The structure of banks to a certain extent bridge these mismatches, which explains why they are the centre of financial markets. However, non-bank financial instruments can be structured to bridge the gap. It is therefore that there is a continual blurring between bank and non-bank finance as they attempt to move into each other's turf. It is also unsurprising that the so-called “crypto community” has ended up re-inventing the structures of traditional finance, since even internet money faces the same economic forces.
Borrowers generally want the following.
Borrow at as low an interest rate as possible.
Long borrowing maturity, with an option to pay back early.
Some borrowers are quite willing to take on large loans and not worry about the consequences of default. (This is not universal, in the decades after the Great Depression financial institutions worked to decrease debt aversion among the public.)
Lenders are more varied, with a division by lending horizon.
In some cases, they want to be able to liquidate the instrument at par value on very short notice. This implies very little price risk, and hence credit risk.
Alternatively, they would like a guaranteed payment at some future date. (For example, liability-matching investors.)
They want as high a yield as possible.
They do not want the loan to be repaid unexpectedly, or defaulted on.
As can be seen, these desires are incompatible. Different instruments exist to fit into niches, and then the borrower and lender needs to negotiate pricing. “Financial engineering” is the never-ending attempt to bridge conflicting interests (and/or evade regulations/taxes).
Aside: “Roles of Money”
If we look at the desires of lenders, we can see a parallel with economist discussions of the “roles of money”: store of value, use in transactions, etc. This is yet another example of how myths about money damage economic theory. If we ignore all the earlier balderdash written on this subject, we can see there are two functions of “money” that matter:
use as a payments system;
short-term store of value.
The problem with economists’ stories on this topic is that mainstream economists (as well as some heterodox ones, like Austrians) use barter-like exchanges as the basis for equilibrium economic models. (In the models, everybody has a starting “endowment” of “commodities” that are then exchanged in trades, with the hope that an “equilibrium” of some sort arises where there are stable “exchange ratios” — prices. “Money” — either a commodity or a fiat currency — is just one of these “commodities.” As such, in these models “money” has to “store value” — since it needs to have value in exchange for recipients — and is allegedly used to intermediate transactions. Since the equilibrium models have all exchanges happening simultaneously, this intermediation is purely a projection by humans on the mathematical model.)
If we accept that equilibrium models are not particularly good guides for reality, we can see that the payments system in a modern industrial economy is somewhat distinct from stores of value. We can make payments without holding any particular asset, and we can hold assets that we expect to be able to trade at par on a market without any belief that this asset can be used to make payments to anyone outside that market infrastructure.
We need to step back to the not very well known concept of the “unit of account.” A unit of account is as the name suggests, a measurement of financial instruments used for accounting. In Canada, we use “Canadian dollars” to measure the size of financial instruments. Coins and bills are what people tend to think about when they hear “Canadian dollars,” but they are just bearer instruments that have a par value stamped/printed on them. They are used in retail transactions, wholesale transactions use the financial system.
We do not need to “hold money” (either government money or bank money) on our measured end-of-day balance sheet to transact.
Retail payments are heavily intermediated by credit cards. The buyer gets an account payable to the credit card issuer, the seller gets an account payable from the credit card company.
The Canadian payments system (pre-2020) operated with all the member banks (including the central bank) having a target end-of-day balance of $0. This implied that Canadian banks kept no “reserves” at the central bank. (Starting in 2020, the Bank of Canada decided to implement “Quantitative Easing,” which meant that private banks were forced to have positive settlement balances.) This meant that these extensive transactions occurred without the banks “holding money” on their end-of-day balance sheets.
With the use of overdraft protection, it is possible to run a chequing account (“checking” in American) and keep the end of day balance close to $0. (I used to do this; I just swept excess into a money market fund.)
Some American money market funds allow the holders to write a certain number of “checks.” Money market funds are not considered to be an instrument in “narrow” money aggregates (although they show up in wider aggregates).
Financial transactions can be funded via the use of margin loans or repurchase agreements (“repo”).
Banks to the Rescue
Traditional banking offers an institutional way of bridging borrower/lender mismatches.
Bank deposits are backstopped both by the central bank and legal institutions: in modern economies, bank deposits clear at par. (This was not always the case.) They thus meet the needs for immediate transaction needs, with guaranteed value.
Banks issue term deposits and bonds to meet some demand for relatively low risk future payments.
Banks liquidity management is done on a portfolio basis, and so they can offer loans that can be prepaid early.
They are supposed to invest considerable amounts into credit analysis as well as the management of defaulted borrowers (the latter of which is expensive and avoided by most bond portfolio managers). Since they can have a good view of client’s liquidity, they have informational advantages for credit analysis versus portfolio managers.
Banks Not Enough
Despite the ability of banks to bridge many of the mismatches, they cannot cover everything.
Banks liquidity and loss risk management is done on an actuarial basis (i.e., like an insurance company). The idea is that they expect to get a certain portion of their loan book impaired, and they just keep those losses at acceptable levels versus their loss-bearing capacity. However, credit losses are not exactly like insurance losses, which are normally assumed to be due to mishaps. The economic cycle can impair an entire sector of borrowers, and large corporations could represent potential losses that are too large versus bank equity.
Banks have a short life expectancy if they are borrowing at rates that are higher than they lending at. This means that they cannot provide the spread pickup over government bonds that liability-matching investors demand.
Banks generally allow borrowers to pre-pay debt, so they have limited capacity to borrow at the ultra-long maturities (e.g., 30-years) that liability-matching investors demand.
Since not all investors want to just hold bank (or central government) liabilities as assets, we need something like a bond market to allow for circular monetary flows. Traditionally, the bond market absorbed the concentrated risks associated with fixed investments (e.g., railway bonds). Countries might have domestic financial systems dominated by banks, but larger borrowers would tap into global bond markets (mainly in London). In the modern environment, securitisations allow the banks to concentrate on servicing debt, while offloading the funding operations to the financial markets.
The End of Financial History?
The financial system pays people a lot of money to invent new products, so there is always a whirlwind of new ways to gamble or evade taxes or regulations. However, from the perspective of financing investment, one might argue that the fundamental picture has been stable for some time: non-bank finance is increasingly trying to offer loans in areas that were dominated by traditional banks, and banks rely on getting risk off their balance sheets in the non-bank financial markets (“shadow banking system”).
Malcontents with a wide variety of political persuasions are unhappy with this state of affairs, arguing that fundamental changes are needed. However, unless those reform proposals take into account the mismatches being bridged for the current system — and they rarely do — those proposals will not be taken too seriously.
Believers in “crypto” might argue that we are on the edge of a revolution of finance. However, the crypto ecosystem has been great for financing gambling and ape jpegs, but not fixed investment. Given that the crypto community is oblivious to this reality after years of “development,” I see no reason for this to change.