Today we will catch up on a topic that I first addressed in 2019, regarding new regulations for banks arising from Basel III, and bank reserves, which are a form of base money and a liability on the Federal Reserve’s balance sheet.
October 21, 2019: Basel III and Demand for Bank Reserves
October 29, 2019: Basel III and Demand for Bank Reserves #2: Return to Normalcy
The conclusion of those items were that new regulations arising from Basel III and finalized in November 2010, and phased in through the end of 2019, led banks to hold much higher levels of bank reserves (Federal Reserve deposits) than had been common previously. This was actually a return to old-fashioned banking principles, of the 1950s and decades previous. At the end of 2019, bank reserves were low compared to this new requirement, setting up a minor crisis as banks scrambled for reserves that didn’t exist. This drove short-term lending rates over 8% in some instances.
In 2020, the Federal Reserve increased bank reserves dramatically, by buying Treasury bonds. This resolved the shortage that existed at 2019-end, and also, added more, which was required because banks’ balance sheets also expanded dramatically, leading to a need for more reserves. By early 2021, banks’ reserve requirements seemed to be entirely satisfied. The drama then became: what will happen when the Fed expands more than is necessary to accommodate these new Basel III requirements? But, this didn’t happen. The Fed was on track to do this, but canceled out (“sterilized”) further expansion through the expansion of the Reverse Repo function, which reduced the monetary base.
This brings us to today, where it seems that banks have fully satisfied their reserve requirements, but things have not obviously exceeded that level by very much. Unfortunately, the precise Basel III requirements are rather complicated, and vary from bank to bank depending on their assets and liabilities, so it is hard to say exactly what is required and what is in excess. Perhaps someone with more expertise in these matters will inform us in time.
The significance of this is broader. It provides the primary answer to the question: “How can the Federal Reserve create so much new money, without much inflationary consequences?” Actually, there were inflationary consequences. The Fed’s change of direction and expansion in late 2019, although necessary to meet the new reserve requirements, was nevertheless accompanied by a decline in the dollar’s value vs. gold from around $1250/oz. to around $1800/oz. today. That is actually a pretty big move, and the inflationary effects of that are being felt today. But even this is, you could say, somewhat modest compared to the size of monetary expansion in 2020.
Banks began to adopt their new reserve policies somewhat informally, in the middle of the 2008 crisis. This resulted in a huge expansion of Federal Reserve discount lending. The 2010 agreement was basically the formalization of this change. Banks were running super-low reserves basically to maximize profitability. Managements tended to be drawn into this risky situation because, if they didn’t, they would show lower profits, and could be fired. So, banks agreed to impose requirements on everyone, so that they wouldn’t have this competitiveness issue.
This desire, and later requirement, to hold much larger reserves led to the creation of new bank reserves, with the Federal Reserve buying Treasury and Agency securities to accomplish this goal. This was basically a free windfall for the Federal government. The Federal government could have used this windfall to effectively reduce its outstanding debt, which would have been nice. It didn’t have to spend any more money. But that didn’t happen. Instead, sensing an opportunity to issue bonds in unprecedented (since WWII) quantity, since the Federal Reserve was going to be buying them, the Federal Government ran huge deficits. I see this as a sort of instinctual, animalistic impulse of the political system, whose members didn’t really understand what was happening, but sensed an opportunity.
Let’s see how it looks now.
We can see the minimal levels of bank reserves prior to 2008. This was really crazy, and a major factor in the financial crisis in 2008. (I’m using the monetary base “ex-Treasury,” without including the Treasury’s account at the Federal Reserve. You can argue, legitimately, that you should include that, so go ahead if you want to.)
You could think of bank reserves as “money in your wallet” — banknotes. Plus, you also have, in your wallet, a debit card (debt that the bank owes to you), and a credit card (an offer to borrow money from a bank). You have only $20 of cash, but since you have a debit card and credit card, you can make many payments through the month. During the 2008 crisis, banks effectively stopped lending to each other. They canceled each others’ credit cards. This left them with just the $20, plus the money that other banks owed them. The other banks also had only $20. Banks asked: “Hey, that $10,000 I lent you, can you pay it back? I need the cash.” But the other bank only had $20 too, and couldn’t pay back the $10,000.
In the end, they borrowed the $10,000 from the Federal Reserve, from the discount window. This led to the huge expansion in bank reserves after 2008. This discount lending was later replaced by purchases of Treasury and Agency bonds, in QE beginning in 2009.
Here we see what banks’ balance sheets have looked like, since 1973. “Cash” consists of Federal Reserve deposits (base money), plus callable short-term lending. We can see that there was a longstanding pattern of holding about 10%-12% of Assets in Cash during the 1970s (and also, in previous decades, which I will get to). During the 1980s and 1990s, this “cash” was drawn down, to improve profitability. Today, cash is again back to its traditional 10%-12% level, and actually somewhat above this now, to around 15%. Also, unlike in the past, when “cash” consisted of a mix of Fed deposits (bank reserves) and short-term lending, “cash” today is nearly 100% Fed deposits. It looks like the transition to the new framework of high cash (Fed Reserves) for banks is complete, even moderately in excess.
We can see Bank Reserves at the Fed falling to absolutely teensyweensy levels in the 2000s, but today they are nearly 100% (1.0) of “cash.” This is good. It is a nice, conservative way to run a bank. There is plenty of cash on hand, in case of crisis. This was the way banks were typically run, before about 1960.
Here is Interbank Lending, the primary component of “cash” besides Fed Reserves. As we can see, this plummeted after 2008 and has nearly disappeared, with this data series discontinued at the end of 2017.
Here we have the ratio of Fed bank reserves to the Total Assets of depository institutions, going back to 1950. We can see that this began the 1950s around 9%, and then fell over time to very low levels. The recent peak, on this chart, is higher than the chart above, because it includes both Domestic and Foreign banks. (The last year on the chart is 2019, which is why it doesn’t show the recent expansion.) I don’t think Foreign banks were significant in the 1950s and 1960s, so the ratio from that time can be considered representative for Domestic banks alone. This is what I mean by “going back to 1950s-style banking.”
Here are the reserves in excess of requirements, at least back to 1959. In those days too, banks held high reserves basically because they were required to. Their excess reserves were about 3% of total reserves.
Here is the same reserves/total assets ratio, from 1916-1941:
Unfortunately, I don’t have handy info for the period 1941-1950, but you get the idea.
We can ignore the period before 1918, because the Federal Reserve was just getting started.
We see that there was about a 6% ratio of reserves/assets during the 1920s. This then rose to very high levels in the 1930s, as banks took a very risk-averse stance. The ratio even spent a little time above 20% (it is around 15% today).
So, we can see that 6% (1920s) or 9%+ (1950s) was a common reserve/asset ratio in the past, even climbing above 20% in the difficulties of the Great Depression. We can see that the situation before 2008 was certainly unusual, and perilous.
A little description of the pre-1914 era is here:
Deposits are not the same as Total Assets, but they are pretty close. The high reserve/deposit ratio of the 1950s was also common in the pre-1914 era. It would be nice to play a little bit with more detailed banking statistics from the pre-1914 era, which I think are available via the Office of the Comptroller of the Currency.
This expansion of bank reserves since 2008, although done to address a real need, nevertheless has been accompanied by a decline in the USD’s value vs. gold, which I take to reflect a real “inflationary” decline in USD value. We see that the USD was around $900/oz. of gold in 2008, and is around $1800/oz. today, a neat 50% decline in USD value vs. gold, which I take to be a pretty good indicator of “stable value.” There was a period of stability in 2013-early 2019, which I’ve called the “Yellen gold standard” period. The breakdown in late 2019 and early 2020 corresponds to the Fed’s move toward an easier/expansionary stance, to relieve the bank reserve shortage at the end of 2019, and the expansion in 2020.
Thus, the overall effect of this dramatic expansion of the base money supply, since 2008, has been a certain amount of “inflation,” but not nearly so much as you might expect. A 50% decline over 13 years is actually a pretty big deal, but spread over a long enough time that it is not easily perceptible.
I don’t think the Federal Reserve has had a very good idea of this notion of banks requiring a lot of new reserves due to Basel III, which were phased in over time, with full phase-in in 2019. We can see that is the case because the Fed was messing things up pretty seriously even as late as the end of 2019. They wouldn’t have done that, if they had a good idea of banks’ new reserve requirements. So, it was a sort of “feeling your way in the dark” process I think. Today, I think the Federal Reserve probably does understand things, at least as well as our exposition here, and probably a lot better since they can just ask banks about it directly. Bankers tend to be not very good about these broader macroeconomic, systemic or central banking topics, but they have a damn good idea of their own requirements of business, down to the last penny.
Now, let’s come to the political system, or fiscal policy.
This expansion by the Federal Reserve, necessary to return banks to a “1950s-style” balance sheet with high reserves, was accomplished by the Federal Reserve buying a lot of Treasury bonds and MBS. Since MBS falls into the category of “low risk, liquid” debt, which is required even by regulation of a lot of pension funds, life insurers and so forth, buying of MBS effectively increases the demand for Treasury bonds.
We can see that the Fed’s Total Assets rose about $1.3 trillion in 2008-2009.
Wonder of wonders, that was just about the same size as the Federal Deficit in 2009. Although I don’t think there was ever any outright coordination, the size of the Federal Deficit “just happened” to be the same amount as the amount of money the Federal Reserve created or “printed” that year. The same thing happened in 2020. Plus, there was steady expansion of the Fed’s balance sheet in between (as Basel III was progressively phased in), which was accompanied also by persistently high deficits/GDP even though the economy was reasonably healthy.
This was a deficit of just about 10% of GDP in 2009, the largest since World War II. In 2020, the deficit was even larger, around 15% of GDP. These are ridiculous numbers.
From the standpoint of a politician, and the advisors to politicians such as MMT advocate Stephanie Kelton, it seems like “we printed up a bunch of money and got away with it.” They want to try it again. I think it is safe to say that one impetus behind ramping up spending still more, in recent proposals, is the idea that it will be financed — somehow, we know not how — by the Federal Reserve, just as has been their experience for the last 13 years.
But, at least to the extent that monetary expansion was justified by banks’ new regulatory requirements, that process is done, and even, somewhat in excess. This sets up an interesting conflict, between a Federal Reserve that, I think, has an understanding that they have reached the end of the line, and further expansion will create significant risk of a “loss of confidence” that may produce dramatic currency and bond market moves, and a political system that will pressure them to pick up the tab for their ongoing spending extravaganza. This may be a central drama of 2022.
