Summary. Conventional banks combine two activities, lending and accepting deposits. That combination is essentially fraudulent. The harmful effects of that fraud are remedied to some extent by deposit insurance and lender of last resort. However no public purpose is served by those forms of state assistance to money lenders any more than offering that form of assistance to non-bank corporations would serve public purpose or bring social benefits. Ergo those two activities, lending and accepting deposits, should be separated.
Private banks have always liked to merge the above two activities, that is, have the same institutions (private banks) perform both functions. Reason is that that “merge” enables private banks to engage in “borrow short and lend long” (i.e. maturity transformation). And that’s profitable.
Unfortunately the above merge is risky. As Adam Levitin (professor of banking law) put it in the first sentence of the abstract of this paper: “Banking is based on two fundamentally irreconcilable functions: safekeeping of deposits and re-lending of deposits.”
The irreconcilability or fraud consists of the fact that banks promise depositors they’ll get their money back at the same time as putting that money at risk by lending it on. That tactic is bound to fail at some point, and when it does depositors are fleeced big time, as happened in the 1930s. Or if government run deposit insurance is in place, which it wasn’t in the 1930s, then it’s still someone other than the miscreant bank who pays, which is nice for the miscreant bank and those funding it. A second form of protection offered by governments / central banks for the above merged system is lender of last resort (LLR).
So there are two possible bank systems. First, a system under which entities which accept deposits lend on those deposits. In view of the risks involved there, taxpayers have to stand behind private banks. Second, there is a system which solves the above problems by having the state only guarantee money lodged with the state. As to lending, that is funded by people who specifically choose to have their money loaned on, and who buy shares or similar stakes in money lenders, and those shares are of course not the same thing as deposits.
A “merge” system protected by deposit insurance and LLR might seem to make sense. In fact it’s a system riddled with flaws, as follows and numbered.
1. The system involves moral hazard: the temptation to take excess risk, keep profits when that works and send the bill to the insurer when it doesn’t. Indeed, it was precisely that that contributed to a significant extent to the crisis in 2007/8.
2. One common excuse for having the state insure money lenders and those who fund them (cited for example by the UK’s Vickers commission) is that doing so encourages loans and investment. Unfortunately exactly the same argument applies to every other industry and those funding them. That is, the latter “encourages investment” argument could equally well be applied to all stock exchange quoted shares and corporate bonds.
Of course there’s no harm in letting those who buy stakes in money lenders or any other industry insure themselves against loss, and a certain amount of that type of insurance takes place. However, to justify the state getting involved in that insurance, there must be some very clear social benefit or public purpose served.
3. The idea that governments or regulators are actually able to work out likely risks and charge an appropriate insurance premium is a joke in view of 2007 crisis.
4. Banks are bound to lobby politicians for an unrealistically low insurance premium.
5. When there are bankruptcies in any industry, it is positively HEALTHY for those who funded the industry to lose out. Bankruptcies tend to indicate the industry is too large, and that resources should be diverted to other activities. Thus far from any public purpose being served by deposit insurance, harm is actually done: that is, the effect is to recompense those who fund money lending, which encourages them to engage in more money lending, or “debt creation”.
I.e. when the FDIC reimburses depositors in a failed bank, depositors put the money in another bank. Failing to deal with malinvestments is a misallocation of resources: it reduces GDP.
It could of course be argued that if state sponsored deposit insurance does harm or reduces GDP, then the same applies to the private insurance (mentioned above) that is sometimes taken out for stakes in industries other than banking. That argument is not totally invalid: that is, possibly the latter form of insurance should be banned.
On the other hand, private insurance is not as “sure” as state sponsored insurance in that private insurers can go bust. I.e. private insurance is in a sense not insurance. Also it’s a generally accepted principle that people should be allowed to do whatever they want (e.g. insure their own legs) unless some very clear harm comes what they do.
So it’s debatable as to whether private insurance of investments like stock exchange quoted shares should be allowed or not.
6. Where money lenders are funded by shares, the entire system is more robust. As the former governor of the Bank of England, Mervyn King put it: “..we saw in 1987 and again in the early 2000s, that a sharp fall in equity values did not cause the same damage as did the banking crisis. Equity markets provide a natural safety valve, and when they suffer sharp falls, economic policy can respond. But when the banking system failed in September 2008, not even massive injections of both liquidity and capital by the state could prevent a devastating collapse of confidence and output around the world.”
In short, banks funded by equity are more resilient than where they are funded by debt (e.g. deposits).
7. Bank regulators the world over have expressed approval of the principle that banks should be treated like any other industry: i.e. that ideally governments should openly declare that no form of assistance will be offered to banks in trouble. Of course what regulators and politicians say in private is very different. As John Kay put it, "With crass hypocrisy, political leaders have set their public faces against future bank rescues while their operatives have reassured markets that they do not mean what they say."
But assuming a government does openly and explicitly declare that there will be no assistance for failing money lenders, then the stakes in those lenders (deposits in particular) ipso facto become shares or bonds or something of the sort: so those stakeholders stand to lose money, whereas depositors are guaranteed not to lose out.
8. Another point which casts doubt on any idea that any public purpose is served by having the state insure those with stakes in money lenders (or any other industry) is the fact that there is a nonsense at the root of that type of insurance, as follows.
The only reason that those who fund money lenders (or any moderately risky industry) get a larger return than is obtainable from near risk free loans, is that more risk is involved. Now if the risk consists for example of a one in X chance of losing all your money in any one year, then the appropriate annual insurance premium would be 1/X of the sum insured. But that wipes out the profit derived from taking the extra risk (never mind the fact that the insurer will want some sort of profit on turnover and capital employed).
In short, insuring those who fund money lenders is as daft as insuring your own legs: to repeat, it probably doesn’t do any harm to let people take out daft forms of insurance, but there certainly isn't any public purpose or social benefit to be had from insuring those who fund money lenders or other industries.
9. In addition to deposit insurance, another form of assistance for private banks is LLR. The rate charged for LLR loans are supposed to be the “penalty” rates suggested by Walter Bagehot. In practice, and thanks to political pressures and bribes payed by bankers to politicians, those rates are nearer zero than “penalty”: a form of subsidy for banks.
10. Contrary to popular belief, Bagehot did not advocate LLR: in the final chapter of his book “Lombard Street” he simply said he thought LLR was so entrenched that it would be very difficult to remove.
11. It is often argued that if the PROPORTION of funding for money lenders that comes from shares is sufficiently high (i.e. if bank capital ratios are high enough), that solves the problem: that is, the chance of depositors losing out can be reduced to a vanishing small level. Thus there is no need for 100% of those who fund money lending to be exposed to risk, i.e. depositors can still to some extent fund money lending.
A problem with that argument is that if those depositors are in fact totally safe at say a 30% capital ratio, then they’d also be totally safe of the rules of the game were changed and their stakes in money lenders became say preference shares. And in that case, those shareholders would charge the same for funding the bank as those former depositors (because the charge made by funders is related to risk, and the risk in both cases is the same). Thus if bank capital ratios are raised to whatever level makes banks totally safe (say 30%), then the cost of funding banks will not rise any further if the ratio is raised to 100%.
Arguments against separation.
The main argument put against separating lending from deposit accepting (put for example by the UK’s Independent Commission on Banking) is that the result would be large amounts of money not being used: effectively, if you like, money sitting in metaphorical safe deposit boxes. The result, allegedly, would be a fall in demand and rise in unemployment.
The answer to that is that the deflationary effect of the separation can easily be countered by standard stimulatory measures, e.g. simply having the state print money and spend it (and/or cut taxes). Thus there is no reason for separation to increase unemployment.
Another poor argument put against separation is that it would raise interest rates which allegedly would reduce investment and cut economic growth.
The answer to that is that as is widely accepted in economics, GDP is maximised where there is an absence of subsidies, except where there is a clear social case for a subsidy, as is doubtless the case with for example kid’s education. Or in more general terms, there is a case for subsidies and taxes where those subsidies put right an instance of market failure. (Taxes are imposed for example on alcohol because it is thought, rightly no doubt, that social harm derives from excess alcohol consumption.)
However, the onus is on those advocating subsidies and taxes to prove the existence of social benefits, market failure and so on. To repeat, the general and widely accepted rule is that GDP is maximised where market prices prevail, i.e. where there is an absence of subsidies or taxes.
In short the free market rate of interest is probably somewhat higher than the currently prevailing rate, thus moving to that higher rate ought to increase GDP, not reduce it, as claimed by the ICB and others.