Arnold Kling  


Morning Commentary... The Origins of Money...

Richard Green writes,

depositories are not capable of holding long-term fixed-rate mortgages, because it subjects them to too much duration risk: mortgages are assets with long duration (i.e., have values that are sensitive to changes in interest rates), while deposits are liabilities with short duration.

...the basic MBS [mortgage-backed security] was and remains an ingenious product, and will continue to be an important instrument of housing finance in the years to come.

We are not exactly in agreement here. Some quick points.

1. The mortgage security does not make duration mismatching go away. To the extent that mortgage securities wind up held by banks (where they get a generous risk rating from regulators), the duration problem is right back where it started.

2. If Freddie and Fannie hold the securities (which they increasingly were doing), then you get a highly leveraged, highly concentrated pile of mortgage securities. This was a financial land mine waiting to be stepped on, and this summer it blew up.

3. If there are natural holders of long-term assets (pension funds or insurance companies), then there is nothing stopping banks from issuing long-term debt. Then banks can more safely fund long-term mortgages.

The financial system I want to see around mortgages is many banks, holding loans that they originated themselves, with decent-sized down payments (you don't get such big housing bubbles when people put down 20 percent), and with sufficient capital to ensure that shareholders and not taxpayers are the biggest losers if the bank messes up. This is pretty much the system that existed in 1968, prior to securitization. It blew up in large part because inflation got out of control, so that long-term interest rates rose, destroying the value of fixed-rate mortgages. So part of my ideal system of mortgage finance is that the Fed doesn't lose control of the inflation rate again. I think that's a lot easier to orchestrate than a system to try to put back together the Humpty-Dumpty of the mortgage-backed securities market.

Comments and Sharing

COMMENTS (7 to date)
aaron writes:

While I've agreed with you for the most part these months, I agree with Green here. MBS can't make uncertainty go away, but they can better distribute risk to people who can afford the loss of income.

Willie writes:

Remember the banks were the ones who originated/packaged the vast majority of these aggressive mortgages and retained the riskiest pieces, emboldened by the ability to raise FDIC-insured deposits. The worst of the mortgage products (think OptionARMs) were not able to be sold in the capital markets and were retained on balance sheet. Do you really want the IndyMac, Countrywide, WaMu and Wachovia's of the world to be the gatekeepers? Say what you will about the credit markets but the biggest blunders were made by the big banks & broker/dealers (Citi, UBS, ML, etc.) who ultimately stuck the taxpayer with the bill. Rewarding them for their irresponsible behavior would be a travesty. . .

Mark Seecof writes:

My immodest proposal: do away with (NRSRO) ratings on MBS and other CDO. Now's the time, since the NRSRO's have lost all credibility anyway. Replace ratings with bond insurance, issued by proper insurance companies. Everyone, including bank regulators, could look to the insurance coverage (amount and quality of issuer) to assess the risk of particular bonds. High-risk bonds would have partial or dubious insurance coverage, low risk would have complete and very credible coverage.

I realize a CDS is sort of like insurance, but a CDS doesn't run with the bond, and it is clear that many financial firms wrote "naked put" CDS's without having enough capital to cover them.

Unlike NRSRO's, insurers' interests are mostly aligned with bond purchasers' even though insurers are paid by issuers. Insurance regulation, imperfect as it is, is more likely to keep insurers responsible and solvent than the (obviously inadequate) oversight of NRSRO's. Insurers would have the incentive, and generally the ability, to estimate bond risks more accurately than NRSRO's or downstream folks like bank regulators.

Banks would not need to buy insurance on loans they didn't syndicate into MBS or CDO. For loans banks hold on to, they have the right incentives as well as the ability to manage their own risks (and the performance of their underwriting and servicing operations). Bank regulators can assess banks' own loan portfolios easily enough.

My scheme would undo one of the big problems Prof. Kling has written about, the wrong notion that it is less risky for some bank to hold a bond representing loans written by some other firm than to hold a loan written by the bank. My scheme would fix that problem two ways: first, there would be no bond ratings to mislead regulators; second, with no need to pay for insurance, banks' own loans would usually be more profitable than bonds representing other lenders' loans-- and when credibly insured bonds were cheaper, that would indicate that a bank's own loans were too risky!

Lord writes:

What causes bubbles is lending based on appreciating property values and not the ability to repay.

Bill Woolsey writes:

I would think that fixed rate, long term to maturity, mortgages could be an appropriate part of a well-diversified portfolio of a depository instituion.

An S$L-type institution, that specializes in using deposits to hold fixed-rate, long-term mortgates, is a mistake.

While last time they failed because of an interest rate spike, they would also have problems with a price bubble like today.

The commercial banking system as a whole looks to be too heavily into mortgages. (My figures showed
40%, and that wasn't counting mortgage backed

But that doesn't mean they can't hold any.

Oddly enough, "base deposits" are a pretty secure source of longterm funds. Of course, there is interest rate risk. (And credit risk.) It is just that they can't hold many of them.

Francisco Aboim writes:

It seems to me that when we view these institutions pooling mortgages, and levering themselves on top of these maturity mismatched assets (we could say the only thing that changed was the "fruit" on the MM maturity mismatch tree), we have the same problem which we face with banking and general currency. Particularly in this case true because these pooled mortgages were federally "sponsored" (I don´t care about the underlying assets, with the FED behind me, i´ll give my personal AAA rating here). In this view, though oversimplified, down payments and other limitations would be, therefore, simply analogue to a fractional reserve requirement system. Credit expansion would naturally follow, albeit without much parameter for a ceiling since "reserve requirements" were based not only on downpayments but future values which, when the bubble burst, led to "margin calls" which reversed the cycle. Given there are differences, namely that the dollar is the reserve currency of choice and the dynamics of the money market have become much more complex then that of MBS, would it be then too much of a stretch to extend your conclusions to the whole monetary system?

Bill Woolsey writes:

Banking and leverage.

Banks are financial intermediaries. They are middlemen. Just like a retailer purchases goods from manfacturers or wholesalers and then sells to consumers, the bank borrows money from depositors and uses the funds to make a variety of loans-- consumer, commerical, and mortgage.

Middlemen generally have to pay for a location, employees, and the like. They have to finance this activity somehow. Middlemen usually have to finance the goods they plan to resell as well.

Banks are in the unusual situation that what they are "buying" is money and doesn't require financing exactly.

Anyway, the _core_ of the banking business is pure leverage. Infinite leverage.

In reality, of course, banks have traditionally financed some of their activity with capital. It provides reassurance to depositors. Early bank regulation generaly required a fixed amount of capital to start a bank. Modern bank regulations impose capital ratios. The bank owners must finance a certain share of the bank's total assets. The idea is that because depositors who are insured don't care about their banks' capital, the deposit insurer must insist on capital.

Perhaps it is my imagination, but it seems to me that some people whose perspective on this comes from the realm of personal finance treat leverage like a boogyman. If your broker suggests that you buy a stock because it will rise in value, and then suggests that you borrow money from him to buy extra and so "leverage" your potential profit, it sounds very risky. If the stock goes down, your losses are magnified as well.

There is no doubt that banks with 50% capital ratios would be very safe. But always remember the core business of financial intermediaries before you start to think about banks using "leverage."

Comments for this entry have been closed
Return to top