How increasing banks’ capital reduces lending

This is a controversial topic. Since the beginning of the financial crisis, there hasn’t been a single week without someone calling for increased banks’ capitalisation. What does it mean in practice?

Banks fund their loans and investments through several main channels: customer deposits (retail and corporate), interbank deposits, short and long-term wholesale borrowings, and equity. Equity represents around 3 to 7% of banks’ funding structure in developed-markets, i.e. equity funds 3 to 7% of the bank’s loans and investments. This has led many to say that banks are ‘over-leveraged’, as the rest of the funding structure is effectively debt, under one form or another. Under current and future Basel 3 rules, banks are also allowed to count some form of loss-absorbing hybrid debt and preference shares as complementary capital, on top of shareholders’ equity. This ‘Tier 1’ capital should reach 6% of risk-weighted assets (not of total assets, see my previous post) by 2019 (up from 4.5% previously).

Many people think it is not enough. In the UK, Sir John Vickers proposed that equity should fund 20% of the banks RWAs (up from the 10% he recommended with his Independent Commission on Banking). In the US, calls for higher capital requirements are also very common (see here, here, here). Most of them point to the fact that banks’ equity level used to be much higher in the past than it is now. I’ll come back to this claim and many others on capital in another post.

Today I only want to address the main claim backing the ‘more equity is always better’ argument: that increasing equity (or more generally, regulatory capital) does not negatively impact the banking sector’s lending ability. Research by Anat Admati and Martin Hellwig has provided the main intellectual foundation to proponents of increased capital requirements (see also their now famous book, The Bankers’ New Clothes, which has received positive comments from regulators and most of the financial media, and which I’ve read and will try to review when I have the time). Their argument about capitalisation looks convincing and has not really been challenged so far on theoretical grounds: capital does not represent money set aside for safety which could otherwise be used to lend, as some foolish bankers would like us to believe. Why? Because equity is also already used to lend. Therefore, we can increase equity and lending will not be constrained as a result. They are right (as highlighted at the beginning of my post). However, they are also wrong.

Let me take a very simple financial system, only comprising banks as financial intermediaries. Here are my assumptions:

– High-powered money supply (money base) M0 is fixed and there is no physical cash

– M1: money supply following deposit expansion through the money multiplier

– Reserve requirements (RR): 10% of deposits (even though in most developed countries this figure is closer to 1 or 2% or even non-existent nowadays)

Before considering two different scenarios, let’s remember that we live in a fractional reserve banking world. When you deposit money in a bank, the bank lends out a portion of it to other people (or invests it in securities). Effectively, the bank never has all of its depositors’ money at the same time. However, deposits are still considered as part of the money supply. Why? Think about your own reaction when you put money in a bank: you consider money as yours and redeemable on demand. So you hold your physical cash balances at a minimum, considering that you can go get the rest of your money whenever you want anyway. And your consumption pattern reflects your whole money holding (cash on hand and deposits), not only the cash that you have in your wallet.

As a result the money multiplier applies, according to reserve requirements. What is it? In such a system, banks can effectively lend out multiple times the funds that have originally been deposited, as long as they keep enough of them to satisfy daily withdrawal. As such, banks create money (through new deposits). In a world with 10% reserve requirements, the money multiplier is 1/RR = 1/0.1 = 10, meaning the banking system can potentially multiply the original deposit base up to 10 times through lending.

This is where things differ regarding equity. Equity is not a deposit. It is not subject to reserve requirements (and hence to the money multiplier). Once you’ve invested in a bank’s equity, you don’t consider this money as yours anymore (you only have a claim on it that cannot be used for anything else). The only thing you can do is sell your stake at some point in the future to generate cash. As such, equity funding is a kind of 100%-reserve banking system, i.e. equity transfers money instead of creating it. A banking system 100%-funded through equity would be similar to a 100%-reserve banking system, with banks essentially becoming some sort of mutual funds (and deposits not being used at all for investments).

Alright. Now let’s now see what happens if banks’ capital requirements is 10% of their assets. Ex-post, after applying the deposit multiplier (when the system if ‘fully loaded’), M1 should be comprised of 10% of equity and 90% of deposits. Consequently, to figure out the original ability of the system to lend and create extra-deposit, we need to work backward in order to find out the ex-ante money supply structure. Rebasing the 10%/90% M1 structure gives a 53%/47% M0 structure after dividing the deposit base by the deposit multiplier, equivalent to M0 = 19% of M1. It also means that the system’s fully-loaded state (M1, after monetary expansion through fractional reserve lending) represents 526% of the original money supply M0.

What if banks’ capital requirements are raised to 20% of their assets? Ex-post, after applying the deposit multiplier, M1 should be comprised of 20% of equity and 80% of deposits. The 20%/80% M1 structure gives a 71%/29% M0 structure once divided by the deposit multiplier, equivalent to M0 = 28% of M1. It also means that the fully-loaded state (M1) represents 357% of the original money supply M0.

Clearly, increasing capital requirements led to a lower potential money supply as banks were not able to lend as much as before as a larger part of the money supply was not subject to fractional reserves anymore. Admittedly, this is a very simple scenario that might not accurately reflect the effects of the various near-moneys and injection of reserves by central banks of the real world. However, it does show that Admati and Hellwig’s claim is not that simple and straightforward. They also never explain how capital requirements can be smoothly increased. In real life, the transition process can be quite painful as we witness every day at the moment (banks cutting lending in order not to issue new equity, etc).

Don’t get me wrong though. I am in favour of banks holding more equity. But I don’t want to force them to do so. There are various historical reasons that explain why equity as a percentage of assets is low nowadays, and most of them are due to….the influence of government’s policies. Surprising heh?