"You and I can only lend money if we've got some to lend; but banks can create money and debt simultaneously at the stroke of a pen". Might commercial banks be the cause of the increase in debt over the last few years?
Maybe partly, but basically no. You can't just look at one side of the market for debt; you have to consider both supply and demand. And when you are talking about macroeconomic things, like total amounts of debt and money, you cannot just do partial equilibrium analysis; you have to do general (dis)equilibrium analysis and look at the economy as a whole. It's the only way to avoid fallacies of composition.
This is the third (and last) of three posts on why debt increased.The first post ignored money. The second post introduced money, but ignored commercial banks. This third post introduces commercial banks.
A financial intermediary is a firm whose primary business is borrowing and lending. A bank is a financial intermediary whose liabilities are money, by which I mean a medium of exchange. A commercial bank is a bank whose aim is to maximise profits (unlike a central bank which has macroeconomic objectives). And a fractional reserve commercial bank is one which promises to redeem its money liabilities in central bank money, but keeps less than 100% reserves of central bank money ("fractional" could include 0%). Those are the banks I'm talking about.
So, if you could create money and loans at the stroke of a pen (and banks can, by simply crediting the borrower's chequing account) why wouldn't you do so? And do an unlimited amount too?
There are 3 things that may impose a limit:
1. Capital reserves. Lending is risky. The bank may not get paid back. Banks ought to watch the ratio between their capital reserves and total amount of risky lending, and that ratio may impose a limit on their lending. I'm going to ignore that limit.
2. Currency reserves. If one bank lends Andy $100, by crediting his chequing account with $100, the individual bank may need to have $100 currency in reserves because it is likely that Andy will use that $100 loan to buy something from Bill, and if Bill banks at a second bank the second bank will want to collect its $100 reserves from the first bank. But for the banking system as a whole there is no loss of reserves, unless someone withdraws currency. Currency reserves impose a limit on aggregate bank lending if individual banks desire a reserve/deposit ratio above 0%, or if people desire a currency/deposit ratio above 0%. (This is the limit on expansion of bank lending and money creation typically discussed in ECON1000). That is not the limit I'm focusing on.
3. Profitability. The demand for bank money and the demand for bank loans together determine whether it will be profitable for banks in aggregate to create more money and loans. This is the limit I am focusing on.
Start with the model from my second post. Suppose it's an even year, in which red apples grow well, and green apples grow badly. So the red growers have temporarily high income, and want to lend; and the green growers have temporarily low income, and want to borrow. So there's debt, because the greens borrow from the reds. And everybody holds (base) money (currency), because money is a medium of exchange and debt isn't.
What happens when we introduce banks?
Banks earn their income from the interest rate spread. Suppose we have a constant money stock, and no real growth, so the inflation rate is zero. So currency pays zero interest, both nominal and real. But debt pays a positive rate of interest. So there's a spread between the interest rates on bonds and currency. If people are willing to hold some of their money as bank deposits, at a rate of interest lower than the rate of interest on bonds (and they should, because by assumption bank deposits can be used as a medium of exchange, and bonds can't), then banks can earn income on that spread. If that spread exceeds banks' administrative costs, profit-maximising banks will expand loans and deposits until at the margin the spread just equals the marginal admin cost.
What does the new equilibrium look like?
Banks create money. The nominal supply of money will be higher, and so the price level will be higher, unless the central bank reduces the supply of currency to keep the price level the same (an inflation-targeting central bank would do so).
The real money supply will probably be higher. That's because bank money either pays interest, or is more convenient in some respects than currency. (It must be better in some respect for some people to hold part of their money as bank money, otherwise they would stick to holding only currency).
Banks create loans. Does that mean the total quantity of debt will be higher? No. Or rather, not obviously. Before banks appeared, reds lent directly to greens. Now some of that lending goes through banks; reds lend to banks, and banks lend to greens. If bank-intermediated lending simply displaces direct lending, the total quantity of debt in real terms will stay the same. (Except each $1 of intermediated lending gets counted twice as debt: once as a liability of the bank; and once as a liability of the ultimate borrower.)
In order to get an increase in real debt (ignoring the double-counting issue), we need ultimate lenders to want to lend more (which means consuming fewer apples themselves), and ultimate borrowers to want to borrow more (which means consuming more apples themselves), at the same time. A rise in the rate of interest would do the first but not the second, and a fall in the rate of interest would do the second but not the first. How can banks do both? How can the introduction of banks make lending more attractive to the lenders, and borrowing more attractive to the borrowers?
The answer is clear. If the lenders lend to banks, their loans are a medium of exchange, and so are more attractive than direct loans. This can lead the red savers to save more, and lend more, even at a lower rate of interest. And a lower rate of interest can lead the green borrowers to consume more, and borrow more. It is exactly as is the banks, for a fee, could stamp direct debt with a certification of moneyness and make it more attractive to the saver and lender who was buying it. It's like removing an illiquidity tax on borrowing and lending. Or reducing transactions costs.
If either ultimate borrowing or ultimate lending were totally interest-inelastic (as they would be if consumption were totally interest-inelastic) introducing banks would not affect total real debt. But if both depend on interest rates or liquidity, then introducing banks will increase real borrowing and lending.
That's the long run effects of banks, when prices are flexible. What about the short run, when the price level is fixed?
It's a bit bizarre to think of the introduction of banks as a short-run experiment. Suppose instead that banks expand, because their administrative costs fall, for example.
Well, if the central bank takes no offsetting action, the money supply will expand, and so aggregate demand will expand, and output and employment will rise. The effects will be much the same as if the central bank increased the money supply, by printing more currency. Since that was covered in my second post, let's suppose the central bank does take offsetting action, and reduces the amount of currency in order to prevent any expansion in aggregate demand (and subsequent rise in prices).
The answer is then rather boring. As administrative costs fall, profit-maximising banks will compete to expand their loans and deposits, and spreads will fall. Banks' lending rates will fall, and borrowing rates (on deposits) will rise. Red lenders save more and deposit more with banks; green borrowers consume more and borrow more from banks. Just the same as the effect of introducing banks, writ small.
OK. Last post from me for a while on debt, and what caused it to increase. Time for a change.