Thursday, August 18, 2016

QE and the amazing demand for money

Since late October 2008, I have been arguing that the main objective of the Fed's Quantitative Easing program was not to provide monetary stimulus, but rather to satisfy very strong money demand. I've fleshed out the argument many times over the years, with a few examples herehere, and here.

A short version of the argument goes like this: The true objective of Quantitative Easing was not to "print money." Instead, it was to convert notes and bonds into bank reserves (now effectively T-bill equivalents, since they pay interest) in order to meet the economy's tremendous demand for liquid, default-free, interesting-bearing securities. QE dramatically expanded the supply of "money" at a time when the market's demand for money was almost insatiable. Milton Friedman taught us that inflation results when the supply of money exceeds the demand for it. The evidence to date speaks for itself: we have seen no increase in inflation despite an enormous increase in the monetary base, so it must be the case that QE served mainly to supply money that the market wanted to hold.

In any event, QE could never be stimulative, because you can't expand the output of an economy by expanding its supply of money; pumping money into an economy in excess of what's demanded simply serves to inflate the price level. So it makes little or no sense to argue that the Fed (or any other central bank that has engaged in QE) is powerless to stimulate the economy. QE wasn't a failure, it was absolutely necessary, and it worked: if the Fed hadn't undertaken QE there would have been a serious shortage of money, and that could have prolonged the recession and created lots of deflation.

QE was a response to something that triggered a huge demand for money, and that "something" was a major shock to confidence. Terrified of the prospect of another financial market meltdown and another global recession, people—and especially banks—wanted a safe place to hide, and there were not enough T-bills to go around at the time QE launched. (Recall that by June 2008 the Fed had sold virtually all of its holdings of T-bills in response to the market's demand for them.) Moreover, in the years following the Great Recession, Treasury significantly reduced its issuance of T-bills in favor of longer-term securities in a rational response to the historically low level of Treasury yields.

Even though confidence has gradually returned as the economy has recovered, it's quite likely that banks' appetite for excess bank reserves will remain very strong because of increased regulatory requirements, as I noted earlier this year. At least half of the $2.2 trillion of excess reserves currently held by major banks likely is needed just to meet risk-weighted capital requirements imposed by Dodd-Frank and the Basel Accords, plus the Fed-imposed requirement that banks hold highly liquid assets (mainly bank reserves) equal to 100% of the amount that Fed stress tests indicate they would need to survive another liquidity crisis. As for the rest, banks are still—obviously—reluctant to lend anywhere near as much as their abundant holdings of excess reserves would allow.

Here are some up-to-date charts that illustrate and explain the increased demand for money:



The chart above is my preferred way to measure the demand for money. It compares the level of the M2 money supply to the level of nominal GDP. The ratio of M2 to nominal GDP is akin to the percentage of annual income that the average person holds in the form of "money." For many years, from 1960 through the late 1980s, this ratio was relatively stable. It declined in the 1990s due to the fallout from the S&L crisis. Since 2001, however, it has risen steadily and is now at its highest level ever.



The chart above shows the current composition of M2: retail money market funds, small time deposits, currency in circulation, retail checking accounts, and bank savings deposits. The mix of these components has changed over time, but the sum is a good representation of the amount of highly liquid, easily spendable "money" that is held by the public.



The chart above compares the level of M2 to the level of nominal GDP, with both measured using a semi-log scale. While the ratio of M2 to GDP has wobbled quite a bit over the years, over a period of almost 60 years the two have increased by almost the same amount. (Note that the two y-axes are offset by a factor of 1.75.) This makes intuitive sense, since the larger the economy, the higher are incomes, and the more money is needed to make everything work. The gap at the right hand side of the chart suggests that there is about $2.3 trillion of "extra" M2 money in the system, most of which is held in the form of bank savings deposits, which now represent about two-thirds of M2, as compared to about 50% in the early 2000s.

In other words, people are happily "hoarding" some $2.3 trillion of extra money relative to their incomes, most of that in bank savings deposits which pay almost nothing in the way of interest. (That the public is happy to hold some $8.6 trillion in bank savings deposits despite the fact that they pay almost nothing is a powerful illustration of just how strong the demand for money is.) Should they lose the desire to hold that money, and decide instead to spend it, nominal GDP could grow by as much as $4 trillion (2.3 * 1.75) if past relationships were to reestablish themselves. Much of that growth would likely come in the form of higher prices, while some would come from stronger real growth—because declining money demand, the flip side of rising confidence, would likely go hand in hand with increased investment, which has been sorely lacking during this recovery.

The following charts demonstrate that despite all the trillions of bonds the Fed has purchased, the amount of money in the economy has not grown any faster under a QE regime than it has over the past six decades:


We start with the expansion of the Fed's balance sheet, a simplified version of which is shown in the chart above. The blue bars represent positions at the beginning of 2007, well before the onset of the 2008 financial crisis and the launch of Quantitative Easing. The red bars show positions as of the Fed's latest report. Thus, the era of Quantitative Easing, which began in October 2008, saw the Fed on net purchase almost $3 trillion of Treasuries and MBS. This resulted in an increase of about $2.4 billion in bank reserves and an increase of about $600 billion in the amount of currency in circulation. (Banks can exchange their bank reserves for currency whenever they want.)


The chart above shows the growth of bank reserves, which grew from almost $100 billion in September 2008 to just over $2.4 trillion today.


Because of the expansion of the money supply (which increases required reserves) and the modest shrinkage in the Fed's balance sheet over the past few years, excess bank reserves have declined by almost 20% since their high two years ago. And the sky hasn't fallen.


The monetary base, which is the sum of bank reserves and currency in circulation, and which in effect represents the high-powered money that the Fed has created via its purchases of notes and bonds, has expanded from about $900 billion in September 2008 to just over $3.8 trillion today.

The Fed massively expanded its balance sheet and massively grew the supply of reserves and high-powered money. But contrary to popular belief, there was NOT a similar explosion of growth in the money supply. What happened was that banks were happy to lend their deposit inflows to the Fed, rather than to the private sector. As I explained here, banks took in $4 trillion of savings deposits and used that cash to purchase notes and bonds which were in turn sold to the Fed in exchange for bank reserves. That is why Fed purchases did not create massive amounts of new money: banks were happy to boost their holdings of bank reserves.



The chart above is the proof. For almost six decades, the M2 money supply has grown at an annualized rate of 6.9%. During the period of Quantitative Easing, M2 grew at about the same rate as it has grown since 1960.


Looking more closely, in the above chart, we can see that M2 has grown by an annualized 6.5% since early 2007 and since QE began in late 2008. A slower rate, in fact, than it has averaged since 1960.

The Fed's Quantitative Easing regime is broadly misunderstood. It was not about stimulus, it was instead about satisfying an amazing and unprecedented increase in public's and the banking system's demand for safe, interest-bearing assets.

It may all end in tears, if the demand for money should decline and should the Fed fail to take timely and sufficient measures to offset the decline in banks' demand for excess reserves (by shrinking its balance sheet and/or raising the interest rate it pays on excess reserves), but for now things look OK.

46 comments:

NormanB said...

Give this man a Nobel Prize in Economics!!!!!!!

Benjamin Cole said...

Terrific post. I do disagree about some parts, but so be it.

If you think the QE program in the U.S. is flabbergasting, then consider Japan.

The Bank of Japan now owns one-third of Japan's national debt outstanding ( and they have a doozy of a debt level). The BoJ is also on course to become the largest shareholder in Japan in most large companies.

Japan is still fighting deflation! And the yen is appreciating!

My own unpopular conclusion is that a straightforward program of helicopter drops or money financed fiscal programs is perhaps the best approach today. The Bank of Japan is literally running out of assets to buy.

If Scott Grannis is correct, that without QE there is the possibility of a long-term deflationary recession, then helicopter drops make sense in Japan.

Interestingly enough, Japan largely avoided the Great Depression through helicopter drops.

These are some of the most interesting economic times imaginable. Consider someone like James Grant of Grant's Observer. Here is a guy incredibly intelligent and erudite. And yet for three decades he has been predicting an inflationary holocaust. In recent years he has stated he still believes a hyperinflationary catastrophe is pending, but he will no longer say it is imminent.

My conclusion is that the trade of macroeconomics has become ossified and practitioners genuflect to hoary, encrusted totems of the past.

Something has changed out there.

Unknown said...

Scott, for any middle class or upper middle class family over the past 8 years +, to imply (or others imply in the threads)that we have not experienced any "real" inflation across the board (groceries, con-sum discretionary, property taxes, sales taxes) is simply ludicrous from my vantage point (and I actually have savings)!

Scott, what are your thoughts, if any, on Paul Craig Roberts? My God, his pitch and books are so seemingly compelling...

We really cannot elect Clinton to save our lives as she will enact all her constituent desires - to bankrupt America at the middle class' expense - and possibly get us into nuclear war with Russia. The Billionaire elites IMHO want to see a reboot of the US (to our middle class' expense/sacrifice). As much as I have thought that this thought process was too conspiracy theory in years past - it seems to be unfolding in front of our eyes daily without challenge.

If I was not 22 year veteran of the tech industry that now cannot get decent in-face interviews much less a job offer back in this sector, I would not necessarily believe the dire straits we seem to be in and normalcy bias I have experienced. Your thoughts much appreciated...

steve said...

My brain hurts...

Anonymous said...

Another excellent post Scott. Thanks.

Rob said...

Scott, this article says the Fed wants to raise rates but can't because it knows the housing market, stock market etc will collapse, since Main Street is already over-extended and QE has made assets more expensive than they should be. I'd love to know what you think of the author's arguments !

http://news.stocktradersdaily.com/news_releases/FOMC+The+Risk+of+Higher+Interest+Rates+ProShares+UltraShort+Lehman+20+Yr+TBT+81616.html#

Scott Grannis said...

Rob: The author of the article believes the Fed has stimulated the economy by pushing rates artificially lower. I don't believe that, as I've tried to explain here and in many other posts. I think rates are low because the market is still very cautious. I don't think the market is over-extended. Real housing prices today are 22% below where they were at their peak. In real terms, housing prices are back to where they were in late 2003, when 30-yr mortgage rates were 5.5%. Today mortgage rates are less than 3.5%, but that hasn't been enough to get prices back up to where they were when rates were significantly higher. Households are much less leveraged today than they were in 2003/2004. Check out this post with info on household balance sheets, net worth, and leverage:

http://scottgrannis.blogspot.com/2016/06/the-us-is-richer-than-ever.html

I think what will drive a rate hike decision is evidence that confidence is increasing and the demand for money is declining. We're seeing some of this, but not a significant amount, so I think the Fed is on hold for now. More confidence and less money demand will go hand in hand with more investment, and that will strengthen the economy. A stronger economy can easily sustain higher interest rates.

Unknown said...

Scott--I like that you repeat these charts over the years. Great context. Question--the M2 vs GDP chart--tight fit up until the 1990s, the M2 dips below for a long period then post-crisis M2 is now trending higher than GDP. Signal? If it were to continue, what does it signal?

Tal

Scott Grannis said...

Re Paul Craig Roberts: I was a fan of his back in the 1980s. But then I decided he had gone off the deep end into conspiracy theories and fringe thinking. I stopped following him a long time ago, so I can't say whether he's gotten better or not.

As for inflation, of course there has been inflation, but there has also been significant deflation in durable goods prices for the past 20 years. A flat screen TV that originally cost $8000 now costs $500. Computers have been getting cheaper and much more powerful. Smartphones are universally owned and extremely cheap for all the things they can do.

So there is a lot of good news, but so is there lots of bad news. We could be doing a whole lot better if economic policies were more growth-oriented and business-friendly. And if the government were smaller and less intrusive.

Scott Grannis said...

Tal, good to hear from you! The rise in M2/GDP is an unambiguous signal, I believe, that risk aversion is still very much alive and well in the world. People want to hold more money because they are worried about a host of issues, many of them political in nature. Rising M2/GDP has coincided with a long period of weak business investment, another sign of risk aversion. Households have been deleveraging even as interest rates fall, because they are risk averse.

Unknown said...

Bee honey mat ong rung nguyen chat | mat ong rung | gia mat ong rung | mat ong nguyen chat | mat ong rung thien nhien | mat ong | mat ong nguyen chat lam dep | ban mat ong nguyen chat | lam dep tu mat ong nguyen chat | ban mat ong rung nguyen chat tphcm | mat ong rung nguyen chat gia bao nhieu quality

randy said...

Very thought provoking. Have you noted what assets the rise in cash (savings deposits) displaced? And/or do you relate the rise in cash to reduced consumption? It may not be irrational, or likely to change anytime soon, for John Doe to have subdued consumption habits relative to the 80s-90s. It is also of course not irrational to hold cash instead of a mid/long bond fund earning nothing. Trying to consider what might result in reduced demand for cash.

Unknown said...

The demand for money appears to be inversely correlated to the velocity of money. So is demand for money going up because money is getting less traction so to speak? https://fred.stlouisfed.org/series/M2V

Unknown said...

The demand for money appears to be inversely correlated to the velocity of money. So is demand for money going up because money is getting less traction so to speak? https://fred.stlouisfed.org/series/M2V

Benjamin Cole said...

Scott: if investors actively wish to be in cash, then what is the purpose of the Fed's interest on excess reserves?

In Japan, of course, there is now negative interest on many deposits.

Is IOER an example of "regulatory capture"?

A filip for commercial banks?.

Anonyx said...

Scott, why do you suppose the Fed would not more directly communicate that QE is satisfying a demand for liquidity and that it is not intended to be stimulative?

Tom Nugent said...

Let's take another perspective. The Fed bailed out the economy by buying distressed mortgage securities. So they effectively increased a tax on the private sector by taking interest bearing securities in and giving out non interest bearing reserves. So now the Fed is paying about $80 billion in interest received on those securities back to the Treasury. So, the Fed's policy has been an effective tax on the private economy.

Benjamin Cole said...

Tom Nugent:

You raise an interesting point.

But I think it is a mischaracterization to say the Fed only swapped reserves for Treasuries or MBS.

Remember, the Fed only buys or sells securities through the 22 primary dealers. The Fed does NOT buy or sell securities directly from commercial banks.

The commercial banks carry only a relatively small amount of Treasuries in inventory. Remember, commercial banks are largely intermediaries.

The 22 primary dealers (using freshly digitized funds borrowed from the Fed, through the Primary Dealers Credit Facility) bought Treasuries or MBS through market operations, and paid fresh digitized cash for the securities. The institutions or people who sold Treasuries or MBS received fresh digitized cash for their sales.

So, the final picture is this: $4 trillion in fresh digitized cash in the pockets of bond and MBS sellers, another $4 trillion in reserves, and $4 trillion on the Fed's balance sheet.

(Where do the reserves come from? When the primary dealers sell the bonds and MBS to the Fed, their commercial bank accounts are credited).

The strange part of this is the we do not know what bond and MBS sellers did with their freshly digitized cash. They can deposit the money in banks, spend it, invest in securities or property, or convert the digitized cash into paper cash. It looks like they did all of that.



randy said...

Anonx - Bernanke made a stab at being explicit - sort of - here:

https://www.brookings.edu/blog/ben-bernanke/2015/03/30/why-are-interest-rates-so-low/

Scott Grannis said...

Benjamin: You still do not understand how the Fed and the banking system work. The Fed does not create money or "digitized cash." The Fed pays for anything it buys by crediting the reserve account of one of the major banks it deals with. Only those banks can sell bonds to the Fed, but those banks can source the bonds from all other banks, acting as a sort of agent. The banking system received strong inflows of cash seeking bank savings deposits. The banks used that cash to buy bonds, which were then sold to the Fed in exchange for reserves. It's that simple.

Scott Grannis said...

Annex: There are several answers to your question. The Fed billed its QE program as "stimulus" because:

1) It fit the narrative. "Everyone knows" that when the Fed buys bonds it pushes interest rates down and that stimulates the economy. What everyone does't understand, however, is that this is not an accurate description of how monetary policy works.
2) It was easy to explain. Just look at how difficult it is for people to understand and accept my explanation.
3) It was consistent with the Fed's mandate, which is to attempt to fine-tune the economy while also keeping inflation low. The economy has been weak and the Fed feels pressured to make it better.
4) To do otherwise would have raised a lot of issues. Who was in need of more T-bill equivalents? Why? Who was getting bailed out? Why was the banking system so desperate for more T-bill equivalents? (hint: the politicians imposed a lot of ill-considered burdens on the banking industry)

Scott Grannis said...

Tom: You're overlooking the fact that QE also ushered in the era of IOER. The Fed pays interest on the reserves it issues. Before QE it didn't. There is a huge difference between the two regimes. What the Fed pays to Treasury is the net of the interest it pays and the interest it receives.

Let's stipulate the Fed is taking money from the private sector and paying $80 billion to Treasury every year. Is that a net drain on the economy? Tough to say, I'd argue, since we don't know what would have happened if the Fed had never done QE. Maybe they saved the global financial system from collapse, and that would have been a gigantic burden.

Scott Grannis said...

One of the unfortunate things to come from QE is the uncertainty it creates. It's also unfortunate that the government has imposed huge new regulatory burdens on banks, which in turn has created a tremendous demand for bank reserves.

Scott Grannis said...

Re: what is the purpose of the Fed's interest on excess reserves?

Before QE, the banking system was forced to hold reserves at the Fed to collateralize deposits, but those reserves paid no interest, so they were a deadweight loss on banks' balance sheets. Now that reserves pay interest they are an asset. Before QE the Fed regulated the amount of money banks could lend by restricting the supply of bank reserves. Now banks have all the reserves they could possibly want. If the interest the Fed pays on reserves is not enough to make the banks willing to hold the reserves (so holding reserves looks better on a risk-adjusted basis than make loans to the private sector), then banks might be encouraged to lend way too much and this would result in a huge increase in inflation.

Having greatly expanded the supply of reserves, the Fed needs to pay interest on those reserves in order to prevent a massive expansion of the money supply.

Scott Grannis said...

erik g: the demand for money (as defined here as M2/GDP) is precisely the inverse of the velocity of money (i.e., GDP/M2). Money demand is rising (and velocity is falling) because the world wants to hold more money relative to income. The world is "stockpiling" money.

Scott Grannis said...

Re "what might result in reduced demand for cash?" Answer: anything that increases people's confidence and/or anything that increases the after-tax reward to taking risk.

Unknown said...

Thank you for answering my question. Didn't know it was the exact inverse equation!

Benjamin Cole said...

Scott:

As you know, the Bank of Japan has been conducting a QE program larger than that of the Fed's (relative to the size of the economy) and now owns one-third of Japanese national government bonds outstanding.

The BoJ does not pay IOER, and in fact has negative interest on reserves in some regards.

What explains the deflation in Japan, alongside negative IOER?

Is the IOER not really necessary, but rather a subsidy for US commercial banks, disguised as monetary policy?

Presently, though IOER and the reverse-repo program, the Fed appears to be artificially propping interest rates up, not down.



Scott Grannis said...

Contrary to popular opinion, Japan does not suffer from deflation. What Japan has enjoyed for the past 20 years is price stability. The Japanese consumer price index today is at the same level it was at 20 years ago. Two decades of zero inflation. I consider that a remarkable achievement.

I think Japan's growth problem today is a function of 1) demographics (the labor force is shrinking and the population is aging) and 2) oppressive fiscal policy (way too much government spending which suffocates the private sector).

The Bank of Japan is powerless to "stimulate" the economy, and the same goes for the Fed and the ECB.

Better fiscal policy is the answer to Japan's woes. The same goes for the US and the Eurozone.

Benjamin Cole said...

Scott:

"Japan’s GDP has been on a declining trend since 1997 when it was 523.5 trillion yen. The decline is due to low real growth (0.6% per year on average between 1997-2012) and outright deflation (-1.2% per year on average between 1997-2012)."

That is from Japan Macro Advisers.

https://www.japanmacroadvisors.com/page/category/economic-indicators/gdp-and-business-activity/gdp/

Japan's CPI has been nearly flat since 1993 or so, but the CPI tends to overstate inflation. Japan's GDP, in nominal terms, has been shrinking.

Inflation-adjusted per capita income grew rapidly in Japan in the post war era, but slowed dramatically in 1993, and never bounced back. It is higher today than in 1993. so there has been slow growth in the deflation era, but it ain't what it was.

In all, I would say the Japan experiment with minor deflation is disappointing. Sure, Japan has structural impediments, and every developed nation has structural impediments.

The US managed 3% real growth from 1982 to 2007 with 3% inflation, with a roughly equal collection of structural impediments.

I think moderate inflation helps lubricate the gears of modern economies. The private-sector rapidly evolves new good and services, generating bottlenecks, which are resolved through the price signal. That is good!

In addition, commercial banks have 80% of their loan portfolios in property. It is just reality. Deflation would destroy our banks. Even zero inflation would wreak havoc for many banks (some tea estate would go down), and we see what happens to the US economy when banks go down.

In a perfect world, maybe there is price stability, as measured. When central bankers die, they go to Deflationland and experience eternal euphoria.

For now, give me moderate inflation and robust growth!

Scott Grannis said...

Correlation is not causation. There is no mechanism that says that mild deflation causes weak growth. But there are plenty of obvious candidates for why Japanese growth has been slow: fiscal policy, excessive government spending, huge deficits, etc.

Scott Grannis said...

I recommend this post by John Cochrane which explores possible reasons for slow growth in the US, as I think it applies to Japan as well:

http://johnhcochrane.blogspot.com/2016/08/micro-vs-macro.html

Benjamin Cole said...

Scott: with about 80% of commercial bank loan portfolios placed in real estate, what would be the impact of deflation?

Once real estate values started to deflate, commercial bank would extend lower loan to value (LTV) ratio loans on property. The restriction of credit on property loans would lower property values even further.

How many businesses and property owners would choose to walk away from their property rather than honor a loan that exceeds the value of the property?

If you want to destroy America's commercial banking system, just institute real estate deflation for a couple decades!

Jean-Pierre Deslandes said...

Thanks you Scott for those (as usual for me) enlightening remarks…

If I understand correctly the massive amount of reserves now held by the banks make it difficult for the FED to raise interest rates just by raising the federal funds rate, it would be ineffective because the banks have so much reserves they don’t need much borrowing from each other. Therefore raising the federal funds rate would be mostly psychological, that is it would make for big headlines but would have little direct effect on the credit market.

To effectively raise rates the FED would also need to raise the rates it pays on reserves, which would cost money so they won’t do it lightly, and might even be a hard sell politically.

Am I understanding this correctly?

bt1138 said...

"people are happily "hoarding" some $2.3 trillion of extra money relative to their incomes"

This state of affairs underlines the fact that more tax cuts are probably not simulative in this environment. Corporations and wealthy individuals are already having trouble figuring out what to do with all the money they are sitting on.

Scott Grannis said...

Unknown: on the contrary, the public's unusually strong demand for money and safe assets is evidence of risk aversion. That money could be invested in new businesses, but people aren't sufficiently attracted to the risk/reward prospects of investing. Business investment has been unusually weak as the demand for money has increased; the two go hand in hand. What the economy needs is more investment, which in turn would result in more jobs and more demand. The best way to increase investment is to reduce the after tax rewards to investing. Make investment a compelling alternative to cash and you will see the demand for cash decline as the economy strengthens. Corporate taxes should be the first thing to cut.

Scott Grannis said...

Jean-Pierre: Before the Fed decided to pay interest on bank reserves, it would add or subtract reserves from the banking system. Reserves then existed only for the purpose of collateralizing banks' deposits. Reducing the supply of reserves would make them scarce, and that would usually result in higher interest rates, since banks would compete with each other for a scarce resource. Now reserves are abundant, so draining reserves would make no difference to their price. But now the Fed controls directly the interest rate on reserves by simply stating the interest rate it will pay. Raising the interest rate it pays is not difficult at all. In order to create an abundant supply of reserves, the Fed bought notes and bonds that yield an interest rate that currently is much higher than the interest rate the Fed pays on reserves. Last year the Fed transferred the difference, about $80 billion or so, to Treasury. Raising the interest rate it pays might result in the Fed earning less on its portfolio, but as long as the yield curve is positively sloped, the Fed will make money on its portfolio (which is equivalent to owning notes and bonds by borrowing money at a floating interest rate).

The decision to raise reserves is difficult only because the Fed, and the market, worry that higher interest rates might slow or otherwise damage the economy. No one knows what the correct level of short-term interest rates is, so the Fed has to experiment. By signaling its intent to raise rates, the Fed can judge by the market's response whether that will be good or bad. It's an imperfect system, but so far it has worked reasonably well.

Jean-Pierre Deslandes said...

Thank you Scott, I get it now, this is fascinating stuff.

As you say so far so good, instinctively this might be a bit surprising considering the massive amount of assets involved, one might easily think there needs to be a huge downside somewhere, no wonder so many people think so.

This is important to know, so many people believe that the US economy is growing slowly DESPITE massive stimulus and that in fact it’s desperately weak or even doomed. I’ve learned here exactly why this is not the case.

I don’t think I would have come to understand this without your blog, a precious resource for me, thanks again.

pg said...

Very interesting article Scott.
I would like to share some calculations I've made and get your thoughts on that.
Instead of thinking in a 'narrow' M2 definition of money, I did define
money = currency + sight & time deposits + money market funds + t-bills + commercial papers + MBS

I did my best to avoid double counting (e.g. t-bills accounted are the ones outside mmkt funds, etc).
This 'broad' definition could be motivated by the very low interest rates (which makes them closer substitutes) or by financial engineering, which allows for almost instantly liquidity in all this measures of money.

With this 'broad' money definition at hand, one can see that it grew at 9.8% per year from 1995 to 2000, then again by 9.8% from 2005 to 2007 and it has only increased by 2.6% since 2010.

With the real economy growing at close to 2% since 2010, it does not surprise me that inflation has been low. Any thoughts?

Scott Grannis said...

pg: Defining money very broadly (because financial engineering makes things like MBS very liquid) muddies the waters too much, in my opinion. Liquidity is one thing, stable prices are quite another (which MBS lack), and mixing the two together does not make the whole pot a money substitute.

Steve Fulton said...

Scott, great posts as always. What do you make of the recent $libor vs Euribor rate moves. Seems like there is another $ funding crisis brewing?
Steve Fulton

Scott Grannis said...

Steve: I don't have a good answer. I note that the TED spread has jumped to 50 or so, but has been flat for the past month. Ditto for OIS spreads and Euribor. 2-yr swap spreads are up to 25, but also flat for the month. Something's going on but it doesn't look serious yet. What do you think?

Steve Fulton said...

I'm not quite sure, I was going to check $/€ and $/¥ currency swap rates, those usually tell us if there is a $ funding issue but I haven't done it yet, but 12 month libor is now pretty much flat to the 10yr and that's a bit unusual especially if the fed is actually going to raise rates this year.

Jon said...

Scott,

I think you're saying that if banks were no paid interest on reserves at a sufficient rate, they would otherwise decided to loan out $$ against those reserves and make some net interest margin profit instead.

Why can't banks get paid interest on reserves and also loan out funds against those reserves? How is interest on reserves preventing a huge increase in the supply of money out there? I figure I am missing something obvious, but can't figure it out.

Scott Grannis said...

Jon, re IOER:

Prior to IOER, bank reserves paid no interest. They were a dead-weight loss on banks' balance sheets, yet banks were forced to hold them to collateralize their deposits. Banks thus wanted to keep their holdings of reserves to an absolute minimum, strictly as required to collateralize their deposits. This allowed the Fed to regulate bank because the Fed directly controlled the amount of reserves available to the banking system.

When the Fed switched to a QE regime (in which massive amounts of excess reserves were created), it became necessary for the Fed to have some way to affect banks' willingness to hold excess reserves, otherwise they would be encouraged to lend without any practical limit.

Under the new IOER regime which began in late 2008, the Fed pays interest on bank reserves that is similar to the interest rate on 3-mo T-bills. That means that banks no longer have an incentive to minimize their holdings of reserves, because reserves are now an interest-bearing asset. In fact, reserves are functionally equivalent to T-bills: a very high-quality, risk-free asset. Bear in mind also that today there are only $1.75 trillion of T-bills outstanding, and about $2 trillion of excess reserves. In contrast, total bank credit is currently about $12.5 trillion. Banks hold little or nothing in the way of T-bills, and neither does the Fed. T-bills are actually in short supply these days relative to total assets, which measure in the hundreds of trillions. I've argued in the past that QE was a way for the Fed to "transmogrify" notes and bonds into T-bill equivalents, thus satisfying the world's extreme demand for safe assets.

Because they earn interest, bank reserves today function as an attractive asset class, effectively expanding the relatively small amount of T-bills that are outstanding. Banks are still somewhat risk-averse these days, so holding reserves is an attractive proposition.

Of course, banks can continue to hold the $2 trillion of excess reserves that exist today, and at the same time greatly expand their lending activities, but that would amount to taking on increased leverage. Not many are willing to do that, it would appear, since there is no evidence of excessive lending or monetary-induced inflation.

Banks' incentive to lend more money on the margin is a function of the difference between the expected return on new lending and the risk-free return that the Fed pays on excess reserves. Banks will increase lending so long as they have sufficient reserves to do so and so long as new lending does not result in excessive leverage or a shaky balance sheet. Banks must also pay attention to bank capital requirements, because those requirements assign different weights to assets held depending on their perceived risk.

Bank reserves are considered to be of the highest quality, along with Treasury bills. If banks were to increase their lending, their balance sheets would become riskier, because lending to the private sector is more risky than lending to the government and because more lending would lead to a more leveraged balance sheet.

If the Fed were to stop paying interest on reserves, or if they were to pay a rate that was substantially less than the going rate for risk-free assets, banks would have an extra incentive to leverage their balance sheets, and that might not be desirable.

It is also important to note that paying interest on reserves is the only way the Fed has to set or target short-term interest rates, or to tighten or ease monetary policy. By increasing the interest it pays on reserves, the Fed effectively causes all short-term interest rates to rise, and that in turn increases the demand for money by making borrowing less attractive. Paying a rate that is too low would encourage lending throughout the economy, and that could lead to excessive money creation and ultimately rising inflation.

Jon said...

Thank you for the lengthy reply, it helps clear up my confusion.