« Breaking Drudge News!! | Main | Google Plays Offense - The Latest from the Net Neutrality Front »

June 23, 2007

Comments

"Because there is little trading in the securities, prices may not reflect the highest rate of mortgage delinquencies in 13 years"

So why don't people short these instruments? Because they can't get access to the market? If so, it seems anticapitalist or something.

So why don't people short these instruments? Because they can't get access to the market? If so, it seems anticapitalist or something.

Well, there's no open, standardized market for them to draw from to do shorts.

Not to get too conspiratorial, but I've heard that Wall Street firms are developing all these complicated and non-standardized derivatives because that's really the only action they can keep to themselves.

So why don't people short these instruments? Because they can't get access to the market? If so, it seems anticapitalist or something.

Access is one issue. It's a function of having the money to bear the risk as well as other things. Plus what Jonas said. You have to find someone to lend you the securities to deliver short.

One thing that is troubling is this quote from the guy at Sanford Bernstein:

"An auction that confirms concerns that CDOs are overvalued may spark a chain reaction of writedowns that causes billions of dollars in losses for everyone from hedge funds to pension funds to foreign banks."

My bolding.

Why are pension funds investing in this stuff? Let's see. You're a trustee, responsible for scads of other people's money. How in the world do you invest it in thinly traded, highly complex, instruments that you probably don't understand at all?

Why are pension funds investing in this stuff? Let's see. You're a trustee, responsible for scads of other people's money. How in the world do you invest it in thinly traded, highly complex, instruments that you probably don't understand at all?

Pressure to produce returns. If you are the manager for a fund that provides defined benefit pensions for state employees, and the fund requires an average return of 7.5% to be sound, you are under pressure to meet or exceed that return target. If you're in a state where the Legislature is required to make up the difference with tax dollars if the fund falls too far behind the target, you're under a LOT of pressure.

So if you're prudent, you've got this year's expenses in cash or equivalents, the next couple of year's worth in short-term fixed-income kinds of things, and a great big pile of money that won't be needed for from several years to decades. What do you do? You diversify, and try to put the money into a variety of uncorrelated assets. If you can show me a pension fund that has put 20% of their assets into derivatives based on sub-prime mortgages, I'll be outraged along with you. If they have 2% of their assets in such derivatives, I'm probably not offended at all.

I've long thought one of the most stressful jobs in the world must be fund manager for one of these big public pension funds. Every month, there's another $10M or so that comes in that you have to invest no matter what the market looks like.

You diversify, and try to put the money into a variety of uncorrelated assets.

Sure.

But if you don't understand the asset how do you know it's uncorrelated? It seems to me that subprime mortgages at bottom have a pretty high systematic risk. If the economy tanks, or rates rise, they're not going to do so well.

So now you're into more complicated stuff that might go up if rates rise, say. But you are probably relying on what the investment houses are telling you.

If you can show me a pension fund that has put 20% of their assets into derivatives based on sub-prime mortgages, I'll be outraged along with you. If they have 2% of their assets in such derivatives, I'm probably not offended at all.

If you can show me a manager who really understands the nature of these investments my outrage level will drop.

Remember Orange County and derivatives, anyone?

Same stuff here, different sauce.

And yeah, a lot of pension funds have stuff in CDOs and stuff called "alternative investments." Pressure quite often from BOD and companies to Get The Returns Up No Matter What, considering the shortfalls that have been occuring.

The smart ones will have invested only a few percent--it's unfortunately the ones who are under the most pressure who will have taken the riskiest positions of 10% or more.

FT has had several good articles looking at the whole Bear Sterns case. There's one major difference between the potential default pain at present vs. what happened with LTCM: amount of leverage.

I'm enjoying this conversation. Even with my limited knowledge of macro-economics, or maybe because of it, this stuff scares me as much as anything. I get a lot of my info from Ritholtz over at the Big Picture, who has been all over this issue in an informative, and pessimistic, way.
Also, I wonder if it is coincidence that I was listening to my Lesley Gore collection and The Bubble Broke was playing as I read the comments? That's what keeps life fun for me.

"Over $1.8 trillion, supposedly, in securities backed by subprime mortgages, and no one really knows how much they are worth."

Hilzoy, while you might find it difficult if not impossible to understand how mortgage securities are valued I can assure you that the people who trade them do not find it that difficult. Their problem right now is one of uncertaintly regarding the amounts that may soon be put up for sale and what UST interest rates will be six to twelve months from now.

If they knew, or had a solid indication, on these two item the valuation of the securities in question becomes easy. No one wants to be the first with a winning bid only to find a no end in sight the same product hitting the market right behind you. And if they get stuck holding inventory and they were wrong on interest rates they could lose money on their spread as well cause spreads tend to widen as rates go up.

One more additional point. These are mortgage securities made from large pools of subprime mortgage collateral. It is highly unlikely, near impossible in fact, for the collateral to become worthless. It will always have some value and that value depends, in addition to the two factors mentioned above, on the prepayment speed of the pools in question. My guess is that when these securities were orignilly priced the assumption was that prepayments would occur at standard PSA rates or some reasonably modified version therof due to the different credit of the buyer. With the option to refinance these loans now removed this probably means that the pools will have unwelcome duration extentions. The value is reduced as duration is extended on these kinds of instruments. But until traders see what the new prepayment speed may be and what the duration is they will again be cautious in stepping up to bid on large amounts of these securities.

And finally, these securities are backed by single family homes and even if the homeowner defaults it does not become worthless. So even a 100% loan on a house that goes to foreclosure and is sold at 70% of it's original purchase price is not going to lose the lender 100% of their loan.

And, to the extent these securites were AAA rated due to the loans being purchased after origination by FHA, FNMA, or FHLMC prior to securitization this all becomes academic anyway since both principal and interest are gauranteed.

The fraction of these homes that go into forclosure that are not in gauranteed pools are spread out in other pools of loans that are themselves spread out amount many different institutional investors.

I think that this whole thing is managable and is way overblown by the public.

Ken,

I can assure you that the people who trade [mortgage securities] do not find [valuation] that difficult.

Apparently, they do. Read what Hintz said. Snark aside, valuing thinly traded complex securities is tricky, and the underlying models can fail in panic situations.

Their problem right now is one of uncertaintly regarding the amounts that may soon be put up for sale and what UST interest rates will be six to twelve months from now.

But that just means they are hard to value. You never know what interest rates will be in six to twelve months. And as you yourself say, we don't have a good handle on the prepayment rates and durations. Hard to value.

Finally, you argue that the underlying collateral is sound, or at least not totally worthless. But the funds buying the loans appear to be highly leveraged. So a 30% (or even smaller) drop in the value of the mortgages is a disaster for the fund, and as we see, potentially unpleasant for the fund's lenders.

It's also not clear that the underlying collateral is sound. I mean, it's presumably sound in the sense of 'not literally worthless', but in that sense loans backed by Tulip bulbs in the Tulip mania were also sound, since a bulb is after all something, and not entirely devoid of value.

As I understand it, to make the simplest version of one of these, you take a large pool of mortgages and create securities based on varying bits of what happens to the pool. The least risky tranche might involve buying (essentially) the first year of the mortgages, most of which will almost certainly be repaid; riskier tranches involve buying pieces further out, when the risk of mortgage-holders defaulting gets higher.

Unexpectedly high rates of default will not make these tranches worthless, but it will also make them worth a lot less than one might have thought going in. If you have a highly leveraged fund, that will hurt you a lot.

Bernard, I think I could make it a bit clearer if I point out that once the information required to value the securities is in the securities are relatively easy to value.

The missing information: volume, interest rates, and prepayment speeds are all the kinds of things that working harder to obtain will not rush the information along any faster. (This is not intended to be a complete list but are, to me, the most obvious missing elements.)

These things take time. It takes is time for interest rate forcasts to reach consensus, it takes time for prepayment speeds to become clearer, it takes time to see what kind of volume is developing.

Being hard to value implies that making a greater effort to value them will hasten the solution. But this is not true.

The missing elements are not things that effort alone can determine. More than effort, what is needed is patience to wait and see what develops.

And regardig Hintz. Isn't he merely saying that right now he does not know how much of this stuff will come to the market if other funds blow up? I think that is pretty much what I am saying.

That is why many traders are waiting before offering anything but throw away bids. What they are saying is "Hey if you want a bid right now it's 10 cents on the dollar. But if you can wait a while until I get better information I can make you a better bid."

As to the leveraging of the securities you are indeed correct. That does magnify the problem for the hedge funds and for their investors. But unless that is typical of how the majority of these securities are held I do not see it as having too great a ripple effect. Once Bear Stearns tosses its stones into the pond it will no doubt make a great splash, but beyond that I do not see it having much effect.

Ken,

But some of the missing information is always missing. Yes, we may get better data on prepayment rates after a while, but the quality of interest rate forecasts is not going to improve. After all, people have been making such forecasts for a very long time, for all sorts of purposes.

Once Bear Stearns tosses its stones into the pond it will no doubt make a great splash, but beyond that I do not see it having much effect.

You may well be better-informed than I on these markets, but I don't see that as the end. The splash from Bear Stearns will drive down prices, and lenders will be moved to withdraw money from other leveraged funds, producing more splashes.

hilzoy,

Firstly the risk profile of single family home in the USA today is about a far away from a tulip bulb in 17th century Holland as one can imagine. An investor in large pools of mortgages is reliably 100% certain that the underlying collateral is not going to lose all or even most of its value. That cannot be said of tulips.

Secondly when the pools of mortgages are securitized it is the *entitlement* to the pools cash flow that is cut up into pieces (called tranches) and sold as seperate securities. The pool itself actually remains intact and is held in a trust. Only the cash flows are sold in the form of CMOs

(Pieces of mortgage pools themselves can be sold in different forms or deals called REMICs but those are not as popular today because of the long tail on mortgage collateral. You could have started with piece representing one million dollars in mortgages and ten years later end up with only several thousand dollars for the next twenty years.)

Back to CMOs. The first tranches have the most stable cash flow charactoristics and the last tranche has the least stable charactoristics and is what is leftover after stablility has been assigned to the senior tranches. It is these last tranches that are sometimes called *toxic waste* and they need to be sold before any of the other tranches can be sold otherwise the entire *deal* itself cannot get off the ground. So the amount of stabilty they are left with depends upon the market and varies from deal to deal and from year to year depending upon what is required to get the deal done.

Thirdly this toxic waste represents a small but important part of the mortgage market. It is also the part that is most often abused by unscrupulous brokers and sold to unsuspecting retail customers. But since large institutions also buy this stuff they can insist that on the *deals* they participate in that the *toxic waste* not be as toxic as the name implies. It is very complicated to explain because every deal can be different due to the market changing during the time it takes to put another deal together and bring it to market.

Needless to say the more toxicity (or instability) the last tranches can be sold with the more stable the senior tranches become.

The bottom line is that subprime loan backed securities are not a form of dark matter that people who trade in them do not understand and know how and where they fit into the overall risk profile of huge multi billion dollar pension or endowment portfolio. A few people got carried away with highly leveraged bets on some of these securities and they may end up losing.

But as always, their loss will just be someone elses gain.

ken,

What is the significance of the difference between an entitlement to a specified piece of the cash flow and the identically defined share of the pool? Ultimately somebody has to wait for the later payments to dribble in.

What you say about collateral, etc. is true of traditional CMO's. I think the concern here is that the subprime market is qualitatively different - more volatile, less well-understood, etc., not to mention the fear that its very existence is a consequence of an unsustainable bubble in real estate. So the usual considerations regarding CMO's may not apply so fully.

Now maybe the subprime market is here to stay, and a decade or more from now things will be better understood. But that doesn't negate any of today's problems.

Bernard, the difference is in how cash flows coming into the pool are distributed out among the tranches, ie. what entitlement each has.

Initially most of the cash flow to the pool is interest and only a small part is principal. Over time this ratio changes until eventually most of the cash flow is principal and a much smaller part is interest.

Now the more senior CMO tranches created out of that pool have the most stable cash flows. They may be designed to be interest only for the first several years and then begin recieving principal payment starting in year three or five or seven or whatever.

Further, depending on the deal, one or more of the senior tranches may be entitled to recieve all the principal as well as its share of the interest cash flows in sequence.

So tranche one may be entitiled to recieve all principal payments made into the pool until that tranche is fully redeemed. Then tranche two may begin recieving all principal payments, then tranche three, etc.

The formulaes used for dividing the cash flows amongst the senior tranches are limited only by imagination.

The actual design depends upon what the investor want. On one deal they may want an equal amount of principal and interest to be delivered on the senior tranches. On another deal they may want pricipal lockout for a period of years. Whatever is wanted is is how the deal will be structured.

But the design of senior tranches cannot be completed without first finding a way to dispose of the balance, or toxic waste, tranches which carry the greatest risk. All or most of the uncertainty regarding cash flow is passed off to them.

But since these most risky tranches must find a home it is the market not some nefarious scoundral that determines the amount of risk that will be placed there.

So yeah, someone must wait till the balance of the pricipal dribbles in but that is pretty much known in advance when these deals get done. It is only when this stuff gets sold to an innocent retail investor or overly greedy money manager that it becomes problem too big to handle.

Also even the most senior tranches can become unstable if the payment charactoristics of the underlying pool changes radically from the original underwriting assumptions.

In theory it is possible to make every tranche exactly equal but that is not what happens.

Bernard, now with regard to the subprime market.

The charactoristic of this market are not so radically different from the prime mortgage market. Same collateral, same requirments for fire insurance, same need for termite inspection etc.

And when mortgage insurance is tacked on deal even the borrowers credit is brought up to prime, at least as far as the lender is concerned.

I am not saying there is not a greater risk but I just do not see how it is paradigm shifting, earth shatteringly different.

And with few exceptions I think the risk has been accounted for by the markets when this stuff was originally sold.

Now markets can change and some investments can move out of favor as quickly as they move into favor. So with bad press it may become *prudent* for institutions that would normally buy this stuff to hold off for a while until calmer heads prevail.

But for those with both lots of cash on hand and the freedom to act I also think it will be a good time to pick up some primo deals at bargain prices.

I hear that the reports of the tulip crisis are wildly overblown, something I first began to suspect when I discovered that my many overeducated Dutch friends had never heard of it.

ken,

Yes, I actually knew everything in your 2:37. Also, my understanding is that the toxic tranches are often sold to retail investors who understand them poorly, so there may be a touch more nefariousness than you claim, but let that be.

My question was about your 11:24 where you attached importance to the fact that the pool remained in a trust while only "entitlements" were sold. This strikes me as a possibly useful administrative or legal device, but I don't see that it has any financial significance.

As to the riskiness of subprimes, your analysis assumes that they are just one piece of a normal mortgage market; that there is no reason to think that we ought to evaluate the risks any differently, allowing for differences in credit quality and so on. That might be right, but I think it is possible that it is wrong.

Defaults, for example, are probably more closely correlated in the subprime market than in the normal market, and are also probably more strongly correlated with economy-wide problems. Subprime borrowers are, by definition, at their limit. Even a small financial setback, caused maybe by an economic downturn, can lead to a default. In other words, the systematic risks are high.

Another possibility is that we see a lot of subprime activity at particularly dangerous times. Low short term rates make it possible to offer attractive teaser rates. Rapidly rising home prices make people - both lenders and borrowers - feel they can stretch and will be bailed out by continuing rises. They also make borrowers feel they must get on the bandwagon now or prices will run away from them.

Sometimes it works. Lots of borrowers have refinanced and gotten themselves on better footing. But it's very easy for the bubble to burst, and then there are problems. My suspicion is that rising subprime activity is a sign of a bubble, and that makes them not quite just ordinary mortgages with higher rates to compensate for added risks.

Bernard, I believe that the toxic tranches, while a relatively small part of the overall market are still too large to be comfortably put away into retail accounts and that it is still the institutional investor that drives how much risk they entail. Still, it happens all too frequently I'm sure.

You make a good point on the corrolation between subprime defaults and the overall economy. And you are right about how the various scenarious can and do influence the making and the valuation of these mortgages.

I think however, that most if not all of this had to have been anticipated and taken into account in the original pricing. The answer to this will become clearer only with the passage of time. We can speculate all we want but this is one of those situations when the best answer is just to say that 'only time will tell'.

Also, as a side note, keep in mind that a large portion of the subprime loans had mortgage insurance as a requirment on origination. This insurance transfers the credit risk from the borrower to the insurance company. So to the extent they were insured the credit was even better than that of prime mortgage borrower.

Anyway, my overall point is that this whole thing is overblown by the media and by the public who are too easily panicked by the doom and gloomers. If you drill down to what lies at the bottom of the securities we are discussing I think you will find some pretty decent collateral, not perfect but decent for the price and the amount of risk taken on.

ken,

Anyway, my overall point is that this whole thing is overblown by the media and by the public who are too easily panicked by the doom and gloomers. If you drill down to what lies at the bottom of the securities we are discussing I think you will find some pretty decent collateral, not perfect but decent for the price and the amount of risk taken on.

I hope you are right.

rilkefan,

I've heard the same about the tulips. There is a new book on the subject. I haven't read it, but I gather it argues that the situation was not as bad as the stories suggest, but that there was some excess in the tulip market.

The comments to this entry are closed.

Blog powered by Typepad