Recent Posts

Wednesday, July 22, 2015

Book Review: The Solution is Full Reserve/100% Reserve Banking

Ralph S. Musgrave's book argues that The Solution is Full Reserve/100% Reserve Banking. However, even after reading the book, it is unclear what real-world problem full reserve banking is the solution for. Despite its length (130 pages), the reader is expected to wade through the literature elsewhere to find the details.

Book Description

The book is available at Amazon in paperback form, but it is also available as a free PDF download, with the link at Ralph's site.

Although it is technically a book, it is really just an extended multi-chapter article in format. My major complaint with the format is that Musgrave quite often just gives small excerpts of others' arguments, and the reader is expected to track down the references. This is acceptable for an article, where the emphasis is on brevity. But a monograph length work is expected to be at least somewhat self-contained, which this book is certainly not. As a result, I still have to guess at how a "Full Reserve" system would work in practice.

What Is Full Reserve Banking?

There are a number of different schemes proposed by various authors, unfortunately the reader is largely expected to read those external references to get the details. But the key unifying principle is that the banking system would cease to exist.

Two existing types of financial intermediaries are assumed to replace the present banking system.
  1. Treasury Bill Money Market Funds.  These are the only entities that are allowed access to the payments system, have a fixed par value on deposits, and clear cheques. They would Treasury bills and/or have deposits at the central bank (only) -- two instruments that are for all intents and purposes equivalent from the perspective of the private sector. (For a floating currency sovereign at least, for the euro area, the two are quite different.) These are the only entities that will be a counterparty to the central bank. Since they are 100% invested in default risk-free instruments, all deposits effectively have full deposit insurance.
  2. Debt Mutual Funds. All other lending entities have to be essentially funded by equity, or mutual fund (unit trusts) units -- which are assumed to be equity-like. To paraphrase Polonius: not a borrower and a lender be. Since debt positions are funded by equity, lenders cannot be insolvent (but see comments below).
The answer to "How do we get there?" appears straightforward: existing banks will be forced to sell all their assets to the central bank, or else to the new debt mutual funds. This would obviously create the opportunity for very large distributive effects, being a transfer of wealth to (or from) bank shareholders. Ralph Musgrave repeats Milton Friedman's assertion that the transition would be easy; however the new lending entities are unlikely to have sufficient analytical capacity, and so the transition could easily be difficult due to the constraints on lenders. (Since the existing bank funding model would no longer work, the lending capacity of banks would not easily be hived off to new entities.)

What About The Non-Bank Financial Sector?

Most of the supporters of full reserve banking appear to come from the United Kingdom, and they exploit the British practice of referring to all of the participants in the financial sector as "bankers." Conversely, in North America, people will quite often distinguish between "bankers" and "financiers." This is not just a linguistic tic; it means that many in Britain attribute to "banks" the problems during the Financial Crisis that were actually caused by the non-bank sector (including the non-bank subsidiaries of banks). In fact, traditional banking comported itself very well during the crisis. For example, Canada has 0% reserve requirements, and had only a mild economic impact from the Financial Crisis.

Reforming the banking system to deal with problems caused by the non-bank financial sector makes no sense. But the non-banking system was not forgotten, Musgrave's solution is to eliminate the non-bank financial sector (other than equity trading). In Section 2.15, he says:
But the simple solution to that is to regulate ANY ENTITY above a certain size that amounts to a bank.
Basically, all entities -- including non-financial companies and rich individuals -- would have to be strictly monitored to prevent them from acting both as a borrower and a lender. They would be forced to fit into the two categories of entities listed above. Moreover, capital controls would almost certainly have to be imposed to prevent financial activity from moving offshore.

Anything other than such draconian surveillance would allow "pseudo-banks" to form, and all that would be accomplished is a very expensive relabelling of the components of the financial system.

What Are The Advantages?

According to Musgrave, the advantages of full reserve banking are (page 13):
  1. No more bank failures. Since banks effectively cease to exist, this seems non-controversial. However, we will have failures in the non-bank financial system if regulators are unable to suppress its existence.
  2. Improved stability. Regulating financial leverage will presumably prevent purely financial crises. That said, instability in the real economy (such as the Tech Bubble) will be largely untouched. The business cycle is driven by the pro-cyclical nature of fixed investment, and this is driven by national accounting identities (the Kalecki Profit Equation), and not the form of finance. Moreover, there will be speculative waves that pass through the new credit system; since it is market-based, asset bubbles would be more easily financed.
  3. Bank subsidies vanish. This claim is repeated throughout the book, and is presumably the most important point.
  4. Depositor/Investors choose. Retail deposits are not used to finance risky assets. Most regulators who understand the banking system would view this as a defect, not an advantage, as discussed below.
  5. No limit to deposit guarantees. Outside of the euro area, who cares?
He does mention in other passages the idea that resources would be freed up as we would need less bank supervisors. Given that all commercial activity would have to be tightly regulated as I discussed above, I seriously doubt that there would be any savings on bureaucrats' salaries.

The claim that "subsidies reduce GDP" is repeated throughout the book. The statement is unproven and dubious.

The subsidy that is provided by access to lender-of-last-resort operations is a structural feature of the economy, and cannot be measured in terms of a financial cost. Since any realistic assessment of how capitalist economies function shows that investment and financial activity is pro-cyclical -- regardless of the form of financing -- lender-of-last-resort operations will always occur if the authorities aim to head of a depression.

Also, one can draw the analogy with another structural subsidy in the transport sector. Trucking companies are subsidised versus railroads, as they are provided a road network by the state, whereas railroads have to maintain their own track. It seems likely that a country that prevented trucks from using public-owned roads would have lower growth than those following current practices. In any event, since the structure of the economy would be radically different, it is extremely difficult to isolate the growth impact of the "subsidy."

Pros And Cons Discussed Within The Book

The book has a huge range of arguments both for and against a Full Reserve banking system. There are 45 broad complaints that Musgrave identifies, and which he addresses. Unfortunately, they are introduced by cryptic partial quotes, and so the reader would have to track down the references to see whether the author is accurately capturing the critiques.

Given the huge number of points, I cannot hope to even begin to address them. I would note that I think that most of the critiques (as Musgrave presents them) are in fact weak. That said, I still think a Full Reserve system would fail; my criticisms are not properly captured by the 45 he lists.

Would It Work?

Although I believe such a system will not work, that statement should be qualified. Once the difficult transition phase was completed, the economy would be able to function with a simplified financial system. It would be more difficult for smaller firms to raise capital (as discussed below), and so there would be bias towards larger firms. In my view, this would be unwelcome, as the economy would tend to become too rigid. That said, it is unclear whether this would obviously damage economic performance.

However, the system would be under continuous pressure to evolve entities that are essentially banks, even if they are labelled differently.
  1. Citizens and businesses will have to obey the draconian financial regulations, and not find ways to arbitrage them. Arbitrage and economic forces would otherwise cause the rise of a new system of entities that are for all intents and purposes banks.
  2. Politicians would have to follow the strict rule of watching their economies spiral down into depression instead of engaging in lender-of-last-resort operations. Any such lender-of-last-resort operations would christen the recipients as becoming the core of a new "banking" system.

Points Of Failure

The Treasury Bill fund based payment system would obviously function. The only question is whether the stock of government liabilities would be large enough to support the demand for base money (or bills). That said, the government could monetise random assets to create base money. (This would only have the cost of redistributing wealth to the sellers of said assets. This is a subsidy, which we are told reduces GDP.) An additional problem is the question why anyone other than an oligopoly would want to offer such a product to the public. It would be tightly regulated, require large investments in the payment infrastructure, and profit margins would be negligible. Banks make such investments in the payment infrastructure because deposits are a cheap source of funding, but that incentive would no longer exist.

The more serious problems will occur in the lending entities. They would have a hard time extending finance to smaller firms.

The book repeatedly argues from authority, citing economists such as Friedman, Fisher, Kotlikoff, and Cochrane, that there will be no problems. This sounds impressive until we realise that these economists represent a fringe that believes that market-based solutions solve all economic problems. As a Minsky-ite economist, I have serious doubts that this is the case. 

The economists he cites tend to assume that markets are perfect, continuously liquid, self-regulating and could never misprice any assets. Correspondingly, the analysis skips over the real world issue that a great deal of private sector debt is illiquid and there is effectively no price for it. Loans to smaller firms is the most dominant form of such illiquid assets. The types of private sector debts that can plausibly be traded in markets have been securitised or issued in bond format a long time ago. 

If there is no market, there is no way of assigning a price to a loan (at least until it is paid back or written off). Banks have developed practices over centuries to deal with this problem. The best practice is to fund such illiquid assets with "sticky" retail deposits, and not wholesale funding. The entire objective of full reserve banking is to prevent this optimal funding structure from occurring. On paper, long-term investors should be able to hold illiquid securities. But in practical terms, most investors are chasing after short-term performance metrics, and have only limited capacity to hold illiquid assets.

The inability to price assets makes the repeated appeal to the Miller-Modigliani theorem irrelevant. Since we cannot price the asset side of the bank properly, we cannot say too much about the relative pricing of the liability side.

The inability to price illiquid assets means that the "debt mutual funds" discussed by Musgrave will not work.
  • If the funds are "open" -- unit holders can redeem units at calculated NAV -- they are still vulnerable to runs. The first people to redeem will get the proceeds of the liquidation of liquid assets, and the remainder will be stuck with units that are backed by illiquid garbage that cannot be valued. And this is not a theoretical argument; this happens to hedge funds during crises. The run on funding units would cause a collapse in real activity, just as we saw in the Financial Crisis.
  • If the fund units cannot be redeemed, there is no way for unit holders to discipline poor managers of funds. Moreover, the units would have to be traded in the secondary market, at prices that may bear no resemblance to their NAV. It is hard to see why any competent individual would invest in a non-redeemable fixed income product that has a market value that is completely detached from fundamental valuations.
Investors who want to own fixed income products with transparent pricing characteristics would be forced to lend directly to entities issuing bond-like products. Since one can structure bonds to look like deposits (such as the failed Auction Rate Securities in the United States), we could just end up with bank-like institutions.

Meanwhile, it is unclear that mutual funds or similar market-value oriented investors are able to allocate credit in an efficient fashion. Although some investors may be disciplined, the bulk of investing activity can be characterised as trend-following. A flow of capital into an area will reduce spreads, raising the market value of holdings. This will validate the earlier flood of investment, and raise the investment capacity of those who want to keep pouring capital into the sector. The financial sector accentuates the booms and busts that are a natural part of capitalism. Admittedly, banks follow trends themselves, but those bubbles are quite often associated with regulatory failures (such as the Savings & Loan crisis in the United States).

An additional problem with this scheme is that it is essentially impossible to offer credit lines. (I see no way of offering credit lines in an economical fashion while being consistent with the objectives of Full Reserve banking.) Commercial paper and bond issuance are now backed with credit lines to prevent "runs" on issuers. Pulling those credit lines will mean that "runs" are the rational response to any perceived illiquidity risk. Credit lines are not a minor piece of financial trivia; they are a critical component of the private sectors' coping mechanism to deal with liquidity risk. (Lender-of-last-resort operations are essentially just credit lines offered by the central bank.) Fundamentalist efficient market analysis that relies on all securities always having a well-defined price (which can be borrowed against in perfectly liquid lending markets) ignores the importance of credit lines, and Musgrave's analysis inherits that flaw.

From a bigger picture perspective, mismatches between the desired liability and asset structures of the private sector implies the need for "maturity transformation." Banks have been designed to supply that maturity transformation. Abolishing banks will just create an economic force that will push other entities to take over that role. The form does not matter, only function.

Connection With MMT?

Both Full Reserve Banking and Modern Monetary Theory (MMT) are well represented upon the internet. There are some superficial similarities between them, given the emphasis on central bank reserves. This similarity appears misleading, however.

The MMT view is that the central bank and the Treasury should be consolidated, and that the government should stop issuing bonds. This would mean that government liabilities would only consist of non-interest bearing deposits at the central bank, as well as government-issued notes and coin. Since there is no distinction between the Treasury and the central bank, there is no distinction between fiscal and monetary policy; they are integrated.

Full Reserve Banking proponents believe that government liabilities would largely consist of deposits at the central bank ("reserves"). They seem to believe that a third agency will be able stimulate the economy by adjusting the amount of those reserves. This is seen as some sort of independent agency, like the Monetary Policy Committee of a central bank.

This committee would be redundant. The amount of government liabilities is determined by fiscal policy, and the mix between reserves versus bonds/bills is determined by private sector preferences and central bank policy. Correspondingly, the new committee's decisions about the monetary base are meaningless. As a result, the switch to full reserves would not create any new policy space.

More generally, I have a hard time seeing is any advantage of a political alliance between MMT and "Full Reserve" banking proponents. (Please note that I do not speak for anyone other than myself; others in the MMT camp may look at things differently.)

Modern Monetary Theory proposes incremental changes to how governments finance themselves, as well as a reform of the Welfare State (a Job Guarantee). These are reforms that could plausibly be supported from across the political spectrum, and have at least a chance of improving economic performance.

The Full Reserve Banking reforms consist of regulating most of the financial system out of existence, and generate no corresponding tangible benefits (other than satisfying those who think banking is unfair). Why use spectacular amounts of political capital on a programme that accomplishes almost nothing?

Concluding Remarks

There are obvious defects to the current financial system. Simplification is probably the best way forward. Reverting back to the pre-1980s financial system, which rigidly separated the types of financial firms, would probably be the optimal outcome. But at the same, the reforms that were put into place since the 1980s did not occur just because of "neoliberalism." Private sector balance sheets have grown, and the non-bank financial system has grown in response. Bank regulations had to adapt to the rise of the shadow banks. Unless there is a plausible means to reduce private sector balance sheets, reforms will just change details of the form of the system, but not the substance.

As a result, I can sympathise with some of Ralph Musgrave's complaints about the structure of the banking system. But at the same time, I do not see clearly how the proposed system could cope with the economic forces driving the financial system.

(c) Brian Romanchuk 2015


  1. I think banks must be implicated in the crisis along with the nonbank financial firms for the following reason. Large banks increased returned on equity prior to the crisis by adopting the originate to distribute model where the bank would originate the loan, package it in asset backed securities pool, and sell those ABS instruments to nonbanks with credit insurance. In part, if banks had not issued such credit insurance, then banks would not have been forced to borrow from Fed in the crisis to repurchase many loans in the nonbank ABS portfolios.

    One must carefully study the interaction between banks and nonbanks to understand how the system becomes corrupted by a spike in defaults and a liquidity crisis. One theory of why banks were dragged into the crisis is that banks effectively provide credit default insurance to nonbanks and Fed provides credit insurance to banks as Lender of Last Resort. Experts who thought banks were ring-fenced or isolated from the bad credit in nonbank portfolios were caught by surprise when banks were not isolated at all during the crisis.

    1. Agreed, but I believe that this is consistent with what I wrote. I did not have space to write a history of the crisis, nor a comprehensive explanation of how I believe the financial system should be reformed. However, I do not believe that bank use of CDS was the source of the crisis. For the banks, the problem was their off-balance non-bank funding subsidiaries. The problematic role of CDS was the demand for securitised loans that was created by the synthetic CDO market.

      The "banks" in the real world are corporations that consist of traditional banking subsidiaries and non-bank financial subsidiaries. Such a merger would have not been allowed under the pre-1980s rules in most countries. In Canada, this was referred to as the "pillars" of the financial system, and the pillars were firewalled. Over recent decades, regulators allowed the merger of bank non-bank functions, which appears to have been a mistake. Eliminating the safest part of the financial system in response to a botched deregulation makes little sense.

      If countries moved to the Full Reserve model, lending would have to be market-based, that is, ready to be securitised. There is no guarantee that the entities that have the operational capacity to lend would coincide with funding ability; they would have to lend, and sell most or if not all of their loans to the entities which have the capital. In other words, all lending would largely have to be on the inherently dysfunctional originate-to-distribute model.

    2. I think Warren Mosler advocates strict regulation of bank assets (risk management) and fiat government insurance (insured deposits & lender of last resort) for bank liabilities.

      I don't think credit default insurance, deposit insurance, or other risk transfer schemes can reduce systemic risk in the private sector. Aggressive executives have an incentive to increase risk, increase bonus & compensation, and transfer risk to others, including bank equity owners, if the bank/firm must be resolved. These moral hazard incentives are called Agency Problems.

      Since some default rate is statistically certain in a complex financial system this means some loans will be written-down, reducing the growth rate of equity in banks and nonbank financial firms, and when the write-down is large, there is a net loss of equity. Government bailouts occur because investors will not rollover deposits or place new paid-in equity in a firm where the assets may include a large future write-down, and the real economy will suffer from a lack of credit formation if investors must take the write-down.

      I think billionaires can afford the write-down but how do you force them to accept the loss when their holdings are mixed with ordinary pensions and savings of the less wealthy retirement community? So I refer to the financial system in crisis as The Gordian Knot.

      William F. Hummel has a proposal for a National Depository System:

      in which both banks and nonbanks hold transaction accounts at the central bank, so there are no bank reserves, just deposits at the central bank. Banks would have special regulations and special borrowing privileges with the central bank that are not available to other nonbank units.


Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.