Friday, September 26, 2008

Mortgage securities oversold by 15-25 percent

Below are some quotes which support the view that mortgage assets are currently undervalued by the market. Yes, the market is inefficient -- it overpriced the assets at the peak of the bubble (greed), and is currently underpricing them (fear). Both Buffet and ex-Merrill banker Ricciardi below think the mispricing is about 15-25 percent. That is, the "fear premium" currently demanded by the market is 15-25 percent below a conservative guess as to what the assets are really worth. This is the margin that can be used to recapitalize banks, perhaps without costing the taxpayer any money, simply by providing a rational buyer of last resort and injecting some confidence into the market. Note to traders: yes, this is obvious. Note to academic economists: this is yet another market failure -- but of an unprecedented scale and complexity.

(Actually, 15-25 percent is not bad, and just shows that credit markets are generally more rational and data driven than equities. During the Internet bubble and collapse you had mispricings of hundreds of percent, even an order of magnitude.)

Warren Buffet interview from CNBC:

Government intervention necessary to restore confidence in the market.

If I didn't think the government was going to act, I would not be doing anything this week. I might be trying to undo things this week. I am, to some extent, betting on the fact that the government will do the rational thing here and act promptly.

Mispricing is about 15-20 percent:

...all the major institutions in the world trying to deleverage. And we want them to deleverage, but they're trying to deleverage at the same time. Well, if huge institutions are trying to deleverage, you need someone in the world that's willing to leverage up. And there's no one that can leverage up except the United States government. And what they're talking about is leveraging up to the tune of 700 billion, to in effect, offset the deleveraging that's going on through all the financial institutions. And I might add, if they do it right, and I think they will do it reasonably right, they won't do it perfectly right, I think they'll make a lot of money. Because if they don't -- they shouldn't buy these debt instruments at what the institutions paid. They shouldn't buy them at what they're carrying, what the carrying value is, necessarily. They should buy them at the kind of prices that are available in the market. People who are buying these instruments in the market are expecting to make 15 to 20 percent on those instruments. If the government makes anything over its cost of borrowing, this deal will come out with a profit. And I would bet it will come out with a profit, actually.


Christopher Ricciardi, former head of Merrill's structured credit business, in an open letter to Paulson. Note his comments illustrate the role that psychology, or animal spirits (Keynes), plays in the market.

The securitization market worked exceptionally well for decades and was the financing tool of choice for large and small institutions alike. As investments, performance for securitized assets typically exceeded corporate and Treasury bond investments for decades.

Where securitization went wrong in recent years was with subprime mortgages. These securitizations performed disastrously, causing people to mistakenly question the practice of securitization itself.

Decades of historical data were ignored, with the subprime experience exclusively driving market perceptions: The entire securitization market was effectively shut down, and this explains the depth and persistence of the ongoing credit crisis.

Government purchases of illiquid mortgage assets from the system will cost taxpayers significant sums and expose them to downside risk, without addressing this fundamental issue. Billions of dollars held by all the major institutional bond managers, hedge funds and distressed funds are already available to purchase mortgage assets.

However, in the absence of a way to finance the purchase of these assets, such funds must bid at prices which represent an attractive absolute return acceptable to their investors (15% to 25% typically), resulting in typical transaction terms that have significantly impeded the sale of mortgage securities to these funds. If these funds could finance their purchases, especially under efficient financing terms, they would still require similar returns, but would be able to buy many more assets, and bid higher prices for the assets.

Our financial system needs the capital markets and the natural power of securitization to get a jumpstart from the government. I propose using the powers granted to Treasury to create “vehicles that are authorized…to purchase troubled assets and issue obligations” under currently contemplated legislation to more efficiently address the crisis and establish a program which we might call the Federal Bond Insurance Corporation (”FBIC”), as an alternative to simply having the government directly purchase assets.

Comment re: behavioral economics. The preceding housing bubble and the current crisis are very good examples of why economics is, at a fundamental level, the study of ape psychology. On the planet Vulcan, Mr. Spock and other rational, super-smart traders and investors would have cleared this market already. But we don't live on Vulcan. Anyone who wants to model the economy based on rational agents who can process infinite amounts of information without being subject to fear, bounded cognition, herd mentality, etc. is crazy.

When the conventional wisdom is that house prices never go down (people believed this just a couple years ago), you risk little of your reputation or self-image by investing in housing. When the conventional wisdom is that all mortgage backed securities are toxic, you must be extremely independent and strong willed to risk buying in, even if metrics suggest the market is oversold. This is simple psychology. Very few people can resist conventional wisdom, even when it's wrong.

10 comments:

Anonymous said...

I'm confused by those two quotes. I could interpret them as saying that the assets are priced 15-25% too low, or that people wanting to buy those assets are demanding 15-25% yields, which would be very different. Which is the correct interpretation?

Anonymous said...

They shouldn't buy them at what they're carrying, what the carrying value is, necessarily. They should buy them at the kind of prices that are available in the market.

That is the key point. My impression is that in the Paulson plan the government is NOT buying at the market price, but rather paying the carrying price. In which case it is not going to be a money-maker for the taxpayer.

It is indeed absurd that "rational expectations" and other such idealistic "theories" are taken so seriously by academic economists--and they get Nobel prizes for it! Of course, it is no surprise there is market failure (and you get Nobel prizes for discovering that too!). And unlike string theory, it is not even profound mathematics. Any good physicist knows the limits of a model, and how seriously to taken any prediction from such a model. Sorry, but the "big shots" in economics are nowhere near those in theoretical physics, who are truly farthest out on the tail. Thus, I am still not convinced Paulson and Bernanke know what they are doing (And I am not sure if at least one of them has other motives---what was the point of including Section 8? ). I would be more comfortable if you were at helm, for instance :-)

MFA

Steve Hsu said...

If the assets are underpriced by X, then your return will be X if you buy them at the market price and either wait for the market to become rational (and then resell) or hold to maturity.

If I tell a consortium of investors I want to invest in underpriced CDOs and they *demand* a return of X, I had better be able to buy the CDOs at a X less than the "real" value.

I hope that makes sense.

MFA: I think Paulson and Bernanke know what they are doing. It's a good team -- a smart academic and a guy with real Wall St experience. Paulson definitely has conflicts of interest, though.

Anonymous said...

MFA: Why wouldn't the carry price (balance-sheet price ?) be the same as the market price if the securities are already marked to market ?

If the government were to buy the securities at market price, would that lead to sufficient capitalization for the banks to allow the credit markets to unclog ?

If the government were to buy the securities at higher than market price, would there still be a sufficient return to the taxpayer ?

Finally, which are the *investment* banks that still have these toxic securities on their balance sheet ? lehman/bear/merrill are gone, morgan-stanley/goldman/jp-morgan claim they don't have that much exposure. So which big first-tier investment or quasi-investment bank is left (Citi ?). Or are we just talking about commercial banks at this point.

Steve Hsu said...

I don't know if the mispricing margin is enough to really recapitalize the banks without an outright transfer of funds from taxpayers. I would guess not -- somewhere there is about a trillion in housing losses if we go back to historical home prices (another 20% decline) and I think existing markdowns are less than half that (could be much less) -- so that still leaves hundreds of billions unaccounted for.

I don't think there are any true i-banks left and, unless I am mistaken, that means what remains (commercial banks) can use long-term valuations in their accounting -- they don't have to mark assets to market at this instant in time, which was a big part of causing the insolvency. That's good news.

The bad news is that a lot of the entities holding the bad paper are hedge funds, foreign banks, even perhaps foreign central banks. Do we want to spend taxpayer money bailing out hedge funds? If Treasury has an auction and a bunch of hedge funds step forward to unload their CDOs (esp. synthetic ones), won't there be a huge public outcry?

Unknown said...

I find it strange that with all the talk about the possible valuations of all these MBS, no one in the media (or anywhere else, that I can see) has taken a representative example and simply laid out exactly what goes into these things, which mortgages are used, the projected default rate, etc. It would require some math, but with all the stuff taken over by the gov't we should at least have a little transparency.

And in defense of economists. Yes, rational expectations, equilibrium, and the lot often fail in the real world, but it was mostly not academic economists who manned the securitization machine, wrote the code that priced the securities, etc. Nor have I seen many physicist/quants (except Steve) who expressed much doubt before the crisis broke.

And finally: suppose I believed that MBS were undervalued. Is there any way at all for individuals to buy them? Or is it reserved for the Buffetts of the world?

Steve Hsu said...

Here are some posts that really get into the nitty gritty of what is inside a typical CDO. Because individual CDOs are so different, Treasury is going to have to do a lot of analysis themselves (or hire the right people, like PIMCO). The reverse auction model sounds great, but there aren't simple categories of similar CDOs that you can just use as buckets.

anatomy of a cdo

deep inside the subprime crisis

mackenzie on the credit crisis

gaussian copula and credit derivatives

I've been covering this since 2005! (And Fannie derivatives issues since 2004.)

Quants I know have been pretty skeptical about credit structures for a long time. But if you are employed in the industry it is unlikely that you are going to speak in public about such things.

Anonymous said...

...and, unless I am mistaken, that means what remains (commercial banks) can use long-term valuations in their accounting -- they don't have to mark assets to market at this instant in time, ...

This is interesting, why would the i-banks (lehman, bear) want to mark to market. Did margin calls on their highly leveraged positions trigger this ?

zarkov01 said...

Any theory about the value of MBS had better assume that real estate will continue to fall in price and foreclosures will continue as more poorly underwritten mortgages go into default. What are the assumptions behind that 15-29% figure?

Steve Hsu said...

I believe the i-banks were forced to use mark to market or "fair value" accounting for their level 3 assets under FASB 157. Unless I am mistaken commercial banks are somehow allowed to carry assets on their balance sheet using some kind of longer term valuation, which was one of the factors for i-banks wanting to become commercial banks.

http://en.wikipedia.org/wiki/Mark_to_market

http://www.cfo.com/article.cfm/10097878

http://www.bloomberg.com/apps/news?pid=20601087&sid=aoHW3zyqTeAA&refer=home

"As bank holding companies, Goldman and Morgan Stanley will no longer be required to mark all of their assets to the current market values. Assets held in the bank divisions of the firms don't have to be valued at market rates, potentially allowing the companies to allow further writedowns on the value of their holdings.
``We are moving assets from a number of strategic businesses, including our lending businesses, into GS Bank USA,'' Goldman said in its statement yesterday."


A:

I don't know what specific assumptions Buffet / Ricciardi are using, but I believe that even under very pessimistic assumptions about future home prices and default rates MBS are currently undervalued by the market. I think the obvious explanation is fear, and I have seen the same phenomena before.

Blog Archive

Labels