Toward a robust dealer system of first resort

Goldman Sachs takes “A Look at Liquidity”, and tells us what they see.  Suffice it to say that different people see different things, depending on their vantage point, like the proverbial blind men touching the elephant.  Let’s see if we can construct a picture of the animal as a whole from the snapshots provided.

What seems clear is that the bank dealer system of the past is now gone (notwithstanding last-ditch pleading to bring it back).  Banks are largely out of that business, guided by new regulations (specifically the “non-risk-based leverage and liquidity rules”), but also motivated by their own experience with the crisis.  Banks do not want to be in the position of requiring emergency support from the Fed any more than the Fed wants to be in the position of providing that emergency support.

Ever since the crisis, central banks have been standing in for the pre-crisis bank dealer system, flooding the system with funding liquidity.  World-wide QE has essentially bought time for a new more robust dealer system to begin rising from the ashes of the old.  However, at the moment that new system is far from complete, even as central banks (led by the Fed) are signalling that they will not be around forever.  In normal times the central banks supports the market; only in crisis times does it become the market.  What will the new normal times look like?

Steve Strongin talks about how the Fed might respond to the next crisis:  “the Fed might have to buy the distressed assets directly and/or other parts of the government might have to step in” (p. 5).  That is of course how the Fed responded to the last crisis, as I have myself recounted in my book New Lombard Street.  But the necessity for that response shows exactly the inadequacy of the old dealer system–it was not a robust first resort system.  That’s why we have junked the old system.  The question is whether we can rely on the emerging new system to be more robust.

There is a lot of hype about electronic exchanges, and also Exchange Traded Funds, and some of the hype is warranted.  Yes, to the extent that we can make it easier for buyers and sellers to find each other and do business directly, we can do without the now-missing dealer intermediary.  In effect, all such measures work by making the broker function more efficient, which is fine if markets are balanced.  “But the largest problems are likely to arise when markets are not balanced and under significant net selling pressure…” (p. 5)

Just so, consider Blackrock’s Richie Prager on the ETF:  “If supply exceeds demand, the ETF wrapper is unzipped and underlying securities are delivered to the market, thereby reducing the number of ETF shares.  Given these mechanics, even in an extreme scenario, you could unzip the wrapper and be left owning the bonds” (p. 9).  Golly!  The asset manager who wants to sell the ETF is supposed to be calmed by the thought that he now owns the underlying illiquid bonds?  I think not.  Long before that happens, the sponsor of the ETF will be pressured to support its creation by absorbing the net flow on its own balance sheet, which is to say by acting as a dealer not a broker.

The problem is, as Strongin points out, that supporting markets in this way requires the ability to expand your balance sheet on both sides, buying the unwanted assets and funding that purchase with borrowed money.  The ability of banks to do that on their own account is now severely limited.  And it is not even clear that they can readily provide the funding liquidity for someone else, i.e. Blackrock, to do it.  Without Strongin’s “dynamic balance sheet expansion”, the result will be Prager’s “discontinuous pricing and greater volatility” (p. 9).

Why should we care?

One reason is efficient pricing, as Himmelberg and Bartlett point out.  “Without improvements in single-name liquidity, the feasibility of active portfolio management and hence market efficient will necessarily suffer” (p. 7).  For an economist, this argument has an almost irresistible appeal.  We habitually imagine a world of perfect liquidity and efficient pricing, in which only tiny fluctuation in price is required to bring forth demand or supply as needed to clear markets.  Unfortunately, in the real world, the value investor makes the outside spread not the inside spread, buying only when it is a bargain, i.e. when price is clearly less than value.  Value investors are liquidity providers of last resort, but if we rely on them for market liquidity, then prices will be discontinuous, jumping to the outside spread whenever there is imbalance in order flow.

The more important reason we care is that we are worried that discontinuous pricing and greater volatility can get out of hand, and if they do then the Fed is going to have to once again serve as dealer of last resort.  Let me repeat what I said above:  No one wants this, not the banks and not the regulators.  But also no one has a very good idea how to avoid it.  We are in uncharted waters, inventing new things.

At this juncture it would be helpful, I think, to have in mind the bigger picture, which is the evolution away from a bank-based credit system toward a market-based credit system, more suited to a world of financial globalization.  (The enormous expansion of corporate bond issuance since the crisis has only accelerated that shift.)  Today, as before the crisis, marginal credit in the global system is provided by shadow banking, “money market funding of capital market lending”.  In this system, pricing of the capital asset is crucial since typically that asset serves directly or indirectly as collateral for the funding.  And pricing requires market liquidity.  In this system, prices that gap wide from true value during stress times have the effect of freezing up the flow of credit, for everyone.  This is not a guess, but rather an observation from recent experience.

That said, it is also true that discontinuous prices and volatility are strong profit incentives for developing a new and robust dealer system, outside the banks.  And there are signs that such a new system is in fact developing, though such signs are easily missed if our vision is distorted by pre-crisis institutional priors.  What is important is not the survival or profitability of the entities we have grown used to calling “dealers”.  What is important is the survival and profitability of the activity of dealing and market-making, wherever it is happening in the system.   What is important is the existence of balance sheets that can and will absorb temporary mismatch of supply and demand, so reducing discontinuity and volatility, for a price.

The problem faced by central banks today is most analogous to the problem faced in transitioning from war conditions to peace conditions.  They need to withdraw in order to make room for private market-makers, but not so fast that the market collapses and not so slow that no one enters.  That’s the challenge facing Janet Yellen and her colleagues today, as it faced William McChesney Martin in 1952.

11 comments on “Toward a robust dealer system of first resort

  1. Phil Prince on said:

    “And there are signs that such a new system is in fact developing”

    In the new system that is developing, what balance sheet is providing the “dynamic balance sheet expansion” that allows funding to flow from sellers to buyers?

    • Perry Mehrling on said:

      That is the million dollar question, currently being tested in real time. I am watching…

  2. John Rose on said:

    The reprint of this article in Naked Capitalism is titled: “Perry Mehrling: No One Has a Good Idea of How to Keep the Fed From Having to Rescue Mr. Market Again” It caught my eye because I do have a “good idea” for the Fed to implement, although it is much broader and more systemic than the functional problem the article addresses. However, implementing my idea would have such a smoothing effect on the whole economy as to reduce the need for the balancing mechanism addressed in this article.

    My idea is based on two premises, a full acceptance of fiat money with no commodity illusions and the premise that economic instability is based on the disparity of wealth among the participants. By which I mean money not spent on needs and wants or on the production of those needs and wants or on protecting, supporting, enabling that production (mostly government) is excess wealth and essentially destabilizing since it will be chasing more and more scarce sources of increase.

    My proposal is that the Federal Reserve perform its function of regulating the money supply by requiring banks to provide small monthly interest free loans in amounts determined monthly by the Fed to any person with a social security number. These loans would be callable at Fed initiative but only at the same rate at which they are supplied.

    With electronic money transfer systems, the cost would be minimal and the banks could be well, but not excessively, compensated for their function. People would not have to use the money and it might take some time for the system to be understood and incorporated into their spending patterns. But since the money will be supplied in larger amounts in lean times and need be callable, if at all, only in the most prosperous of times, the learning and adaptation would be inexorable. Even any petty fraud that slips through appropriate oversight would not be systemically disruptive since the money would still flow into the system.

    Since, with fiat money, the Fed must supply, on average, an increase in the money supply to match the increase of economic activity, this proposal would vastly increase effectiveness over the present trickle down system of providing that liquidity to the banks for their own choice of uses. It would eliminate the lag time required by the choosing of sound borrowers, if they are even available at the time they are most needed.

    It would also vastly improve the Fed’s ability to perform its second mandate of promoting full employment since that money would be inserted into the economy at a point where it would be used immediately. And since it is money required by the functioning of the whole economy, it is most appropriate that all participants in the economy receive it equally.

    Because of the unique monetary position of the United States in the world economy, this action would have worldwide consequences. Since the dollar is so dominant as a reserve currency, the Fed must supply an increase of money not only to match the economic increase in the US but also to the extent of increase in the other countries that hold dollars in reserve. For the present, those dollars are returned to those other countries, eventually, through our balance of payments deficit and through our military and economic support of world order. So the Fed would be providing substantial service to that world order by smoothing monetary flow throughout.

    I hope this possibility will make its way to being considered in some public or decisive forum. I believe it could be transformative and yet, is not further beyond the mandate of the Fed than other actions they have taken in recent years.

    • Perry Mehrling on said:

      The essence of your plan seems to be a regular small transfer to every citizen, which they can spend or save. Are you assuming they spend, since if they save then it seems like there will be no effect?

      I am not sure why you want to involve the monetary apparatus in this. Would it make a difference if the transfer was funded by issuing Treasury bills? Would it make a difference if all those Treasury bills were purchased by the Fed, using newly created monetary reserves?

      • John Rose on said:

        Hello Dr. Mehrling,

        The essence of my suggestion is the distribution of necessary increases of the money supply directly to all the people who participate in the economy and who thereby create the need for the increase. They should get the profit generated by their activities.

        The mechanism is secondary. My reasons for suggesting this particular one are based on behavioral economics. I suggest using the banks as conduits for the sake of minimal deviation from current functioning and because I believe the Federal Reserve board already has the authority, or can simply exercise the authority to make it happen. They need only supply the money to the banks and direct the banks how to distribute it.

        Involving the banks in this way offers them reasons to not oppose the plan since it increases their customer base in a particularly attractive way. They would be experienced as the source of monthly income! Better than any advertising!

        It would also enhance the status of the Fed as people see them functioning this way and would be an experiential lesson in monetary policy and behavior. I think of these transfers as being very small when little needed and expanding greatly during downturns.

        The learning curve of the public is hard to predict; probably slow at first. But the early adopters will talk to others till more and more sign up. And many will indeed leave the money sit until they become more comfortable with the whole process. The Fed, then, by adjusting the amounts distributed each month according to the extent it is spent, will be provided a direct and finely calibrated tool to do their job.

        The distribution of the amounts as loans is mostly legal fiction to enable the Fed to do the distribution. As fiat money, created out of necessity, it need never be paid back and could be automatically wiped out at death. And when inflation is the problem, it could be effectively countered by minimizing or eliminating the monthly distributions.

        Callability is probably necessary to maintain the legal fiction of the distributions as loans. But it could be useful if the economy were to become so overheated that there would be need to call the loans in order to shrink the money supply. Still that would happen only at the rate they were originally disbursed. And even this could be limited to the individual’s “Fed” account merely creating a negative balance to be restored when payments resumed. People would never have to worry about having to pay it back although, as they gained confidence in the system, they might well want to do so in good times for the sake of quicker income in future hard times.

        The system provides for learning and adjustments on both sides within a context of proper functioning of the banks as retail institutions, fairness to all participants in the economy, as well as enhanced effectiveness by the Fed in the exercise of its mandates.

        I do believe it would be transformative.

        • Perry Mehrling on said:

          See my reply to Bongiovanni below.

          This further elaboration is helpful, as it makes clear (at least to me) that what you have in mind is essentially fiscal policy. A tax rebate to everyone with a social security number is income widely distributed. No need to involve commercial banks or the Fed.

  3. Bold'un on said:

    Thank you for your post. Perhaps what we need is “bullet-proof securitization”, which would create huge (therefore liquid) issues whose quality will vary less than individual issues. Only we need to keep them simple (as compared to pre-2007 versions) by having a single tranche that is not credit enhanced: i.e. if the securitization contains 100 single-A credits then the single-tranche securitization is also A. The investors would give up a little of their coupon in exchange for this liquidity which could reward a decent (say 10%?) equity tranche with an expectred return that is equity-like.
    Securitization involves the SEC rather than the Federal Reserve; ‘bullet-proofing’ might involve a new Securities Act like that of 1933.

  4. To John Rose
    Thanks for your well thought out, kind of ‘mechanical’, approach to ensuring both liquidity and monetary stability, being head and shoulders above anything proposed so far hereabouts, either by Perry or his banker-dealer sources, as they await the pending CRASH!

    But I must ask why you involve the ‘commercial’ bankers in the act of increasing the money supply as to ensure a match between GDP-potential and actual GDP-production.

    This is surely a governmental money-system management role.

    Rather, we have crossed the Rubicon of that modern-day “taboo” of direct money issuance BY the government, as recently advanced by all of Lord Adair Turner, Martin Wolf, Positive MoneyUK, the American Monetary Institute, and the historic legislation developed by progressive Congressman Dennis Kucinich here
    https://www.govtrack.us/congress/bills/112/hr2990/text
    and also the proposal backed by several Nobel(sic) Laureates, and publicly backed by hundreds of economists back in its day.
    http://faculty.chicagobooth.edu/amir.sufi/research/MonetaryReform_1939.pdf

    In other words, we simply need to recognize the mathematical and mechanical potential for money management , as you propose, and come up with a workable, legal plan to make it happen.
    It’s all in the Kucinich Bill.
    Very well thought out indeed.
    Public Money.
    Private banking.
    Waiting for discovery by Dr. Mehrling.

  5. Perry Mehrling on said:

    I appreciate these links, both to 1939 100% money proposal, and to modern Kucinich. There are two issues that I think need to be confronted.

    First, the issue of hybridity. All modern banking systems are part private and part public, and the boundary between these two parts is always in play. The maintenance of par between private and public money is where this all plays out. I don’t think it is possible, or desirable to have a purely public money system, any more than it is possible, or desirable, to have a purely private money system. (MMT and Austrians equally extreme on this point.)

    Second, the issue of “fiat”. I think of money as the highest form of credit, and as such an inside asset. Most people use the word fiat to indicate that they think of money as an outside asset, like gold except with zero production cost. I think of central bank money as analogous to commercial bank money, both of them liabilities of the issuing bank, but at different levels of the system. Because we can move easily back and forth between central bank money and commercial bank money (at par), it is easy to miss the hierarchical character of the system.

    Against this background, my reply to Rose may seem less cryptic. In the history of ideas, there are many examples of people wanting to use “fiat” money as a kind of free lunch, and I thought I could detect some hint of that in the Rose proposal. That’s why I asked him if it would be okay to run his transfer system with Treasury debt rather than Fed debt, and also with Fed debt rather than “fiat” currency.

    He wants the Fed to expand its balance sheet on both sides, issuing new money (liabilities) against new household loans (assets) that are never paid back. I ask whether the same macroeconomic effect could be achieved by Treasury issuing debt which is bought by the Fed (asset) using new money (liabilities) that is then distributed to the population.

    • John Rose on said:

      We seem to agree, Dr. Mehrling, on the usefulness of distributing broadly the needed increases in the money supply. In the scenario you suggest, what would be the mechanism for actual distribution? Would it not require Congressional appropriation, a very unlikely prospect?

      My suggestion builds on the assumption that the Fed can choose how to feed money supply increases into the economy. And I don’t think of those increases as “free lunch” but as legitimate profit from our common economic endeavors, so perhaps indeed as credit, an asset inside our common enterprise.

      Currently that profit flows where? To the banks in the form of interest payments and somewhat to the government by purchase of government debt? (I do appreciate this opportunity to clarify my understanding of macroeconomic mechanics.) To the extent it flows to the banks and ultimately to their executives and shareholders, not to the rest of us, this is a matter not only of usefulness but of justice.

      I may be thinking at an even higher level of abstraction that includes both central and commercial banks as components in the “common enterprise.”

      John Rose

    • Miro Alibasic on said:

      Great post and comments Dr. Mehrling. Thanks for the links Mr. Bongiovanni.
      I see money at present form as useless peaces of paper and future generations would laugh at us! -… what in the world did we carry in our wallets? Bitcoin is proof of that. We’ll have various forms of ‘money’ and actually different services & goods should be payed by different money..
      No one is bigger than the market, not even the Party, which seems more like a pajama party these days..
      very best & thank you.