Even though it has been a year or more since the BoE published this paper, I still see it cited very often by bloggers, often in support of specific types of endogenous money theories. I think this paper is extremely informative, although I don’t think it is perfect, and I do think the authors sometimes extrapolate conclusions that aren’t necessarily warranted. In this post I intend to strip down the paper to its core elements, and analyze what implications (if any) they have for economic theory.
The paper begins by stating “Saving does not by itself increase the deposits or ‘funds available’ for banks to lend.”
This is true, at an individual level. If an individual purchases something instead of saving, this would have probably no net effect on the level of bank deposits. But what about the macro level? To an economist is the ‘savings’ rate of the public, by which I mean in this specific case (because it’s not usually defined as such) the amount of money the public hold as deposits at banks, decreased? No, the money leaves one deposit for another, so nothing changes. Essentially the problem is the authors seem to to be underestimating how abstractly economist are defining ‘savers’, since all deposits are classed as such, this implies purchasing something simply results in one person’s ‘savings’ reduced while another person’s (the seller’s) ‘savings’ increased, exactly offsetting the reduction. This is the case usually at least, given that most purchases are electronic, or at the very least, cash will eventually be deposited in a bank account. The only exception are people who store all of their money as cash under a mattress (mostly criminals), but they are negligible in the grand scheme of things.
The problem is that when economists say banks are intermediaries, they are talking about contrasting this with a world in which banks did not exist. If this was the case, the only money would be base money (cash), or near money substitutes. This would necessarily make it much harder for any firm to attain funds, if there was no institution to intermediate between lenders and borrowers, it would be extremely costly for an individual to lend; they’d likely need to hire a lawyer to draw up a contract, and they would have little diversification and ability to absorb losses meaning the risk premium would be much higher, this would result in very high interest rates and largely reduced lending in the economy.
Banks on the other hand are able to massively reduce the transactions costs, and absorb losses, when they lend. Reducing interest rates and massively increasing the availability of credit. Ultimately, base money is required for this, even though ‘broad’ money is created as new ‘demand deposits’ whenever a bank lends. So if banks were introduced into this economy, and people had the same level of comfortableness with depositing money at banks as they do today in reality, almost everyone would immediately convert their cash into a demand deposit at the bank, effectively making everyone a saver and lender under this technical definition. In which case, banks would hold a large amount of base money (which they might further still hold as reserves at the central bank), in the meantime banks would also be lending, and the amount of credit would be extended significantly in the economy. Add a monetary authority that can create more base money, and eventually you’d have something that looks mostly indistinguishable from the economy today.
So this result of vastly more lending when you contrast between a society without banks and only base money, and a society with banks, is what is emphasized when banks are described as intermediaries. Of course, the end result is that almost everyone is using broad money rather than base money, resulting in the base money held by banks to be more or less constant (unless changed by the central bank), causing this intermediary relationship to break down when considering broad money rather than base money (a transfer of one person’s broad money to another person’s account has no net effect on base money and the level of ‘saving’ as described above), it still remains true that an initial condition for this lending to have occurred is for the public to swap their base money for demand deposits (lending), in the first place.
This means that the statement: “saving does not by itself increase the deposits or ‘funds available’ for banks to lend”, in today’s economy, would not be surprising from any standard theoretical perspective, and it would be a mistake to label this as a misconception. This was a rather lengthy response to one sentence, but it’s a statement I keep hearing so I feel an urge to present my thoughts on this.
The paper then goes on to describe how it’s mistaken to believe that if the central bank changes the amount of base money, the amount of lending must increase with it. This is widely agreed to be false, and it does have implications for the idea of a money multiplier, at least the more simple concept that an increase in base money is always ‘multiplied’ by the banking sector by expanding credit until all excess reserves become desired reserves. I question, however, if any significant amount of people in academia believe in this simple model today, and whether this statement is really surprising.
Still, it does pose problems for any monetarist model that relies on increases in base money causing changes in aggregate demand, because it might rely on the assumption that an increase in base money causes an increase in broad money to move along with it.
What’s more important is that they note that the central bank does not “fix the amount of reserves that are available”. This is more pertinent, because this means the banking sector does not need to worry about finding more reserves if they wish to extend credit, in other words not only is there no current binding constraint on lending, but there is also no practical upper bound constraint on normal lending behavior imposed by the level of bank reserves (I say practical because this still relies on ongoing behaviour of the central bank, if you look at things ceteris paribus, where the central bank does not change the level of bank reserves, then reserves would indeed have an upper bound constraint, but this is not how central banks work in reality). This would call into question anyone who might say, for instance, that ‘while excess reserves don’t cause increased lending today, it does allow for banks to over expand the amount of lending they might do in the future, when the economy picks up again’. This would appear to be false, if the central bank will always increase reserves when the banking sector needs it, then current excess reserves would not affect the ability of banks to lend in the future.
This of course means that the central bank puts a constraint on the amount of lending in the economy using another instrument: interest rates. I think this is interesting, because it does challenge some endogenous money theorists who appear to act as if there is absolutely no exogenous constraint and that the central bank is completely irrelevant. This is not true, according to the bank of England; by controlling interest rates they necessarily affect lending decisions; if the rate of return on safe assets is high, banks will choose to charge higher interest on loans, which will result in less borrowing by the public, and ultimately less broad money being created.
Of course, this has little implication for modern macro models (New Keynesian DSGE), as they generally analyze changes in monetary policy as changes in interest rates, in which case this description of banking would not conflict with these models. They do however conflict with any monetarist who might consider a sudden increase in base money as expansionary ceteris paribus. Not that this seems to bother market monetarists, Nick Rowe for instance would emphasize various additional transmission mechanisms policies like Quantative-Easing could have to move demand, I remain less convinced.
Although speaking of QE, the paper does provide detail on this type of monetary policy and identifies a transmission mechanism. The first thing to note is that it immediately challenges the caricature of QE simply ‘giving money to bankers’. At least in the UK, with QE the central bank mainly purchases assets from the non bank private sector. Banks in this case are simply intermediaries between the central bank and the asset holder, they do gain a large amount of reserves in the process, but these reserves are an irrelevant by-product of the policy, and are matched by an accompanying liability by the former asset holder.
The paper instead describes the transmission mechanism thusly: “the central bank purchases a quantity of assets, financed by the creation of broad money and a corresponding increase in the amount of central bank reserves. The sellers of the assets will be left holding the newly created deposits in place of government bonds. They will be likely to be holding more money than they would like, relative to other assets that they wish to hold. They will therefore want to rebalance their portfolios, for example by using the new deposits to buy higher-yielding assets such as bonds and shares issued by companies — leading to the ‘hot potato’ effect discussed earlier. This will raise the value of those assets and lower the cost to companies of raising funds in these markets.”
This would again seem to challenge some endogenous money theorists who argue that QE has absolutely no effect whatsoever. The Bank of England is not on their side in this case. The central bank can create broad money after all, by purchasing directly from the non bank private sector. The portfolio re-balancing effect is a direct transmission mechanism that should in theory have an effect on aggregate demand. A mechanism that doesn’t rely on bank behaviour at all!
Still, don’t celebrate just yet monetarists. While this does increase broad money, it doesn’t have a substantial impact on an individuals actual net worth as they had to lose an asset of equal value in the process; people are made richer only to the extent that it increases asset values more than they otherwise would have been. Since the assets the central bank are legally allowed to buy are usually already fairly safe and liquid, the increase in their value (or decline in their yield) is mild, so the amount people are made richer by is also mild, which means I think the effect this policy has on AD is ultimately… well, mild. Mild compared to for instance a ‘helicopter drop’, which would be a no strings attached increase in the amount of broad money somebody holds.
So in summary, the implications for economic theory that the Bank of England’s stance on money creation has is as follows:
- It would appear to have no bearing on standard New Keynesian economic models, which analyze interest rates rather than changes in base money.
- It challenges the simple concept of a strict money multiplier, but nobody really uses that anymore.
- It challenges some monetarists who might have a less nuanced theory on the relationship between base money, broad money and demand, and who might expect that sudden increase in base money would have any substantial impact on bank behaviour, even if permanent.
- It challenges some endogenous money theorists who might claim that the central bank does not determine the amount of broad money at all, by identifying how it imposes limits on money creation by controlling interest rates.
- It might also challenge endogenous money theorists who claim that policies like QE have absolutely no effect in the UK, by identifying a transmission mechanism and emphasizing that assets are bought from the non-bank private sector.
- The identified transmission mechanism is not as strong as one that would involve a strict money multiplier on the newly created bank reserves.
All in all, I don’t expect this to substantially change anyone else’s stance on monetary policy, but anyone not familiar (if they still exist) with these concepts presented by the BoE would do well to acquaint themselves with them in order to save themselves from potential embarrassment.