Financial Intermediation and the Financial Crisis

This discusses the difference between direct and indirect finance, and how indirect finance through financial intermediaries increases economic efficiency. The relationship between financial intermediation and the crisis is also noted.

In particular, I use a simple numerical example to talk about pooling and diversifying risk, pooling over time (i.e. borrowing short and lending long), and pooling small deposits to make large loans, and then relate the risk pooling and time pooling functions to the insolvency and illiquidity issues we are seeing in financial markets. I also briefly note two other functions of intermediaries, reducing transactions costs and reducing default risk, and note the failure of intermediaries to effectively assess default risk is a factor in the crisis.

It's fairly classroom like and somewhat elementary, but I hope it's useful to some of you. My main goal was to show how intermediaries pool and diversify risk (and how that makes the economy more efficient), how the mispricing of risk led to insolvency and liquidity issues, and how this disrupted the time pooling function making things far worse.

I also give an example (taken from here) at the very end of how mortgages can be sliced and diced into different risk categories.

Comments

[Matt]

Pooling of risk does not need many samples before you have gotten to a minimum variance in the aggregate risk. Pooling by the Fannies of the majority of mortgages did minimize risk variance, but applied an unnecessary economy of scale. I would cite the law of large numbers. The same holds true for pooling over time.

So, why do we have national pooling systems, mortgages being held by a few, large institutions? Why did we lose our regional pooling system with regional banks when, after all, the regional banking system was set up to pool risk into local regions?

The teasury for mortgage trade has the purpose of having the government absorbing short term volatility, and thus increasing the asset value of the mortgages as the rest of the economy develops past the volatility. But the volatility does not go away, government has it.

The Paulson bailout is simply another version of the Fannies, government aggregating large pools of mortgages.

But even more uncertain is whether government can estimate the actual value of the mortgages in an economy undergoing transition to an export driven economy.

[salao85]

Great vid - thanks Prof.

Something that eludes me is the nature of risk pricing.

What do economists think about the risk on risk? (I'm no economist) Whatever Heath Robinson device we produce to facilitate intermediation, we still have no way of pricing in systemic change. What is the banking system supposed to do if today a variable that yesterday was insignificant suddenly changes? Even if we untangle the system, make its components less connected and the risk pools smaller, if they all follow the same strategy we'd still have a crisis wouldn't we?

Can anyone recommend a book which discusses society's relationship to risk and the role of banks as the repositories thereof? It seems to me that we have a very poor understanding of systemic risk, exemplified by Greenspan's often repeated line that it's not really possible to tell if an asset is over priced but then I don't really know what truth there is in that.

[Pool]

"So, why do we have national pooling systems..."

Lack of local deposits. Mortgages have to be pooled so they can be securitized for sale overseas. They are pooled nationally because GSE guarantees are necessary for the securities to be marketable.

The old local model relied on pooling regional deposits, and loaning them out regionally. This is not possible without deposits sufficient to meet regional demand.

[ken melvin]

All well and good until things get bad.

[Pool]

Yeah, domestic deposits are generally more stable than foreign loans. Foreign loans tends to move into and out of economies en masse. Nations that depend on foreign loans tend to have boom/bust cycles. I guess large nations are not immune to this principle.

[Matt]

Nice video lesson. Thanks, I learned a lot.

[RC]

Thanks a lot. That was very useful.

[Ronald Rutherford]

Very good lecture Dr. Thoma!!!
I do have a couple of points though:
1. When talking about time pooling you probably should have included the concept of liquidity concerns. You mentioned it later but still could have been introduced earlier for a better understanding of the two problems in solvency and liquidity.

2. Although kind of technical, I still hate when people say that mortgages will have 0 value when the Federal Government takes them over. They will lose a lot like 50% of the nominal loan amounts according to what I read. They do represent an actual home or condo or some tangible asset. Unless of course it is total fraud and in which case I would hope the Government can figure those things out.

3. When talking about tranches, you could also talk about prepayments and how that can affect the payouts of the tranches.

4. Lastly, a link that my boss sent out to nearly all the staff:The link to the recent Fed paper on the subprime mortgage fiasco
http://www.newyorkfed.org/research/staff_reports/sr318.html

Sincerely,
Ronald Rutherford

[DRR]

I appreciate the video Professor. Thanks.

[Roger Chittum]

Excellent.

You just touched on the problem of cascading difficulty amongst institutions that borrowed from and lent to each other. There is a risk even for Very Prudent Bank whose assets are 100% loans to tenured professors with rock solid credit, and no loan losses or late payments since 1960. If some of their deposits or short term borrowings are from troubled institutions, those funds can be suddenly withdrawn and not replaced. It strikes me that being a bank in today's world is like having unprotected casual sex, in that one is exposed indirectly to the diseases of an unknown number of people with unknown histories. Risk spreading and contagion seem inseparable. Also, there was further mispricing in that the contagion risk was underestimated and/or externalized.

As other commenters noted, contagion is no longer regional and the global spreading of risk has introduced a systemic risk of cascading liquidity and solvency problems that is beyond the Fed's capacity to handle. In what ways should the next Congress address the system? Can spreading the credit risk of loan defaults be limited to the point where the contagion risk exceeds the benefit of further spreading of credit risk?

One quibble: A bank can fix a liquidity problem--but not an insolvency problem--by taking in more deposits or other borrowed money. An insolvent bank needs to recapitalize by getting in money it does not need to repay, or is at least junior to the depositors and other general creditors,--i.e., common or preferred equity.

I love your blog.

[Roger Chittum]

Ronald Rutherford:

Please, which paper exactly--by number or title? The link is truncated in your comment as printed?

[Ronald Rutherford]

Roger, maybe this might help, even if unlikely that you will see this post:

Understanding the Securitization of Subprime Mortgage Credit