I read an August 2020 Bank of England Staff Working Paper (No.883) – Does quantitative easing boost bank lending to the real economy or cause other bank asset reallocation? The case of the UK – recently, which investigates whether the large bond-buying program of the Bank stimulates bank lending. They find that there was no stimulus to lending. Which would only be a surprise if one thought that mainstream monetary economics had anything useful to say. Modern Monetary Theory (MMT) economists were not at all surprised by this finding.The reality is that the lack of bank lending during the GFC had nothing to do with a liquidity shortfall within the banking sector. It had all to do with a lack of credit-worthy borrowers – which should tell you that bank reserves do not constrain bank lending. The fact that mainstream institutions such as the Bank of England are now publishing this sort of research, which undermines the mainstream theory is the interesting fact.
The reason this is an interesting exercise is that it uses a unique dataset, which includes the exact banks “that receive reserve injections through the BOEs APP (QE banks)”.
This allows the researchers to conduct what is called a “difference-in-differences” methodology, which essentially pinpoints reasons for variations between observations in the dataset.
The practical importance of studies like this is that they provide real world evidence to allow us to understand the way the monetary system operates and to reject the way mainstream economists construct those operations.
In a recent blog post – ECB nearly comes clean – higher fiscal deficits, higher QE (July 20, 2021) – I discussed a report that the Economic Affairs Committee of the UK House of Lords has issued on July 16, 2021 – Quantitative easing: a dangerous addiction? (July 16, 2021) – where they rehearsed all the usual errors relating to QE.
These are the errors that mainstream economists introduce into the policy debate and their teaching programs.
So real world evidence refuting basic mainstream propositions is always welcome.
I also considered these questions in a series of blog posts 12 years ago – when people were starting to wonder what the impacts of the large central bank bond-buying programs would be.
They were scared that the so-called ‘money printing’ interventions from the various central banks early on in the GFC would be inflationary.
The reason they thought that is because they had either been taught that in economics programs at university or because they had been listening to politicians and their crony mainstream economists relentlessly pushing that message in the media.
This set of blog posts was designed back then to set the record straight.
1. Quantitative easing 101 (March 13, 2009).
2. Building bank reserves will not expand credit (December 13, 2009).
3. Building bank reserves is not inflationary (December 14, 2009).
4. Lending is capital – not reserve-constrained (April 5, 2010).
It is amazing that it took so long for bank officials to acknowledge what Modern Monetary Theory (MMT) economists knew all along.
QE does not increase the risk of inflation in an economy.
If it did then Japan should have been hyperinflating by now given the 20 years or so of very large government bond purchases by Bank of Japan.
The problem that the mainstream economists encounter here is that they start with a flawed view of the private banking system.
They believe that bank lending is constrained by the reserves they have in the vault and that quantitative easing solves this shortage by providing those reserves.
So banks are conceived as being ‘desks’ where officials wait for cash to come in in the the form of deposits, which they loan out, profiting from the difference between deposit and loan rates.
But this is a completely incorrect depiction of how banks operate.
Bank lending is not ‘reserve constrained’.
Banks extend loans to any credit worthy customer they can find and then worry about their reserve positions afterwards.
Remember the role of bank reserves is to facilitate the clearing system for transactions that have cross-bank implications.
So if Bank A creates a loan which simultaneously creates a deposit in its books, the person can either draw down the deposit and spend the cash in a business that banks with Bank A or spend in a business that banks elsewhere, say, Bank B.
In the former case, there is no clearing issue. Bank A simply transfers the deposit funds from the customer to the business.
In the latter case, Bank B will call on Bank A to transfer the funds into the account of its business customer.
Those transfers are what the clearing house is about and there are millions of such transfers being done on a daily basis.
That is what bank reserves are for.
So Banks A and B have accounts at the central bank and the relevant entries are made in those accounts to satisfy the transaction noted above.
The banks never loan out reserves to commercial customers (borrowers).
They sometimes loan out excess reserves to each other to smooth out the clearing system.
If banks are short of reserves (their reserve accounts have to be in positive balance each day and in some countries central banks require certain ratios to be maintained) then they borrow from each other in the interbank market or, ultimately, they will borrow from the central bank through the so-called discount window.
They are reluctant to use the latter facility because it normally carries a penalty (higher interest cost).
The point is that building bank reserves will not increase the bank’s capacity to make loans.
Loans create deposits which generate reserves not the other way around.
The major institutional constraints on bank lending (other than a stream of credit worthy customers) are expressed in the capital adequacy requirements set by the Bank of International Settlements (BIS) which is the central bank to the central bankers.
They relate to asset quality and required capital that the banks must hold.
These requirements manifest in the lending rates that the banks charge customers
But despite what is taught in mainstream courses in monetary economics, bank lending is never constrained by a lack of reserves.
Which is why MMT economists never considered QE to be an appropriate vehicle for increasing bank lending in order to stimulate the economy.
Quantitative easing merely involves the central bank buying bonds (or other bank assets) in exchange for deposits made by the central bank in the commercial banking system – that is, crediting their reserve accounts.
Quantitative easing is really just an accounting adjustment in the various accounts to reflect the asset exchange. The commercial banks get a new deposit (central bank funds) and they reduce their holdings of the asset they sell.
It was always obvious that the reason the commercial banks were reluctant to originate loans during the GFC was because they were not convinced there were credit worthy customers on their doorstep.
Further, after years of lax assessment practices in relation to credit-worthiness, the banks tightened their rules once the GFC threatened their solvency.
The 2020 Bank of England report has reflected on this issue and constructed a unique dataset.
The paper still runs the line that:
Banks receive cheap liquidity, in the form of central bank reserves injections, as a direct effect of the asset purchase programs. This should encourage banks to lend more to households and businesses, transmitting the impact to the real economy.
Which continues the myth that bank lending is in some way reserve constrained.
But their mission – to “the impact of the two main waves (March 2009 to November 2009 and October 2011 to October 2012) of the UK asset purchase program (APP), also referred to as quantitative easing (QE) on UK banks’ balance sheets” and to see whether there was any association with the bank lending practices is interesting.
The empirical data is clear:
1. The Bank of England QE program began in March 2009, which followed the US Federal Reserve launching a program in November 2008.
Of course, the Bank of Japan launched its large-scale bond buying program in 2001. So we already knew what happened.
But the mainstream economists kept dismissing the Japanese case as ‘special’ due to cultural differences.
That was a poor way of saying they didn’t have a clue why the way in which the Japanese monetary system responded violated all the textbook predictions that the mainstream economists made.
2. There was no visible impact on bank lending, which showed a “fall or little to no growth” in the period after the QE program began in Britain.
Loans to non-financial businesses fell sharply between early 2009 and 2014.
The authors assembled a dataset which allowed them to compare “UK banks that received reserve injections from APF, called QE-banks, with those that did not.”
They had 18 years of data from “the second half of 2000 to the first half of 2018”.
The authors note that:
When the Bank of England conducts QE, reserves are credited to the reserves account of the seller’s bank, and that bank then credits the seller’s deposit account with the same amount. Hence, banks involved in QE operations (QE banks) initially receive additional liquidity (as reserves), while other banks (non-QE banks) do not.
They note that the initial boost to reserve balances for the QE-banks could “leak” to the other banks, through the payments system – in the way I described above.
However, due to the institutional structure in Britain the QE-banks “mostly do business with a small subset of banks who are also participants in the Bank’s QE operations” and so the leakage was considered small.
The authors also note that the “money multiplier” does not exist in Britain and:
… the supply of credit is mainly driven by banks’ ability and/or incentives to lend.
Yet they still claimed that QE “improves banks’ ability to lend” but other factors like “depleted capital positions” and other regulatory matters intervened.
Their results are interesting:
1. “the additional liquidity did not incentivise QE-bank to increase lending, relative to the control group. There is no evidence suggesting that these results were driven by changes in relative demand for loans the two groups faced.”
2. They “no evidence that lower lending by QE banks after the two QE waves is caused by differences in the demand for loans between the treatment and control groups.”
3. “Relative to the control group, QE banks increased reserves and reduced lending to other banks after QE1. They also increased holdings of government securities, especially after QE2. This suggests that QE banks reallocated their resources from lending towards government securities with low risk weights.”
This is significant.
There was a lot going on at that time.
The Banks scrambled to increase the quality of their capital base – which is why they increased their holdings of government bonds which were considered lower risk than retail loans.
Banks also reduced their exposure to the European debt crisis by reallocating their assets to British bonds.
So there was asset reallocation going on but no increase in bank lending as a result of the rising reserve balances.
The overall conclusion of the paper is that “if the policy objective is to provide an additional boost to the economy through supporting bank lending in a time of stress and uncertainty, it might be valuable consider using alternative credit easing tools.”
Which really is a confession.
QE was never going to break the deadlock in the loans market that arose during the GFC as the deep uncertainty drove would be borrowers underground.
Households, fearing unemployment, did not want to take the risk.
Businesses, knowing that household spending was constrained, and with excess productive capital, had little incentive to borrow more even at lower rates.
Another piece of evidence that tells any aspiring student not to study mainstream monetary economics, unless you like fiction.
That is enough for today!
(c) Copyright 2021 William Mitchell. All Rights Reserved.