Friday, December 11, 2009

Finreg I: Bank capital and original sin

[This post originally appeared at interfluidity on December 11, 2009. It is reproduced here in its entirety.]

I have always flattered myself that I would someday die either in prison or with a rope around my neck. So I was excited when The Epicurean Dealmaker invited me to write about financial regulation and crosspost at a site called The New Decembrists. But my views on the topic have grown both more vehement and more distant from the terms of the current debate (such as it is), and I'm having a hard time expressing myself. So I'll ask readers' indulgence, go slowly, and start from the beginning. This will be the first long post of a series.

Banks are not financial intermediaries. Their role is not, as the storybooks pretend, to serve as a nexus between savers with capital and entrepreneurs in need of capital for economically valuable projects. Savers do transfer funds to banks, and banks do transfer funds to borrowers. But transfers of funds are related to the provision of capital like nightfall is related to lovemaking. Passion and moonlight are often found together, yes, and there are reasons for that. But the two are very distinct phenomena. They are connected more by coincidence than essence.

The essence of capital provision is bearing economic risk. The flow of funds is like the flow of urine: important, even essential, as one learns when the prostate malfunctions. But "liquidity", as they say, takes care of itself when the body is healthy. In financial arrangements, whenever capital is amply provided — whenever there is a party clearly both willing and able to bear the risks of an enterprise — there is no trouble getting cash from people who can be certain of its repayment. Always when people claim there is a dearth of "liquidity", they are really pointing to an absence of capital and expressing disagreement with potential funders about the risks of a venture. Before the Fed swooped in to provide, 2007-vintage CDOs were "illiquid" because the private parties asked to make markets in them or lend against them perceived those activities as horribly risky at the prices their owners desired. There was never an absence of money. There was an absence of willingness to bear risk, an absence of capital for very questionable projects.

No economic risk is borne by insured bank depositors. We have recently learned that very little economic risk is borne by the allegedly uninsured creditors of large banks, and even equityholders — preferred and common — have much of the risk of ownership blunted for them in a crisis by terrified governments. The vast, vast majority of bank capital is therefore provided by the state. Prior to last September, uninsured creditors of US banks were providing some capital. Though they relied upon and profited from a "too big to fail" option when buying bonds of megabanks, there was some uncertainty about whether the government would come ultimately come through. So creditors bore some risk. During the crisis, private creditors wanted out of bearing any of the risk of large banks. Banks were illiquid because private parties viewed them as very probably insolvent, and were unsure that the state would save them.

The US banking system was recapitalized by precisely three words: "no more Lehmans". All the money we shelled out, the TARP, the Fed's exploding balance sheet, the offered-but-untapped "Capital Assistance Program", mattered only insofar as they made the three magic words credible. At this moment the Fed and the Treasury are crowing about how banks are now able to "raise private capital" and about "how TARP is being repaid" and "losses on TARP investments will be much less than anticipated". That is all subterfuge and sleight-of-hand, flows of urine while the beast lumbers on. The US government has persuaded markets that it stands behind its large banks, that despite no legal right to such protection, all creditors will be made whole and equityholders will live to fluctuate another day. Banks have raised almost no private capital, in an economic sense. They have attracted liquidity on the understanding that the government continues to bear the downside risk.

It's unfair to say that the government now supplies all large bank capital. Stockholders still suffer price volatility and there is some uncertainty that the government will remain politically capable of being so generous. But the vast majority of large-bank economic capital is now supplied by the state, regardless of the private identities and legal forms associated with bank funding arrangements. So long as the political consensus to support them is strong, American megabanks are in fact exceedingly well capitalized, as the US dollar risk-bearing capacity of their public guarantors is infinite. But that is not a good thing.

No iron law of economics ensures that those who bear the risk of an enterprise enjoy the fruit of its successes. Governments have the power to absorb the downside of formally private enterprises (and the libertarian so principled as to refuse a bail-out is very rare indeed). But they have no automatic ability to collect profits or capture gains from organizations that they economically capitalize, but do not legally own. Bearing the downside risk of a project with no claim on the upside is the circumstance of the writer of an option. Private parties who write options, either explicitly or implicitly via credit arrangements, demand to be paid handsomely for accepting one-sided risk. Governments do not. Banks pay deposit insurance premia, but only on a fraction of the liabilities the government guarantees and in a manner that does not discriminate between the prudent and reckless (which creates a public subsidy to recklessness). When governments have lent to banks during the crisis, they have lent at well below the rates even untroubled, creditworthy nonfinancial borrowers could obtain on the market, so that the implicit option premia embedded in credit spreads were somewhere between negligible and negative. Governments paid banks for the privilege of insuring their risks, or to put it more accurately, they acknowledged that they had already insured bank risks and found ways of paying out claims via subsidies sufficiently hidden as to be politically palatable.

As we think about how to regulate banks going forward, we must first be clear about what we are doing. We are negotiating the terms of options that the government will offer to bank stakeholders. It is important to understand that, for both practical and philosophical reasons. As long as the state substantially bears the downside risks of the financial system, by virtue of explicit legal arrangements or de facto political realities, philosophical arguments for deregulation are incoherent. Deregulation is equivalent to a blank check from the state. If you are philosophically in favor of free market capitalism, you must be in favor of very radical changes in the structure of banking, towards a system under which the state would have no obligation to intervene, and would in fact not intervene, to support bank stakeholders even when their enterprise threaten to collapse. The "resolution regimes" currently proposed do not restore "free market" incentives, because those proposals codify rather than forbid state assumption of risk and losses in the event of a crisis. A free market resolution regime would allocate losses among private stakeholders only, and prevent any loss-shifting to the government. For the moment, such a regime, and the structural changes to the banking system that would be required to make it credible, are politically beyond the pale.

So, the options written by government to bank stakeholders will remain in place. All that remains, then, is to negotiate the terms of those options. Framing bank regulation in terms of option contracts underlines a reality that is tragic but true: options are zero-sum games. One party's benefit is another party's cost. Very deeply, there is no confluence of interest we can seek between our best and brightest financiers and the public good. Terms that are good for banks are bad for taxpayers. Negotiating the terms of an option with a wealth-seeking counterparty is an inherently adversarial affair. When President Obama is on the phone with Jamie Dimon, do you think he keeps that in mind? A fact of life that our President seems not to enjoy is that while sometimes there are miscommunications that can be resolved via open exchange, sometimes there are genuine conflicts of interest that must simply be fought out. Bank regulation, alas, is much more the latter.

There are indirect as well as direct effects of option contracts, so maybe I've framed things too harshly. After all, we allow and encourage formal derivatives exchanges because, even though the derivative contracts themselves are zero-sum, they permit businesses to insure against risks, and that insurance can enable real wealth creation that might not have otherwise occurred. We claim that derivatives markets make indirect positive contributions to the real economy, despite the fact that their direct effect is simply to shuffle money between participants. Banking also just shuffles money around, but we generally think that it enables important business activity. So one might argue that banks and the public have a common interest after all, and smart regulation to evolve could from a more consensual process. But, one would be wrong. We all have a stake in the existence of a payments system, and banks provide one, but managing that is a largely riskless activity, conceptually separable from the lending and investing function of banks. We would also like our economic capital to be allocated productively. But the effect of writing a more bank-friendly options contract is to reduce the penalties for poor capital allocation while enhancing the payoffs to big, long-shot risks. Banks destroy Main Street wealth and create Wall Street crises by making foolish and indiscriminate use of the capital entrusted to them. If we desire a better banking system, we must limit the degree to which private stakeholders can expect to be made whole by the state. With respect to regulation, what's good for Goldman Sachs is quite opposed to what is good for America. The point of regulating is to align public and private interests by imposing costly limitations on how banks can behave. Indirect considerations of public welfare reinforce rather than reduce the degree to which bank regulation is zero-sum, a fight that pits the health of the real economy against the distributional interests of bank stakeholders.

However, there is some light in the bleakness. Once we understand that we are negotiating option contracts, we can look to some guidance from the private sector. Option-like private contracts are negotiated all the time, and the issues that surround managing them are well understood. A compare-and-contrast of public sector bank regulation and private sector contracts will prove informative. That will be the subject of our next installment.

Acknowledgments: To be acknowledged by me is like being kissed by a putrescent semi-decapitated halitotic zombie creature. Nevertheless, I failed to weave many links into the above, and my thinking on these issues owes a lot to a bunch of people who are smarter and better smelling than me, so I feel duty bound to mention them. In particular, it was Winterspeak and Mencius Moldbug who fully disabused me of the notion that banks are intermediaries between private parties, which pushed me to think more deeply about the meaning of bank capital, and capital in general. Others who have the misfortune of having influenced my thinking on these issues include supercommenter JKH, Economics of Contempt, Wonkess, The Epicurean Dealmaker, James Kwak, Mike Konczal, and John Hempton. (I'm sure there are more I've missed: count yourselves lucky!) But the views expressed above are my own, and all of the people I've mentioned are far too sensible lend them any credence.


  1. Excellent, thought-provoking post, Steve. Thanks for adding your voice to the conversation.

    I am intrigued by your concept of an option contract at the core of regulation of the banking system, and I look forward to reading more about it.

    I agree that most people play far too fast and loose with the concept of "provision of capital" by banks and investment banks. Properly understood, the core function of both commercial and investment banks is to act as conduits for capital flows between the actual providers and end users of capital. Bank intermediaries perform marketing, aggregation, and information transfer functions as well, but it is not an inapt analogy to describe the pure banking system as a very large and complicated system of pipes, valves, pumps, and reservoirs. In this view, banks do not own the capital that flows within their system, but simply make it easier to distribute that capital more broadly and more efficiently.

    Of course, it is easy in such a system—and very tempting—for banks and investment banks to use their own capital to participate alongside actual investors as risk capital providers, as well. In part, this is actually a desirable feature of any risk capital transport system: traditional "market making" consists of banks using a portion of their own capital to smooth flows and bridge supply and demand on a short term basis. This reduces the volatility and "stickiness" of capital flows and is arguably good for everybody.

    However, there is no bright line between market making and true proprietary trading or investment, and the profits accruing to these latter activities in recent years have been so attractive that banks have increasingly migrated to them. This is where most of our problems came from, and this is where the rub lies.

  2. I think the post and comment are excellent.

    Financial institutions have relied on government guarantees for hundred of years... see Anthony Haldane, from the Bank of England, in his speech "Banking on the State".

    Banks reviewed the "optionality" of the guarantees put forth by the US and other governments and engaged in a cost benefit analysis... here from a Moody's report on banks and the TALF in March...

    "...Banks have access to other funding alternatives that are more cost effective, including access to other government programs such as the FDIC’s debt guarantee program, which was recently extended. The issues that may keep banks on the sidelines with respect to the initial round of TALF are: a lack of immediate need for funding; a desire to observe results of the initial funding, including pricing obtained; and a need to clarify certain outstanding issues and how they are being worked out (e.g. refinancing risk, details of the Customer Agreement contract between dealers and investors). Ultimately, these factors will be evaluated on an individual bank basis, security by security. Prospectively, however, when TALF loans are granted for a broader range of securities, TALF may have a greater appeal to banks, which could benefit in three ways: as originators and, to some extent as broker dealers, but perhaps more importantly as investors.

    Banks, especially large ones, have in the past more heavily relied on CMBS to fund their lending activities (banks have traditionally represented 20% of Aaa CMBS origination) and, to some extent, on RMBS. These securities have also played a role in banks’ investment and underwriting activities. If TALF ultimately has a positive credit impact on the overall market and helps improve demand, and thus pricing, it would be supportive of the economics of issuing new securities. However, we only expect this to happen modestly over time.

    Banks’ participation as issuers in future TALF phases will depend on whether the terms of the program are sufficiently attractive, especially with regards to the duration of the TALF loans.

    Currently, TALF is granting loans on three-year terms, which is typically less than the maturity of the securities that banks issue. This may cause investors to demand a risk premium for the refinancing risk of this mismatch (see earlier section on investors).

    This premium makes issuing under TALF more financially challenging for banks. An extension of loan terms would help address this concern.

    We can't assume a level of purity in the market structure of banking and markets. If we do it will make developing and implementing law difficult.

    Also no one talks about the shadow banking system. Do we assume it's unlikely to come back?

  3. Sorry... I didn't separate my writing from the Moody's piece...

    I wrote "We can't assume a level of purity..."

  4. The new "macro-prudential" buzzwords...

    "Macro-prudential recommendations will not be addressed to individual institutions.

    This remains the sole prerogative of micro-prudential supervisors.

    Instead, they will be addressed to regulators, supervisors or other authorities so as to foster regulatory and supervisory policies.

    These authorities have instruments which can steer variables such as maturity mismatches, excessive credit growth, leverage, or the risk appetite of market participants.

    While it might not be desirable or feasible to target these variables directly, prudential policy instruments – individually or in combination – can affect them indirectly.

    From a speech by by Jean-Claude Trichet, President of the ECB
    The Economist’s 2nd City Lecture
    London, 11 December 2009

  5. TED — I think we are and have just about always been very far from a world in which banks, commercial or investment, are mostly conduits. Your unusually noble (despite your pretensions to ignomy) corner of the business, is an exception. You really do bring buyers and sellers together, without putting bank capital on the line. Yeoman unleveraged brokerage services are also conduit-like, in that even they they bloat a bank's balance sheet, the bank's exposure is perfectly hedged, as fluctuations in assets are exactly offset by liabilities to clients.

    But even the purest IBs let clients trade on margin. They are no longer in the conduits and tubes business. Market-making is the art of providing liquidity by taking on transient risk-of-ownership. Underwriting issues is conduits-and-tubes only when the IB only when the underwriter doesn't precommit to the sale of the new issue, which is now exceedingly rare.

    Historically, however, there has been limited competition in those traditional IB businesses, so IBs set the terms of business in a manner that was both profitable and very safe for the issuer. New issues are famously underpriced, so IBs don't take much risk committing to sell them. Traditional margin accounts, at least 50% equity that can be liquidated at the first hint of shortfall, are effectively zero risk loans, while retail investors pay rates often akin to unsecured credit lines.

    IBs could, historically at least, run themselves as very safe businesses, despite taking on a bit of balance sheet at the margin. Commercial banks have always been a different story. Nearly all client activity at a commercial bank turns ends up on bank balance sheet, and nothing other than luck or cleverness matches the performance of bank assets with the behavior of its liabilities (a matching which under a "conduit and pipes" theory would be perfect). An M&A shop can, rightfully, claim that it is just an intermediary. A Traditional IB cannot, but it can get away with pretending. A commercial bank always did but never could claim that, given the promises it makes and the assets it holds. (Thus the faint whiff of fraud that has always surrounded commercial banking ever since goldsmiths oversold storage receipts.)


  6. [cont'd from above]

    Why then did IBs abandon their near conduits-and-pipes role? Because the people in the business, like the guy you see in the mirror, are not stupid. To be a conduit is safe, and with limited competition you can earn a living, but it ain't minting money. But by farming the optionality at the heart of the financial system, you can make a killing. Why would you not do that if you can? And there are so many ways for "financial intermediaries" to play that optionality. You just have to be able to make promises that are widely taken to be rock-solid, specifically a promise to give people their money back, and then gamble. You don't have to be a formal commercial bank, or an investment bank, or anything. You can be a mutual fund, forchrissakes, that's what money market funds are, and trouble their triggered TARP. Any entity is "too big to fail" if it has represented its liabilities as not-risky and acquired funds on those terms from parties that are politically sympathetic and haven't hedged against the possibility of loss (since this is allegedly safe money).

    Smart people will play this game. They will fool themselves into thinking the promises they make can be kept, because most people don't like to imagine themselves as thieves. But in the back of their minds, they know that there are risks that they can't manage, and from which they profit. They hope for the best as they get wealthy. But they always claim it was something about the system, if everybody is doing it why should I get in trouble?, if that risk they cannot manage turns against them, as it inevitably does eventually.

    Investment banks had a comparative advantage at this game, because they had reputations for making rock-solid promises. I don't worry when I ask my brokerage to hold shares on my account. I already trust them. It's perfect that the financial crisis really began with two Bear Stearns hedge funds imploding. Bear, despite it's "sharp-elbowed" rep on Wall Street, was known on Main Street as a large, respectable, venerable bank whose representations could be trusted. Sure, the prospectus of those hedge funds reminded investors of risk and yadayada, but people went with Bear hedge funds rather than their cousin's garage fund because it was Bear Stearns. Bear loved this business, arbitraging their reputation for importance and solidity into the high margin hedge fund business. Until the funds went bust. And then what happened to the bank was exactly the same thing, just on a larger scale.

    IBs would stick to their role of being conduits much more if everyone perceived them as skanky and shady, which they are, rather than venerable and powerful (which they are too, but that power is not employed on behalf of clients except in the happenstance of a confluence of interest). No entity that accepts funds and intermediates with assets imperfectly matched to liabilities is a conduit, or trustworthy. At least not without the help of a very lucrative option from the state.

  7. lawmaking — i appreciate the cites. of course i'd heard the doom loop line over and over again, but you inspired me to read the Alessandri & Haldane essay, which really is excellent in its description of the problem (i thought it good, but not quite as excellent, in its proposed remedies.)

    there's certainly nothing new, or nothing that bank regulators don't understand, in pointing out that the banking system and a whole and sometimes banks individually extract and option from the state (as your examples suggest). i've belabored the point a bit, for two reasons. first, i think it's important to remind people that there's an element of zero-sum game in bank regulation, that we have reasons to be suspicious of an overly collaborative process; and second, think it's crucially important, in evaluating bankers' claims to freedom as "private actors", and also to evaluate policymakers who are currently crowing about "funds being repaid", that people understand the difference between the formal liabilities of bank balance sheets and the actual economic capital of banks, which largely derives from government, and is not much altered by transfers of funds.

    anyway, thanks again for the useful cites and excerpts, and thoughtful comments. we absolutely can't assume purity. on the contrary, my claim is that we are negotiating rents, plain and simple, and we may want to pay some to lubricate a banking system but we should be very attentive to how much our pockets are being picked.

    re the shadow banking system, in my piece i was deliberate vague about what kind of banks i was talking about. but the argument stated applies to "shadow banks" as surely as commercial banks or investment banks. if a CDO is held by a municipal government that will ultimately be bailed out by the federal government (perhaps via monetization), then that "shadow bank" is as much an option written by the central govt as Citibank. what defines a "tbtf" bank is ultimately who its creditors are, and the political or social cost of letting them bear losses. nothing more or less. organizational structure is irrelevant ex post (although organizational structure might contribute to the incentives that lay the risks on those unprepared to manage them other than via bailout.)