The Money Problem: A Discussion Of Financial-Regulation Choices In The Form Of A Book Review

Summary

• The Money Problem by Morgan Ricks clarifies many of the thorniest issues about financial regulation.
• Monetary design and preventing panics are at the heart of the Ricks analysis. His new "money" is an elegant construct.
• Even if one does not agree with the Ricks solution (I think I don't), his contribution to thinking about financial regulation is enormous.
• I have my own, less radical, idea about how to prevent panics, and I offer an outline here.

A couple of weeks ago, I asked Bob Litan of the Council on Foreign Relations, one of our best economic thinkers, whether he thought public policy ought to try to prevent debt-fueled financial bubbles. He responded that he was involved in reading Morgan Ricks' new book, The Money Problem: Rethinking Financial Regulation. It is the best book on finance in decades, he said. You have to read it. I admitted I had not read it, despite very interesting reviews, because I was annoyed at what University of Chicago Press charges for it, even as an eBook. But I swallowed my pride and bought it.

I will tell you that Bob was right. It is the best book on finance in decades, even though there have been many good ones. The reason it is the best is not because I agree with its conclusions (I think I am not persuaded), but because it is so clearly and sensibly written. It challenges a great deal of orthodoxy very straightforwardly.

If you are not a financial regulation junkie, you may not want to tackle The Money Problem, but if you do, you will be rewarded by excellent analysis and clear writing.

Financial panic as the source of deep recessions

The book's main point is that financial panic was the phenomenon that caused the greatest damage in The Great Recession and that financial panics almost always are the phenomenon that causes the greatest damage. In saying this, Ricks is echoing Gary Gorton in his excellent book Misunderstanding Financial Crises, and Ricks gives Gorton full credit.

Because it is panics that cause the greatest damage, Ricks reasons, it is panics, rather than debt-fueled bubbles or excessive leverage or the like, that financial regulation should seek to prevent. Ricks discusses the many ways in which financial regulation has sought to deal with the panic problem. He concludes that they have not been successful and that, although the post Dodd-Frank architecture may be better than what came before it, it still is not adequate.

Financial panic is a monetary design problem

Ricks' key observation is that financial panic is a monetary design problem. Money-like obligations, by which he means short-term obligations issued by an entity to a large number of obligees, are inherently runnable. He uses game theory to show that holders of such obligations, whether we call them deposits or cash equivalents or money market instruments or commercial paper or anything else, do not do a great deal of investigation about the financial condition of the issuer. They trust the issuer-until they do not. And they do not trust the issuer when they perceive that other holders of similar paper do not trust the issuer, at which time, all the holders attempt to redeem the obligations at the same time, which leads naturally to panic. In game theory, there are two equilibria-no one worries or everyone panics.

This is true, Ricks argues, regardless of what we call the obligor. It is true of banks and non-banks alike. Any issuer of short-term obligations is subject to the same danger.

The Ricks monetary design solution

Ricks' solution is two fold: One, prohibit any entity except a bank (as defined) from issuing short-term (less than one-year) claims (there are exceptions), and two, make banks, as defined, the sole issuers of money, through the creation of money by using new deposits to make loans or buy securities. These new banks will look a lot like current banks except that all their obligations will be deemed to be "money" issued by the sovereign and therefore always money-good, regardless of whether the bank fails. In effect, it is 100% deposit insurance. But even more to the point, the government directly controls how much money the banking system as a whole is allowed to create. (That amount can be allocated among banks by a cap and trade system.)

The regulation of these new banks is a little complex because it involves measures to protect the sovereign from losing too much in a bank failure, charges to the banks for the privilege of creating money (seigniorage), and adjusting the seigniorage charge up or down in order to manage the money supply. But regardless of whether those decisions are soundly based, the banks' obligations will not be runnable, since they are all "money" created under the auspices of the government.

The tradeoffs between protecting the sovereign from losses due to bank failures and optimizing the money supply would be delicate: "the risk constraints need to be permissive enough to accommodate the desired money supply." (Kindle Locations 475-476)

The banks in Ricks' system would hold no derivatives (except for hedging), which I agree is a good idea in any event. Whether the banks could have affiliations with securities firms is left somewhat up in the air. Ricks discusses the pros and cons of securities firm affiliations, and he notes that if affiliations are permitted, then restrictions on transactions between affiliates have to be strictly enforced. It seems to me that the affiliates question is not very different from the same question regarding the current form of banking.

Stated as baldly as I have stated it above, the proposal sounds quite radical-perhaps even outlandish. But Ricks supports his ideas well, and along the way, quite demolishes many ideas (myths?) about financial regulation and the nature of money.

Why banks (and others) need to be regulated

Among Ricks' contributions is a clear explanation as to why banks need to have limited investment powers and comprehensive capital requirements. The answer lies in the nature of the license and protections the government has granted, regardless of whether we are discussing Ricks' new type of bank or a bank like those currently in existence. The restrictions and capital requirements are needed in order to prevent the license and protections from being taken advantage of unduly, which would be to the detriment of both the public and the government. Once the licenses are required, there can be no question of a free market.

Yet, as Ricks points out, under current law, non-licensees that do not have the benefit of the Fed's safety net, also can issue money-like claims, and they are not subject to investment restrictions or capital requirements. That is where Ricks sees the biggest problem. And in a sense, I agree, as I will discuss at the end of this review.

Are panics really the cause of deep recessions?

I am not convinced that panics are the principal cause of deep recessions. I think Ricks is weaker on that point than on many others. I remain of the view that once a debt-fueled bubble has inflated, it is bound to burst and that when it bursts a significant recession will ensue. That a panic also may ensue certainly does make the recession deeper.

Ricks maintains that it is the psychic damage that the panic causes that makes it harder for the economy to climb out of the recession-which is what caused the slow recovery after the Great Recession. I am not persuaded that that is the primary cause of the slow recovery. My analysis begins with the poor state of the economy ex-the-bubble in the period 2003-2006 and continues with the reversal of the home equity extraction that supported the bubble economy. And for those reasons, although I think preventing panics is important (and Ricks-as well as Gorton before him-has added to my understanding of that importance), I also think it is important to prevent debt-fueled bubbles.

Debt-fueled bubbles

I should note that Ricks does not say preventing debt-fueled bubbles is not a good idea. He expressly disclaims any such implication of his work. But he does maintain that any such policies should be seen as secondary and, even more, that his mechanism to prevent panics actually also may tend to prevent debt-fueled bubbles. And I agree with him that his mechanism would tend to prevent debt-fueled bubbles, especially if the kind of international cooperation that he envisions could be obtained. (I appreciate that Ricks has made the distinction that I have made for a long time between bubbles fueled by debt and other bubbles-e.g., the housing bubble-dangerous-versus the Dotcom bubble-not so dangerous.)

The following schematic illustrates the kind of contribution to clarity that Ricks has made. It shows three ways that we could think about the relationship between the collapse of a bubble, a panic, and a severe recession. Although I do not agree with Ricks' conclusions, I found his schematic (Figure 4.2) contributed to my thinking about the subject.

Monetary policy is more important than preventing panics

Because I think there are less radical solutions to the problems that Ricks illuminates very clearly, I come away wondering whether the style of monetary policy that the Ricks' new form of banking would not only enable but require would be better than the current style. Ricks does not really address this issue directly, which is in way a shortcoming and in a way a mercy. The question of how to conduct monetary policy is separate from the question of how to prevent panics-and it is, in my opinion, far more difficult and diffuse.

The Ricks solution to the runnability problem would make it far easier to define the money supply-and therefore, presumably, to manage it. But it would place far more power in the hands of the monetary authority, with little wiggle room for private actors to soften the monetary authority's decisions. Mistakes would be magnified.

I cannot judge whether the Ricks solution would make monetary policy better. But since monetary policy is at most times more important than preventing runnability, I am reluctant to embrace the proposed solution without understanding more about how monetary policy would be conducted.

A less radical approach to preventing panics

My own current tentative suggestion to redress the runnability/panic problem is two-fold. First, I would address the runnability of specific types of institutions that Ricks calls shadow banks. (I like his definition of "shadow bank": It is an entity that issues short-term claims that are like money. The definition is descriptive, not pejorative.) Ricks has an excellent (though hard to read) figure that shows the various types of money claims outstanding not issued or backed by the government (Figure 1.2). Institutional MMF shares, Eurodollars, asset-backed commercial paper (OTCPK:ABCP), and repurchase agreements played significant roles in the panic of 2008. Ricks would ban all of them (except MMF shares backed by government securities).

I would argue that MMF shares that traded a constant buck was a regulatory mistake that now has mostly been corrected by the SEC. I would argue that ABCP was a regulatory mistake made by the Fed and its sister banking agencies that also now has been corrected, and there is almost no ABCP outstanding. Eurodollars and repos are different stories.

The Eurodollar question can be addressed only by an international agreement. Ricks is optimistic that such an agreement might be reached through the BIS, but that is not a foregone conclusion, and a lot of toothpaste could escape the tube by that route. But if MMFs were prohibited from buying paper of foreign banks, then maybe the issue would not be so large.

Repo as a major issue for financial (or monetary) design

The repo issue is a serious one. It was a run on repo that bankrupted Bear Stearns, and Ricks draws an important conclusion from that experience-a conclusion that reinforces his exploration of game theory: A repo lender will run even though fully secured. The repo lender does not want the collateral; it wants the cash. The collateral may be "good" all day. It may be government securities of short maturities. But the lender does not want the nuisance of the securities-and there is no upside to it. Therefore why not run at the first sign of trouble?

Repos have been a potential problem for many decades. They featured in the bankruptcy of Drysdale Securities in 1982, when few people outside the securities business had even heard of repo (except for auto repossessions) and the Lombard-Wall bankruptcy shortly thereafter. See here for a description of the history of repos and the role of the those bankruptcies. The key question that the Drysdale and Lombard-Wall cases raised was whether the repo is one transaction (a secured loan) or two (a buy and a sell). If one transaction, then it would be subject to a stay in bankruptcy; if two, then it would not be subject to a stay. The courts eventually ruled in Lombard-Wall that it was two transactions, and Congress has reinforced that conclusion. But if Congress reversed that conclusion, the repo market probably would disappear.

Would that disappearance be a good thing? I do not know. Ricks would make it disappear by prohibiting the short-term borrowing. An alternative, however, and one that Ricks might embrace, would be for Congress to make repos subject to the stay in bankruptcy and for the Fed to offer to borrow using its balance sheet securities as collateral (reverse repos-or RRPs). The Fed now does that for some kinds of institutions. Why should that opportunity not be open to all? Since the Fed cannot fail, Fed borrowings through RRPs would not be runnable.

What would happen to the hedge funds and securities firms that fund themselves through repo transactions? They probably would find their funding costs would increase. But note that the bankruptcy interpretation of repo as two transactions is a fictional construct designed to get around the usual stay in bankruptcy. Someone in the market pays for that.

Thus I think we can see alternative ways to deal with shadow banking, rather than prohibiting the issuance of short-term claims. But Ricks has done a service by providing a better definition of shadow banking and by pinpointing the danger.

Solving the bank run problem

The bank run problem is even more straightforward to deal with than the shadow bank problem. Although Ricks devotes considerable space to discussing the holding company resolution provisions of Dodd-Frank (and finds the single point of entry strategy a useful one), he does not discuss what I see as the biggest benefit of Dodd-Frank-the annual stress testing of banks and their holding companies. Such annual stress tests (and the publicity given to their results) have revolutionized the regulatory capital regime by making it forward-looking rather than historical. The severe scenario tests the bank's capital in exigent circumstances, and thus it leaps over the typical dilemma of capital regulation, which is that leverage ratios give an incentive to take greater risk and risk-based ratios tend to lead to gaming the system. With stress-tested capital, the accuracy of the risk-based system no longer matters much, and the tendency of the leverage ratio to promote risk-taking is pulled back by the adverse results of the strategy in the severe scenario. I am surprised that Ricks did not discuss this important effect, especially since he recognizes that his new type of bank still would need capital regulation in order to protect the sovereign from excessive risk-taking.

Like Ricks, I would make all of the bank's liabilities guaranteed (in effect). But I would do so by having the Fed declare that any bank that had a specified level of capital after application of the severe scenario would be deemed "solvent" for the next year, and therefore would be eligible for loans from the Fed for up to the full amount of its assets for that period. Banks would, naturally, manage themselves to qualify for this privilege, since it would reduce their cost of funding (which probably would offset whatever additional capital costs they incurred or whatever risky business they turned away).

If a bank fell out of compliance, it would be only one bank, and although that might cause a run on its uninsured liabilities, it would not cause a systemic run.

The Fed's effective guarantee would not, of course, apply either to the bank's holding company or to its sister companies, such as securities affiliates. Those still could fail. But the bank itself would be protected. Without the ability to fund itself through repos, it seems likely that the securities affiliate would have issued less runnable paper.

My solution is not as conceptually elegant as Ricks' monetary redesign, but it would be far less radical.

How should policy deal with the securities business?

In all the analyses, the securities business seems to be the most problematic. It was the trigger for the panic in September 2008, it was the source of the largest frauds that pumped up the housing market, it was the source of the illegal price fixing of LIBOR, EURIBOR etc. What should one say about the need to bring it under control? It is at the same time the great market that funds American business better than any other funding mechanism in the world and the source of great angst and damage in exigent economic times. It is the "Wall Street" that the left loves to hate.

One readily can see how to bring traditional banking under better control. The Ricks mechanism or my mechanism could do that. But how should the securities business be brought under control without destroying its ability to fund business better than any known alternative?

I will leave that question hanging and will hope to address it in a future article.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.