Kling vs. the International Monetary Fund

The status quo wasn't working, so we need to reform it with a bold two-pronged plan:

1) Add more status

2) Add more quo

(Any similarities to cowbell are purely coincidental.)

Arnold, it seems to me that your views on securitization are formed without understanding of how the financial markets work, or are poorly thought out.

Let me explain: banks make mortgage loans, which are assets. If they don’t sell them to Fennie/Freddie/Ginne or securitize them and sell the securities to investors, they would need to fund them somehow. Surely, then can rely on short-term financing (overnight loans in the inter-bank market, short-term deposits, CDs etc.), but then they would face the duration mismatch (funding long-term assets with short-term liabilities). We’ve been there before: that was the origin of the S&L crisis – not a happy memory. How good do you think thousands of American banks would be in managing their duration risk (the aforementioned mismatch)? Convexity risk? Basis and prepayment risk? Not to mention the ‘Vanna’ and the ‘Volga’.

So, to match their assets and liabilities, they would need to issue longer-term debt. Who would buy their debt? Why, the same investors that now invest in US mortgages. If you think it is difficult to assess the quality of homogenized pools of US mortgage borrowers, how does this sales pitch sound to you: “Herr Weiss, I have for you $50 million of 4.5% coupon, 15-yr bond from Second Peoria National bank! It is subordinated debt, but these guys have been in business since 1972, and they are just magicians when it comes to managing their risk! How about it? (A disclaimer, Herr Weiss: the bond is priced at 101, but if you try to sell it, the bid-ask spread is two points – unfortunately, they are not very liquid, you know.)”

Let me know if you’d like me to explain this in more details.

Ned,
I know about the S&L crisis. I know that mortgages have to be funded. I think that, overall, it is easier to keep a lid on inflation than it is to regulate and manage the securitization process.

If the securitization process is so wonderful, then it ought to be able to survive without government support and without government capital regulations that favor securitization. It has never been able to do so.

I gather you are new to this blog, or you would not have been so condescending in your comments.

Arnold, I have been reading your blog for years; didn’t mean to be condescending, I was just trying to find the best way to illustrate my points (hence “Herr Weiss” – I thought it was just a more entertaining way to point out the problems with bank debt vs. MBS, liquidity, risk management etc.).

Now, back to securitization process: there are large swaths of securitization markets that are not dependent on government in any meaningful way: credit card receivables, auto-loans etc. The so called non-agency mortgage market was of very substantial size, and has dried up (temporarily, in my opinion) because of uncertainty about house prices, i.e. defaults, as well as mass hysteria.

For the ‘agency’ mortgage world, the securitization relies on the government only because the government has set up the GSEs that buy up the mortgage loans. Secondary securitization (creation of CMOs from agency pools) does not rely on government at all. Creation of the CMOs is not about capital regulation – it is about supply and demand for securities with different risk profiles.

As for the government capital regulations, you agree that it is driven by real concerns - bank with one billion of mortgages on its balance sheet and $100MM of capital faces more risks than one with the same capital but only half a billion of mortgages? Given that that’s the case, we can argue what kind of regulation would be best, and there is certainly lot of space for improvement, but the risk ought to be taken by those entities that are best capable to bear it – which is achieved, for example, if mortgages are securitized, thranched, and different pieces sold to different market participants: bank may buy a floater, while a hedge fund may buy highly risky pieces: an inverse floater or an IO. Thus, bank’s capital is secure, and the hedge fund may blow up, which the investors should have known was a possibility.

Cheers, Ned

Ned,
1. Obviously $2 in assets is riskier than $1 in assets. But is $1 in a low LTV mortgage riskier than $1 in a highly-rated structured security backed by junk mortgages? We would agree "no," but the risk-based capital regulations said "yes."

2. The market for securitized loans also depended on risk-based capital regulations that favored securities over whole loans, even with the same underlying risk. If the whole loans and the securities were on the same regulatory playing field, we would see how much value there really is in securitization.

3. As Russ Roberts points out, the capital markets have been artificially distorted by the de facto standard that creditors and counterparties nearly always get bailed out. This takes away their responsibility to exercise discipline. We would have very different financial markets if the participants were less confident of getting bailed out. Interest rates on a whole lot of instruments would be higher. Again, we'd have to say whether the comparative advantage of securitization would be more, less, or unchanged under that scenario.