After an outpouring of excessive and unwarranted humility, TED gave a nice critique of some of the points in my previous post, which I will summarize (probably inaccurately) as (i) Can a system of unlimited liability for banks provide enough capital or does the risk-return tradeoff mean that such a system would hamper growth? (ii) What I meant when I claimed that from a theoretical point of view financial systems don’t require capital was unclear. (iii) TED seems to posit that there is a tension between accurate pricing of risk (i.e. when losses aren’t socialized) and the provision of enough capital to keep the economy growing. I don’t agree that the first and the last points are an accurate description of the tradeoffs faced by society, and rather than address TED’s points directly, in a series of posts I’m going to present my own vision of the relationship between the financial world and the economy – in the process, I hope, addressing the second issue. (And I must admit that it was a sense of immense relief that I realized the piece titled “The Standard Model” was not TED producing a new post on this debate before I’d even managed to reply to the first one, but an entirely different topic. I’m a snail of a writer and look on with envy at the prolific output of bloggers like TED.)
I have long believed that we need to reconceive our understanding of the financial system: the basic ideas on which most of the discourse is based – i.e. intermediation between lenders and borrowers with decisions based on the weighing of risk and return, and the concept that systemic risk originates in a partial reserve banking system’s conversion of risky assets into safe assets – fail to capture the essentials of the financial system.
Faulty modeling means that we fail to understand the nature of the tradeoffs we face (e.g. TED: But lowering the risk of loss we are willing to accept as a society will have ironclad implications on the types of returns we enjoy. Surely there is a happy medium between a low-growth, capital-constrained economy hobbled by unlimited liability to capital providers and the reckless bacchanal we financed with “other” people’s money up to the financial crisis.) and also results in egregiously bad regulatory decisions (e.g. the incentives created by Basel II).
The claim that finance is just a simple risk-return tradeoff ignores the fundamental truth of the industrial revolution. Society went from millennia of washing clothes down by the river (or in a washtub in the house if someone hauled water in) to not even getting one’s hands wet – much less engage in physical labor – over the course of barely more than a century. While the causes of the revolution are debatable, there’s a strong case that it’s all finance: that is, that the risk-return tradeoff exists at different levels and within the context of different financial regimes.
An alternative model: Banks monetize human capital
I want to propose an alternative model: the most important role of banks in the economy is to monetize human capital. When banks fund the working capital of entrepreneurs on an unsecured basis, they make it possible for the economy to realize the value of what’s inside people’s heads – independent of the other resources available to those individuals. If banks could know in advance who would and would not default, the most human capital possible would be realized. Of course, this maximum is unobtainable in practice, so the amount of capital available to the economy is a function of the quality of bank underwriting mechanisms.
The implications of the model are: (i) Banks “create” capital. They don’t simply move capital from one place to another, but are essential to the process by which an intangible and inalienable asset is converted into tangible, alienable assets. (ii) Unsecured debt is a cornerstone of a modern economy.
The traditional models referenced above treat finance as if it’s about stocks of capital and how they are allocated, when in many ways finance is a matter of managing flows of money with the stock of capital only relevant in extreme circumstances. In the simplest framework a disabled landowner, a laborer, and a disabled seed owner can work together to produce food for their themselves, but don’t trust each other. In a world without a coordination device, the land, the seed, and the labor are worth nothing and everybody starves to death. If everybody trusts the bank and the bank is willing to lend (at a spread), then (i) the landowner can borrow from the bank (at x%) hire the laborer and “rent” the seed – returning new seed after the crop cycle, and pay both the laborer and the seed owner with bank IOUs or (ii) the laborer can rent the land and the seed, paying with bank IOUs or (iii) the seed owner can rent the land and hire the laborer paying with bank IOUs. The point is that the bank isn’t lending money that it has or savings that someone has accrued and deposited with the bank, it’s borrowing and lending simultaneously with the result that output, that would not come into existence in the absence of bank intermediation, is produced. In short, banking can facilitate the creation of capital simply by being trustworthy and managing flows without actually “allocating capital” at all.
Maybe the bank is more willing to lend to the landowner or the seed owner, because they have an alienable asset that can secure the debt, but I would argue that historically economic development starts to take off precisely when the collateralized debt constraint is broken; that is, when institutional structures develop such that banks are willing to lend on an unsecured basis and the owners of inalienable capital can get a small line of credit fairly easily at a reasonable rate (e.g. 5%) and can, with careful management, earn the right to have a much larger line of credit.
The role of unsecured credit
The idea that the financial system monetizes intangible, inalienable assets doesn’t apply only to entrepreneurs and human capital. The financial system itself is arguably built on the monetization of such assets. It’s precisely because people trust their banks and the bankers trust each other that the money supply that we have is sustainable. In economic models, this is sometimes called “reputation.” (Existing models however rarely include the liquidity problems that banking is designed to address, and in the absence of such frictions often find that reputation-based equilibria do not create enough value to be stable.)
Implications for regulation
A key goal of financial regulation is to preserve this trust in the banking system. It appears, however, that faulty modeling has meant that regulators don’t have a good sense of the foundations of this trust.
The key here is that lending is unsecured. When lending by banks is secured, what is being monetized is not trust, reputation or human capital, but only the assets themselves. Regulators need to understand that there is a “use it or lose it” aspect to unsecured lending. Unsecured lending forces banks to put in place mechanisms that make unsecured lending reasonable (at least in a world where banks are allowed to fail). These mechanisms then undergird trust in the financial system itself.
When banks are told to seek collateral for their loans to each other (see Alea’s comment here), these mechanisms start to fall into disuse. My concern is that it appears that, as the mechanisms supporting unsecured lending by the banking system disappear, so does trust in the financial system itself. After all, collateralized lending is the easiest and oldest form of lending – it was apparently regulated by the Code of Hammurabi.
In a well-regulated financial system the banks themselves would start the process of shutting down bad banks by restricting their access to credit. The bankers themselves are best positioned to do this: with the movement of employees from one bank to another they can get a very good sense of how their competitors are being managed or mismanaged, and because they compete in the same markets they know when their competitors are mispricing assets. This is exactly the information that is needed to determine which banks are not trustworthy and it would almost certainly be used by banks that (a) know their competitors can fail and (b) regularly extend sizable lines of credit to these competitors.
In our current system it appears that regulators are trying to do the job that banks are better equipped to do. The regulators are searching for some fixed formula (called Basel?) that will be “the” source of financial stability. The underlying problem is that there’s no reason to believe that such a formula exists. Trust, also known as credit, is an amorphous concept that can be capitalized, but when reduced to a simple formula is usually undermined by the existence of a formula that can be gamed. The job of the banker is to stay ahead of the game – possibly by not using simple formulas.
In short, I think regulators should make sure that we have a system where (i) banks can fail and (ii) banks have to lend to each other on an unsecured basis. Bank failures should be a normal enough occurrence that banks are prepared to write off the debt of other banks – and focus on creating safe assets themselves rather than looking to the government to provide such assets in the form of bank liability insurance.
Would this be enough to stabilize the financial system in the absence of increasing the personal liability of the bankers? I don’t know. To return to ideas in my initial post, perhaps in order to address the asymmetric information problems that pervade the financial industry we would also need a policy such that in the event that a bank fails there is a lower standard for creditors to pierce the corporate veil than in non-financial corporations. Imposing the possibility of liability (that would have to be made uninsurable by statute) on directors, officers, employees, and shareholders – to the extent that any of these parties received income from the bank over the previous 10-15 years – may be necessary to prevent misuse of “other people’s money.” Employees should be granted the strongest safe harbors (including, for example, the first $100K per year of income, but not including decisions to gather nickels before steamrollers) and shareholders the weakest.
In addition — or perhaps alternatively — it may be necessary to circumscribe competition in the financial industry. But I’ll leave that to a future post.