The Non-Relationship Between Interest Rates and the Money Supply, Part 2
by Mike Kimel
The Non-Relationship Between Interest Rates and the Money Supply, Part 2
Cross-posted at the Presimetrics Blog.
This post is a bit less about Presidents than usual, but its a follow-up to last week’s post on the non-relationship between the money supply and interest rates. (That post appeared both at the Presimetrics and Angry Bear blogs. In that post, I noted that the Federal Reserve tends to move the money supply monthly and seasonally with no corresponding change in the fed funds rate. For example, the Fed will increase the money supply in December to make sure there’s adequate money in circulation for the Christmas shopping season, and yet interest rates don’t move at all.
This week, I want to expand on that, and point out that I wasn’t entirely accurate. There actually is a relationship between interest rates and the money supply, but its not the the one taught in textbooks. The textbook relationship, as I noted last week, can be described like this:
The Fed determines what the Federal Funds interest rate should be. If the Fed wants to reduce interest rates, it will create money out of thin air and use it to buy bonds. Because there is more money competing to buy bonds, the interest rate bonds have to pay falls. At the same time, the added money sloshing around becomes cheaper for anyone to borrow, whether they’re issuing bonds or not. On the flip side, if the Fed wants to raise interest rates, it sells bonds. That forces anyone else trying to sell bonds to raise interest rates to compete. Additionally, because the Fed retires the money it gets paid for the bonds it sells, that process reduces the amount of money available in the economy, making it harder to come by and hence more costly.
In other words, the money supply and interest rates are negatively correlated. An increase in money supply leads to a decrease in interest rates, and a decrease in the money supply leads to an increase in interest rates. There is an alternative story, but it only applies to high inflation environments. If you’re in Argentina in 1982, for example, the money supply is increasing so fast that any little bit of new money translates immediately into inflation and higher interest rates.
So there’s the theory, what everyone knows is true. But what really happens? For that, as always, we cut to the data. The Federal Reserve’s Economic Database (FRED) reports the federal funds rate going back to July of 1954. For the money supply, we used M1 from 1959 to the present, and the money stock for years before that.
The graph below shows the correlation between the 12 Month Percent Change in M1 and the Fed Funds rate in the same period, one month later, two months later, etc.:
The graph shows that the correlation between the percentage change in M1 over a year ending in a given month and the Fed Funds rate in the same month is about 14%. The correlation between the annual percent change in M1 and the Fed Funds rate in the next month rises to 16%, and so forth. The 1 year change in M1 has a higher correlation with the Fed Funds rate 36 months later than with any lag on the Fed Funds rate. In plain English, increases in M1 lead to increases in the Fed Funds rate, and decreases in M1 lead to decreases in the Fed Funds rate. That doesn’t sound at all the textbooks tell us we should expect in a world that isn’t suffering from hyperinflation.
Now, you may be thinking to yourself… maybe that relationship is a function of the fed trying to react to a slowing economy.
Strip out months in which the economy is in recession, plus the three months leading up to and the 3 months leading out of the recession, and the graph looks like this:
Notice… the correlation drops a wee bit, but the shape of the curve is pretty much the same. Once again, it is fairly evident that this does not conform in any way to the classic textbook story.
So what is going on? My guess is that in the U.S., for the period for which we have data, in general:
1. changing the money supply has had no direct relationship on the Fed Funds rate
2. changing the money supply has had a direct effect on the economy; increasing M1 (actually, real M1 per capita) causes the economy to grow more rapidly, and decreasing the real money supply causes the economy to grow more slowly or contract
3. because increases in the (real) money supply cause the economy to grow more rapidly, eventually an increase in the money supply will lead the Fed to raise interest rates in an attempt to slow the economy (to avoid inflation). On the other hand, because reductions in the (real) money supply lead to slower economic growth or economic contraction, these reductions will eventually lead the Fed to lower interest rates to try to get the economy to grow more quickly.
Step 2 is something we cover in the book, but I hope to write another post showing that soon.
Mike,
Congrats on getting the book out!! When are you coming by DFW for a signing?
As for the post….
Intuitively, the textbook case would work if demand for money stayed constant. We are pretty sure it doesn’t tho. But the whole idea is the Fed wants to influence demand for money by setting the cost of money. When the market doesn’t go along with the central banks command to “Make it so”, I believe Keynes calls it “liquidity trap” and at the other end of the spectrum it is called hyperinflation.
Almost forgot the other effects. Asset inflation and deflation (inevitably both) and ridiculous wall street bonuses. (no clawback there, buy gold with the personal account and lock it up in a swiss bank vault high up in the Alps. French wine is good too.)
Again this is meaningless. The Fed like all monopolists can set either the price or the quantity, and once one is set the other is set. They can’t set both. Since “money supply” is a constantly changing thing and that targeting it causes problems (see 1978-1982) the Fed targets the interest rate. Kash needs to come back…
OK, first, just a reminder that a number of criticism of the first post on money and rates were offered, mine among them. The intro offered here seems to ignore them. If you begin with a weak premise….
Next, I think I see a very, very large logical inconsistency. You have argued that the Fed controls the supply of base money, but not the funds rate. Then, you explain the unexpected sign on the lagged relationship between base money and the funds rate by saying “eventually an increase in the money supply will lead the Fed to raise interest rates in an attempt to slow the economy (to avoid inflation).”
So the Fed does have control over the funds rate? By what mechanism, please?
This effort has run off track. Let me offer a suggestion. Start out with the assumption that the many very bright people who’ve looked at monetary policy may have noticed things that you have not. I realize that this is an unpopular view. Around here, economists are universally assumed to be stupid, no matter who smart they may may be. Still, it is the right thing to do if one wants to understand monetary policy. When explanations for phenomena that you found contradictory or mysterious are offered, take a close look at them before moving on to the next thing. Leads, lags, announcement effects, seasonal changes in demand for money, cyclical changes in regulatory oversight and balance sheets, velocity and the like all are important to seeing how things work.
I realize that the insurgent joy of undermining received wisdom can be addictive. It served you well in the first effort. This effort seems to partake of that same urge to undermine. I would remind you of what Feynman said about conventional scientific belief – He said it’s mostly right. That’s why it’s conventional.
buff,
Thanks. I’ve been told by people that ordered the book that Amazon has shipped the book. I can’t say about any others.
Cedric,
Yes, but as I pointed out last week, both the NY Fed’s trading desk and I believe that, within certain limits (and I think we’ve mostly been within those limits for the past six or eight decades in the U.S.) you could set both the price (i.e., FF) and the amount of money that gets supplied.
Actually this non-correlation is no surprise to those familiar with the MMT view (see Bill Mitchell’s stuff at http://bilbo.economicoutlook.net/blog/ or Randy Wray’s writings at http://neweconomicperspectives.blogspot.com/ ). The conventional textbook view is that The Fed controls M1 absolutely by controlling reserves.
In contrast, MMT, being based on how banks actually make loans and how reserves are actually reported/obtained, realizes that banks are not reserve-constrained in creating new loans (M1). Instead, reserves get created after the fact when banks find credit-worthy borrowers and create new loans.
So the correlation between rising M1 and rising Fed Funds rates would reflect the real-world practice that as economic growth occurs banks make increasing volumes of loans – they create M1 growth. Then they need reserves to put against these new loans – hence they bid up Fed Funds rate searching for reserves for the just created loans.
Rob,
Again, last week I pointed out this isn’t true and the NY Fed’s trading desk (who would be the ones doing the setting) doesn’t think its true. And then you have a positive correlation. Are you really going to argue that as the quantity of money increases, its price rises, even in a non-hyperinflationary environment? Money as a giffen good? Seriously?
kharris,
Sorry if I forgot to address some issues. Lack of sleep has not been kind to me.
“ You have argued that the Fed controls the supply of base money, but not the funds rate.”
No. I have said the Fed controls M1 and the Fed Funds rate, and within a range, a range that depends on the Fed’s reputation, and a range that we’ve been in for most or all of the past six or more decades in the U.S., moving one will not move the other. I quoted from a NY Fed manual pointing out exactly that as well.
The Fed controls the fed funds rate mostly by stating what they want it to be in its announcements. “Guidance” at the regional meetings also helps.
“Start out with the assumption that the many very bright people who’ve looked at monetary policy may have noticed things that you have not. I realize that this is an unpopular view. “
I’m doing exactly that. The difference is, I’m siding with the folks who implement the policy, the NY Fed’s trading desk, over the academic literature. Actually, I recall reading a couple papers on “open mouth policy” in the past ten or twelve years so.
Finally, I’m going further back, to the basics of micro – the economic literature tells me that demand curves slope downward. Which doesn’t seem to be the case here. And money isn’t a giffen good.
Ok, missed last weeks post. I try and find it so we are on the same page.
But I’ll add we have gone way past simple Fed ops to manipulate FF rate. They took the balance sheet up 2.5X with QE trying to bend the entire yield curve of Treasuries and MBS. They also did sterilization moves to partly counter the trillion and a half of money they created as part of QE. Part of it really amounted to swapping bad credit for good credit on bank balance sheets. That was the “stealth” bank bailout that Congress didn’t have to approve.
Things got way more complicated in Fed Land.
Cedric,
Yup. The Fed is definitely in uncharted territory.
Mike,
Also, I recall that sometime back in the early 90s or late 80s the Fed decided that targeting M1 was ineffective and decided they should target inflation. Or they now describe their full mandate as price stability and maximum employment. I read a Fed Primer years ago on the St. Louis site that described a few different methods or market tools they use whether they think they need to tweak FF rate or money supply. If I recall, they favor Repos for money supply and short T-Bills which they use for FF tweaks. Or maybe that was the other way around, I don’t recall for sure.
econproph,
Thanks. I have to spend a bit of time (I don’t have it right now) looking at MMT (modern monetary theory). But off the top of my head, my recollection of MMT is that it contains a few underpinnings I don’t think are backed up by the data either. Forgive me if I’m wrong, but that it includes the debt not mattering. I do agree with the notion that there is an optimal (for growth) amount of money out there, though I would caution that the optimal changes literally by the day.
If anyone is wondering what difference a t bill op or repo op would make, t-bill ops are done in the open market and repos are done with fed member banks.
If anyone wants to reconcile Fed monetary practice with the much vaunted Bill Mitchell and Gang Economic Discovery, I think it works like this:
Someone walks into a bank and asks for a loan.
The bank says ok and the documents are signed and the money instantly appears credited to your account.
At end of day closing the bank realizes they are short Tier 1 capital.
They send a note to the marketing dept to run the free toaster ad.
To satisfy regulatory rules they borrow overnight money from a Fed member bank or any other bank.
If the fed member bank comes up short on capital next day, they do a repo with the fed.
If no new depositors go for the toaster deal, the banks sells some tier 2 assets.
Just guess’n.
Cedric Regula,
That rings a bell.
It’s a common misconception that MMT says debt doesn’t matter. What MMT says, is essentially that:
for a sovereign government issuing it’s own non-convertible, flexible exchange rate currency borrowing in it’s own currency, then debt is different. For such entities (US, Australia, UK, Japan, Canada, etc) financial default is not possible unless the legislature/executive voluntarily choose to default. The track record on such entities under such conditions supports the MMT’ers. The one recorded default of such an entity was Japan, 1942 when for obvious military-strategic reasons Japan chose not to repay to UK/US banks.
For non-currency issuing govt (California, your school board, Catalonia) or for a govt that borrows in a different, foreign currency (Latin America, Russia 1998), or for a govt with a fixed, convertible exchange rate (Greece & the Eurozone), then debt most definitely matters and can lead to default.
For sovereign, non-convertible, currency issuing governments, they can always repay debt. They (with their central banks) after all, are the source of the money base – they can always just credit the accounts of the banks with reserves when the banks want to redeem the debt.
What MMT brings to mind is 2 key concepts: we need to keep seperate the purely financial from the real. The other is that government debt is NOT like ordinary private debt because governments are NOT like households or firms or banks. Governments are the source of our money. I find it more useful to think of government debt as simply government currency that pays interest. The government issues three kinds of money: currency (which doesn’t pay interest), bank reserves (which now do pay interest to the banks in our attempts to re-capitalize them), and bonds (which also pay interest).
It’s a common misconception that MMT says debt doesn’t matter. What MMT says, is essentially that:
for a sovereign government issuing it’s own non-convertible, flexible exchange rate currency borrowing in it’s own currency, then debt is different. For such entities (US, Australia, UK, Japan, Canada, etc) financial default is not possible unless the legislature/executive voluntarily choose to default. The track record on such entities under such conditions supports the MMT’ers. The one recorded default of such an entity was Japan, 1942 when for obvious military-strategic reasons Japan chose not to repay to UK/US banks.
For non-currency issuing govt (California, your school board, Catalonia) or for a govt that borrows in a different, foreign currency (Latin America, Russia 1998), or for a govt with a fixed, convertible exchange rate (Greece & the Eurozone), then debt most definitely matters and can lead to default.
For sovereign, non-convertible, currency issuing governments, they can always repay debt. They (with their central banks) after all, are the source of the money base – they can always just credit the accounts of the banks with reserves when the banks want to redeem the debt.
What MMT brings to mind is 2 key concepts: we need to keep seperate the purely financial from the real. The other is that government debt is NOT like ordinary private debt because governments are NOT like households or firms or banks. Governments are the source of our money. I find it more useful to think of government debt as simply government currency that pays interest. The government issues three kinds of money: currency (which doesn’t pay interest), bank reserves (which now do pay interest to the banks in our attempts to re-capitalize them), and bonds (which also pay interest).
It’s a common misconception that MMT says debt doesn’t matter. What MMT says, is essentially that:
for a sovereign government issuing it’s own non-convertible, flexible exchange rate currency borrowing in it’s own currency, then debt is different. For such entities (US, Australia, UK, Japan, Canada, etc) financial default is not possible unless the legislature/executive voluntarily choose to default. The track record on such entities under such conditions supports the MMT’ers. The one recorded default of such an entity was Japan, 1942 when for obvious military-strategic reasons Japan chose not to repay to UK/US banks.
For non-currency issuing govt (California, your school board, Catalonia) or for a govt that borrows in a different, foreign currency (Latin America, Russia 1998), or for a govt with a fixed, convertible exchange rate (Greece & the Eurozone), then debt most definitely matters and can lead to default.
For sovereign, non-convertible, currency issuing governments, they can always repay debt. They (with their central banks) after all, are the source of the money base – they can always just credit the accounts of the banks with reserves when the banks want to redeem the debt.
What MMT brings to mind is 2 key concepts: we need to keep seperate the purely financial from the real. The other is that government debt is NOT like ordinary private debt because governments are NOT like households or firms or banks. Governments are the source of our money. I find it more useful to think of government debt as simply government currency that pays interest. The government issues three kinds of money: currency (which doesn’t pay interest), bank reserves (which now do pay interest to the banks in our attempts to re-capitalize them), and bonds (which also pay interest).
econproph,
Thanks for clarifying.
Inflation yes? But I would be more worried about things to come after that. Check this guy out, he called the crash back in 2008 and has been spot on with this for the last few years.
http://www.youtube.com/watch?v=gn6kS4l2yFM
In simple two dimensional graphical economic analysis you generally have two curves, one sloping down and the other sloping up. The point where they meet is the equilibrium point. In using this analysis you can move the curves and then read off the X and Y points that correspond to the equilibrium values.
In isolation looking at one curve only you really can’t say anything definitive. For instance, if you increase the money supply and there is no change in money outside forces pushing money demand the equilibrium point will move out and to the right along the money demand curve and interest rates will fall. But then another situation could occur where the money supply curve moves outwards and the money supply curve moves out with it keeping interest rates the same. Or, it could move a little more than the demand curve and interest rates would fall, or move a little less and interest rates would rise.
The point is that you can’t tell anything about what’s actually going on by looking at only one curve but not the other. In a response to the blind use of data in analysis Milton Friedman developed a narrative approach to analyzing monetary policy over history. Christine Romer along with her Husband provided an example of this approach in June’s American Economic Review with respect to the impact on the economy from increasing and decreasing taxes. The Narrative approach uses primary sources to identify every significant piece of federal economic monetary or tax legislation over a period of time. It uses these sources to determine the motivation of each action, and the size and timing of the effects either on interest rates or its revenue effects.
You can’t just look at data in isolation and understand the policy or what was the motivation behind what is observed in the data.
http://www.ssc.wisc.edu/~ldrozd/my_files/102files/slides_9.pdf
“The Fed controls the fed funds rate mostly by stating what they want it to be in its announcements”
Sorry, no. Already did this one. The Fed controls the funds rate through provision of liquidity. Announcements of the target rate are in line with open market operations, which is why they work. You have cited one statement by the desk, aimed at explaining operational detail. You have claimed to side with the desk in opposition to the academic literature, but the desk is not in oppostion to the literature. By ignoring a variety of factors that mediate between open market operations and the funds rate, you are able to make up a tension between the desk and the literature which doesn’t exist.
Once you have addressed seasonality, lags, changes in velocity and the like, you may find an unexplained divergence between what we assume to be the relationship between supply and demand and what can be seen in the fed funds market. So far, you are asserting a divergence, but have not addressed seasonality, lags and the like.
It is as if you noticed a price divergence between cattle auctions at two different locations and claimed a distortion, without taking transportation costs into account. Once you get a handle on the factors at work, the slope of the demand curve won’t seem wrong.
I don’t know why they persist with this drivel. Of course everyone knows a government can fill up a hard drive with worthless currency electrons, hand it to us for payment of debt, and say “we’ll call it even now,right?”
Thx for the news.
No they don’t do the setting. The FOMC does the setting, they carry out what needs to be done to make the target. The quantity of money rises to maintain a fixed price of the interest rate. You’re the one who seems not to understand the basic framework of the money market.
title: “The Non-Relationship Between Interest Rates and the Money Supply, Part 2”
First, you need to specify which “money supply” you are talking about.
I suggest this .pdf.
http://www.bis.org/publ/work292.pdf?noframes=1
From p.9 to p.11 of the .pdf
“At its most basic level, the implementation of monetary policy has two core elements (Borio (1997), Borio and Nelson (2008), Disyatat (2008)). The first comprises mechanisms to signal the desired policy stance (“signalling”); the second comprises operations that involve the use of the central bank balance sheet to make that policy stance effective. Because these operations typically involve managing the amount of central bank funds in the system, they are generally known as “liquidity management operations”.
Before the recent crisis, monetary policy implementation across countries had generally converged on an approach in which the policy stance was defined exclusively in terms of a short term interest rate – henceforth referred to as “interest rate policy” (eg Markets Committee (2008)). In this approach, the corresponding policy signal generally takes the form of the announcement of a “policy rate”, which defines the desired level of the interest rate. In turn, liquidity management operations are designed exclusively to help make that interest rate effective: they ensure that a market “reference rate”, typically an overnight rate, tracks the desired interest rate level closely. As such, liquidity management operations play a purely technical and supportive role. They neither impinge upon, nor contain any information relevant to, the overall stance of policy.3
The fulcrum of the implementation of interest rate policy is the market for bank reserves. This is a peculiar market. By virtue of its monopoly over this asset, the central bank can set the quantity and the terms on which it is supplied at the margin. As such, the central bank is able to set the opportunity cost (“price”) of reserves, the overnight rate, to any particular level, simply because it could stand ready, if it so wished, to buy and sell unlimited amounts at the chosen price. This is the source of the credibility of the signal.
Crucially, the interest rate can be set quite independently of the amount of bank reserves in the system. The same amount of bank reserves can coexist with very different levels of interest rates; conversely, the same interest rate can coexist with different amounts of reserves. What is critical is how reserves are remunerated relative to the policy rate. We refer to this as the “decoupling principle”.4 It is a principle that has far-reaching implications for the rest of the analysis.
There are two types of remuneration scheme. Typically, central banks remunerate excess reserve holdings – ie. holdings over and above any minimum requirements – at a rate that is below the policy rate (scheme 1 in Figure 1). As a result, banks seek to economise on […]
I also suggest JKH’s comment at November 29, 2009 at 05:32 PM from this post:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/11/money-banks-loans-reserves-capital-and-loan-officers.html
Warning, it is a long comment.
Mike Kimel said: “No. I have said the Fed controls M1 and the Fed Funds rate, …”
I don’t believe that the fed directly controls M1.
In most cases when the fed lowers the fed funds rate, they hope banks have enough capital and other rates come down with the fed funds rate. They also hope there are enough creditworthy borrowers who are willing to go into debt when the right interest rate falls. The person buying something gets the demand deposit and the bank gets an asset (the loan with the interest rate attached).
kharris,
I didn’t want to make this post too long, but actually I looked at a number of things you mentioned and got the same results. For instance, if you look at data from Januaries only, or Februaries only, or any month, for that matter, you still get the same upward sloping “demand curve” particularly if you lag the ff. (36 months seems to produce the highest correlation – between 40% and 53%, depending on the month.) I would also note that if you break the sample into pre 1980 and post 1980, the correlations become stronger, whether you look at individual months alone or the entire sample.
All that is to say… there is a positive correlation between the change in the money supply and the fed funds rate, almost no matter how you look at it. The only way that would be feasible given the textbook explanation is if you assume the Fed is a lot more competent than it appears to be at managing the economy.
Templeton,
This is not a situation of two curves. This is one curve, at least if you believe the textbook explanation. You have a good (money) and you have its price. The demand curve should not slope upward.
No More Fed,
I disagree with your handle. But be that as it may, in the post I’m very clear that I’m talking about nominal M1.
No More Fed,
Again, I disagree with your handle. But be that as it may, the doc you cite is missing one key point. Yes, “ in setting the interest rate, no open market operations need be involved at all.” But it also has a strong influence (perhaps to some extent, at the margin, one could call it controlling) over the size of required bank reserves. This comes about through the monthly regional meetings and other sources of communicating their views to the local banks. Let’s be realistic – when the Fed (with the tacit acquiescence of the FDIC) announces that “it would be prudent for agricultural loans in the region to be reduced” one thing will happen with certainty: agricultural loans in the region will be reduced. When the Fed can, ahem, “influence” the amount of loans private banks are going to make, it also “influences” their reserve holdings and thus the money supply.
Which means… the Fed does a pretty good job of setting both P & Q for money.
No More Fed,
See my previous comment about the Fed’s statements at the monthly regional meetings. I’ll admit, I’ve never attended any of ’em. But a buddy of mine used to attend them regularly, and as he saw it, the Fed was pretty nakedly telling private banks to increase or decrease their loans.
No More Fed,
That will take a bit of time to digest. But its missing the bit about how the Fed actually exerts at least some control over both sides (quantity as well as price) of the equation.
i note one other thing: if the Fed set one side of the equation (i.e., either p or q) and the banks went out and set the other side, it should be possible to deduce total revenues for the entire banking system (barring a small (in most months) leakage for bad loans) months in advance simply by knowing the FF rate. And yet there doesn’t seem to be all that much consistency in the revenues or profits of the banking sector.
Rob,
I wrote incorrectly – the NY Fed, by the classic interpretation, makes it happen, and yes, the FOMC does the setting. But the NY Fed is saying that their actions aren’t what makes the FOMC’s desires happen.
“If the Fed wants to reduce interest rates, it will create money out of thin air and use it to buy bonds.”
And, “In other words, the money supply and interest rates are negatively correlated. An increase in money supply leads to a decrease in interest rates, and a decrease in the money supply leads to an increase in interest rates.”
I may be wrong, but to me that implies “money out of thin air” and “money supply” are the same. Do you mean M1 money supply in both cases?
If I understand that correctly, required bank reserves don’t have anything to do with monthly regional meetings and other sources of communicating their views to the local banks.
Required bank reserves are usually about the overnight and supposed to be risk free interest rate.
Let’s say the fed funds rate is 3%. Banks think that lending to farmers for agricultural loans at 8% is a good deal for them and the farmers agree because food prices have moved up. The fed may suggest that maybe it isn’t a good idea because food prices may not stay there. If they fall, the farmers will probably not have enough income for paying back the interest payments and principal. If food prices fall, the real estate (collateral) will probably fall too. Now, the banks have loses.
Just because the fed funds rate is known, does not mean you know the number of creditworthy borrowers who want to borrow.
I’m thinking that a good bit of this discussion is really about the difference between the central bank reserve money supply and the M1 money supply.
If true, that would put a lot of the responsibility for the real estate boom and bust right on the Fed’s doorstep. Most of what I’ve seen in the news is both the Fed distancing themselves from the bust and playing dumb about not knowing what bubbles are, and also “smarter than us economists” saying Fed policy couldn’t have anything to do with the housing debacle or anything else going bad with the economy.
In other words, the Fed fixes things with the printing press. Breaking anything is not possible.