every theory is wrong because it starts by making wrong assumptions.
reason’s query about why we’re having this theoretical arguement brings me back to one of the questions I posed at the beginning of this thread which I don’t think Nick ever answered.
… “why it now no longer works in practice”. The zero bound. Where’s the evidence that if Uncle Ben could get a negative REAL interest rate then we wouldn’t get the desired results?
And btw, Mankiw has jumped on the bandwagon too… http://www.nytimes.com/2009/04/19/business/economy/19view.html?_r=1
A good night’s sleep, and I’m back in:
Adam P: ‘… “why it now no longer works in practice”. The zero bound. Where’s the evidence that if Uncle Ben could get a negative REAL interest rate then we wouldn’t get the desired results?’
I see hitting the zero bound as a consequence of the failure of interest rate control, not just as a a cause of the failure.
We agree that if the central bank could create (big enough) expected inflation, the zero bound would not be a problem, because we could have negative real rates with positive nominal. (And I would add that if the central bank could get people to expect that real AD would be higher in future, consumption and investment demand would increase too, and the financial sector would recover, and these would increase the real natural rate as well).
Think back to the analogy of interest rate control as balancing a pole upright in the palm of your hand. You have to move south if you want the top of the pole to move north. But you have hit a wall (zero nominal bound) to the south, and the pole is falling over to the south. So you change the instrument: turn the pole upside down, grab the other end, and start walking north.
By increasing the price of plywood (gold, whatever), and letting it be known you will keep on increasing it, you raise expected inflation (and expected future AD), and get real rates negative (if needed).
As an aside, there are an awful lot of interest rates, and they are the ones that count for AD, that are decidedly non-zero.
Well, I don’t agree that hitting the zero bound is a consequence of the failure of interest rate control. More importantly though, even if you want to insist that the quantity approach makes sense you still haven’t said how you can change AD without changing the real interest rate. As I said before, you can argue that the changing AD caused the changing interest rates but if you want to argue that changing AD won’t change the real rate (via Euler equation) then you really need to explain that and Nick is just ignoring it.
But then, if you admit that interest rates will move whenever you get your results then using an interest rate target becomes uncontroversial, it’s much more observable then other things like money supply, and if you move it in the desired direction, by whatever action, you’re going to get the right result.
Nick,
I like the pole analogy as a problem analogy, up until the point you construct the solution of turning it upside down. The problem analogy is intuitive but the solution analogy isn’t. I can’t translate it.
There’s a contradiction in the idea that interest rate control is a failure. Interest rates are the sum of real rates and expected inflation. If interest rate control is a failure, why are you proposing solutions that have the objective of a specific constitution for interest rates (i.e. positive expected inflation components to the point they dominate negative real rates?). At the end of the day, any monetary strategy is geared toward an interest rate outcome, and that amounts to effective interest rate control. Whether you announce a target fed funds rate or a target expected inflation rate, which is a component of nominal rates, it effectively amounts to interest rate control. Interest rate control is not a failure. All you have to do is expand the idea of control from nominal rates to the real and constituent components of nominal rates. Picking up unconventional easing at the zero bound is an extension of interest rate control to the expected inflation component, rather than a failure of interest rate control.
JKH:
“I like the pole analogy as a problem analogy, up until the point you construct the solution of turning it upside down. The problem analogy is intuitive but the solution analogy isn’t. I can’t translate it.”
Agreed. The analogy starts to break down at that point. Try this instead:
Pole upright in the palm of the Bank’s hand. Bottom of the pole is the nominal overnight rate, top of the pole is the rate of inflation. There is a string tied to the top of the pole. At the other end of the string is a helium balloon. (The helium is just enough to keep the string upright, not enough to keep the pole upright). The balloon is the price (or maybe rate of inflation?) of plywood (or gold, or whatever). (OK, there are lots of balloons, each representing the nominal price of some good).
The pole starts to fall over to the south. Let go of the bottom of the pole, grab hold of the balloon, and move it north.
Will post this while thinking about Adam P’s and the rest of JKH’s comment (which are the same, and is the crucial point).
On second thoughts, I think I’m going to start a whole new post, to address Adam P’s and JKH’s other point. It’s both too long, and too important, for a comment.
I like the balloon modification.
Targeting (grabbing onto) a single balloon runs the risk of popping it though, with resulting ineffectiveness in terms of inflating the critical mass of the rest of the portfolio of balloons.
Unconventional easing (quantitative, qualitative, credit, … however defined) is an attempt to inflate the full portfolio of balloons, albeit indirectly, avoiding the concentration risk inherent in a single balloon target.
Just one more thing from the original post:
“If you want higher nominal interest rates you first need to lower interest rates, so that inflation starts to rise, and expected inflation starts to rise, at which point you can raise interest rates, and raise them higher than originally, so that inflation and nominal interest rates eventually settle down at some new higher level.
That’s how it was supposed to work in practice.”
Really? That’s how it was supposed to work? I don’t think so. This quote doesn’t make any sense at all. (And as an aside, why would we ever just want higher nominal rates arbitrarily. The central bank is trying mangage rates to stabalize the price lever or inflation rate and maintain full employment.)
Furthermore Nick says: “The standard argument of why interest rate control works in practice is that prices are sticky. The economy can’t jump from one nominal time-path to another. It takes time for monetary policy, seen as a gap between the interest rate set by the central bank and the neutral or natural rate of interest, to affect inflation. Provided the central bank could adjust the rate of interest more quickly than prices can adjust, the central bank can keep the price level determinate.”
Except for the fact that price stickiness is important this is manifestly not how these models work. An arguement that’s closer to correct goes like:
Since price are sticky the central bank can change real interest rates (since changing the nominal rate is a change to the real rate if prices can’t move right away). Changing real rates allows the central bank to affect AD. Buy influencing AD the central bank influences inflation via a Phillips curve.
Now, does that story get you price level determinacy? Not neccesarily. It depends on your Phillips curve specification. And the canonical New-Keynsian Phillips curve doesn’t get you determinancy unless you’re willing to make “we take the only bounded solution” statement which I don’t like to do.
However, the Cochrane paper (http://faculty.chicagobooth.edu/john.cochrane/research/Papers/determination_taylor_rules.pdf) resolves the determinancy question to my satisfaction and my last paragraph then applies to explain how inflation can then be controlled by the central bank.
Adam P: that’s really how it did work in practice, during the 1970’s and 1980’s (only in reverse). At least, that’s the absolutely conventional story of how it did work.
We had high inflation, and high nominal interest rates. The Bank of Canada (and other central banks) raised interest rates (tightened monetary policy). Eventually inflation fell, and then expected inflation fell, and then the Bank of Canada lowered nominal interest rates.
That’s the received doctrine. I am saying nothing new there. There were a few monetary economists who thought that tighter monetary policy (announced in advance, and fully credible) would cause nominal interest rates and inflation to fall at the same time. They were wrong. Maybe they were wrong because it wasn’t a credible announcement (the “credibility hypothesis”), or maybe they were wrong because of inflation intertia, or maybe both. But that is certainly the time path of the variables. Ask anyone who renewed a mortgage in the early 1980’s to describe the episode. Both nominal and real interest rates went temporarily very high when central banks decided to get serious about reducing inflation.
When the central bank WANTED higher nominal rates it first lowered them in order to generate inflation?
Nick: “We had high inflation, and high nominal interest rates. The Bank of Canada (and other central banks) raised interest rates (tightened monetary policy). Eventually inflation fell, and then expected inflation fell, and then the Bank of Canada lowered nominal interest rates.”
This statement seems to me to be prefectly correct but have nothing to do with:
“If you WANT higher nominal interest rates you first need to lower interest rates, so that inflation starts to rise…”
Adam P. @ 12.16.
I thought I was saying the same thing (only with all the signs reversed, of course). If not, it must be because I wasn’t clear enough when I said it the first time.
Let me re-phrase it this way: “If the Bank wants higher inflation and higher nominal interest rates in several years’ time, it must reduce nominal interest rates this year, so that actual and expected inflation increases, and after they do, it can raise nominal interest rates to a level higher than they originally were”
@ 12.06 : I didn’t see that comment while writing my own (we were writing at the same time).
That link doesn’t work, but I think it’s the same John Cochrane paper you linked to on the previous page of comments.
On reflection, I really like that Cochrane paper, but I take away a very different conclusion (both from you, and from Cochrane).
What he shows is that the verbal story New Keynesians give for price level determinacy under interest rate control totally contradicts the formal model with a canonical Calvo-style Phillips curve. That’s important.
If I interpret you correctly, I think you are saying “keep the canonical Calvo NK Phillips curve, but make assumptions about the fiscal regime to get determinacy”. Is that right?
I would take the other horn of the dilemna. I don’t like the canonical Calvo NK formal Phillips Curve anyway, for other reasons. I believe in inflation inertia (which is harder to model formally, but can be done in principle). So I stick with the verbal story. But then, the determinacy of the price level becomes an empirical question, depending on how quickly the Bank can respond, relative to the speed at which inflation and expected inflation adjusts, before hitting the lower bound.
But of course, my maths is not up to it. I’m an old guy, never good at math, and burned out.
No,no. I hate Calvo pricing, we agree 100% on that.
However, I do think determinacy comes from the fiscal theory as set out in Cochrane’s “Money as Stock” paper.
My choice of phrasing was to keep seperate the part of the theory that I think is coherent and generally valid, ie. not to dependent on something ad hoc like Calvo pricing.
How you specify the Phillips curve is still open so it was put in a different paragraph. I’m sympathetic to the form of the NK Phillips curve in principle but like you I don’t buy it’s derivation. Furthermore, it’s fatally flawed. It fails the “Lucas Critique” test because the parameters of the Calvo pricing (average frequency of repricing for the firm) is treated as a physical constant that is exogenous and a constraint on the firm. That’s just dumb, clearly in an unstable environment with high and volatile inflation firms would start changing prices more frequently.
However, even though I hate the derivation I am somewhat sympathetic to the final form of the NK Phillips curve. I think you do need an expected future inflation type term in there and that’s what gives the in-determinacy when you try to get the Taylor rule alone to pin down the price level (without appealing to the fiscal regime).