Thanks for explaining this, especially the bucket and 30-day rolling stuff.

Tanta, if you have time I would beg for an ubernerd post that sums up the history of usages of the term "ALT-A", ALT-B etc, and the whole bit about how subprime is sub-620 to some and sub-660 to others. I am still confused about all that.

"The latter will depend on "loss severity," or what percentage of the outstanding loan balance is not recovered in liquidation. That, obviously, is an issue of home prices and servicer efficiency"

IMO loss severity is where the rubber meets the road and is probably the single most wildly speculated about concern.

Really, an REO is by default going to be a significant loss, as no one will go into default if there's any salvagable equity. The degree of the severity of the loss is what will be in question and CA will finally become a Bright Red State as it crushes ALL competition to rack up losses that will be staggering, probably dwarfing the GDPs of 1/2 of the world's countries.

We're still in the second inning and it's a tie game so far.

Tanta-

thanks for another insightful post. I realize I've been conflating a couple of these terms for quite some time. Guess that makes me an inadvertent weenie.

PS - I don't mean to press, but when are we going to see your manifesto on what sort of reforms are needed for the mortgage market (especially subprime)?

I suspect parts would find their way in legislation surprisingly quickly given the political heat the issue is generating, how popular this site is becoming and the timing of the election cycle.

I think the rubber meets the road in column 3 row 1 of the table (projected defaults of current loans). Everything else can be pretty well modeled. And the difference between the loss of 35% and 50% is smaller than with default of 11% vs 30%.

When the model assumed (based on the past performance) that most of the borrowers in trouble would get current by refinance and the rising home prices would bail out almost everybody, then we have a problem when this assumption fails.

A very useful post to keep things in perspective.

Yes, even though there may be increasing numbers of people not making mortgage payments that doesn't necessarily result in the wholesale destruction of the value of these securities as there is SOME collatoral to recover.

And if a bubble didn't occur in this market of collatoral, if SO MANY didn't grossly bid prices up and then finance them at 100%......then this whole issue wouldn't be that noteworthy. I have no doubt that these would have been great, relatively safe, securities had they been based on 80% mortgages or on fully documented income....like in the "old days".....when the model's data was largely consistent, reliable and predictible.

I think the big concern is that the cycle not only is reversing but unraveling. Too many properties aren't going to catch a bid in too many locations. Too many properties will be vandalized or suffer other damage severely lessening their value. And while they sit, the tax bills keep accumulating along with other "hidden" costs of property disposal.
This is the glitch in the model, an unraveling. Unless someone knows how to get people to buy a couple of million excess units without their prices being cut dramatically (40-50%) to an "affordable" or rental equivalent level, it just doesn't seem logical that there won't be a significant (25%?) decline in the "real" value of these securities over the next year or so.

AllenM, I am equating default with a property that has had foreclosure procedings intiated. I guess it's possible that a property in default might have a tiny shred of equity, but I figure a property would be sold or refi'd by the owner if there was any money in it. Once it goes back to the lender it's a loss, for sure.

Mortgage Credit News - This Week

Pretty good commentary from Lou Barnes today.

Trying to boil the table down a little...

If the fraction of delinquencies is fd, the default rate of currents is dc, the average default rate of delinquences is dd, and loss severity is ls, then the total loss is:

ls * [ (1-fd)dc + fddd ]

The average default fraction as a percentage of all deliquencies in the example given is about 65%. I'm no mortgage lender but that sounds like a pretty reasonable number. It's hard to see that that number should be any lower than, say, 35% (the lowest on the table), and certainly can't be higher than 100%.

As you vary this average number from 35% to 100%, the loss varies from 5.2% to 9.1% (it started out at 7.1%). So varying the default rate of delinquencies over its entire range gives about 30% variation in the final answer (total loss of the pool).

The final answer is linear in the loss rate assumed (vary the loss rate by 30% and the final loss of the pool varies by 30%).

I think I agree with poszi above that the biggest question mark is the fraction of currents that eventually will default. From the numbers in the table, I think this was simply set as fd*dd, which is the steady-state solution.

If you assume a steady state solution (as I'm pretty sure Fitch did), the answer at the end of the day can be boiled down further to

ls * fd * dd *(2-fd)

which shows pretty clearly the dependence of the final loss to the pool as a function of the inputs.

The reason you need a whole table is that I've swept a lot under the rug into "dd", and the table is effectively calculating dd.

If you assume non steady state (i.e. that defaults will increase over time) the numbers look quite different and the simplest formula doesn't work.

Our actual mortgage experts can comment on whether 65% is a reasonably stable number for the default rate averaged over delinquencies.

Barely, Yal and others above touch on the importance of severity. IMO, severity is the key given the structure of leverage.

Think of the AAA tranches as withstanding a 100% default rate with a 20% price-related (ex-foreclosure costs) severity.

"Go ahead, don't pay," the AAA-guy says, "I own your home if you do."

"But don't you have to sell it?" you ask.

"Yeah," he replies, lazily glancing at his daily P&L print-out. "And I'll NEVER sell at a loss of more than 20%, because home prices don't fall, and you can take THAT to the bank."

"Gee," you say. "If home prices don't fall, and defaults don't affect you, how much leverage do you have?"

Finally he smiles; a wolfish grin. "50:1's where I stop, but only 'cause it just isn't fair for me to make ALL the money."

Imagine AAA-guy if he thought home prices would fall. He'd dump his holdings faster than you can say, "redemptions."

And that's why the sticky home price hypothesis is so important to these folks.

Pearson, That's the best one I've seen in a long time. It's a game of chicken with a locomotive and the starters' gun goes off when RE values drop... around 10% or so.

Alec, "Flagstaff should be relatively stable, as they couldn't build that much because of all the national land surrounding it. Rates haven't risen enough to really put downward pressure on prices, just demand"

Aren't you forgetting something, important, in this credit bubble market? Even places that failed to see spikes in either building or RE prices are seeing default rates rocket higher. See, those owners in fly-over-land watch the same TV ADs for $499/mo refis, get the same mailers and watch "Flip This + That" on HGTV. They didn't get the MEW vaccine. They are almost as likely to be facing resets they can't handle and should see price drops too - just perhaps not quite as severe. Depends on the leftover demand, if there is any...

Joe, lived in Northern Az for a decade- it is out of phase with Phoenix. When Phoenix starts getting closer to the bottom there will be more deals. B-I-L owns a shack in Munds park and acknowledges he should have sold last year when the party was still going. A lot of inventory is being trolled, but not much activity, even with the prices being 100k off of last year. But second homes/college housing take a while to move downward.

barely- doing a refi requires there to be enough money in the property to float the fees and that a lender is willing to do it- the latter is getting darn near impossible unless you have significant (over 30%) equity in the property that will remain intact after the refi. So, guess what? There are houses going to the block with equity, per se, but if nobody is willing to pay them for it on the courthouse steps, unrealized equity. Still, most of the houses on the steps are going reo now, unlike last year where investors were still buying quickly.

Someday this war's gonna end...

Tanta, u better be careful. if ukeep talking about the sophisticated methodoligies the rating agencies use, people might start to think they know what they're doing!!!

The best hope is that capital returns to more productive pursuits and keeps jobs loss related to the burst of the housing bubble to a minimum.

Too much capital is already tied up in RE and RE derivatives and will die there.

Thanks for the info to all about Northern Arizona.

I have lived in many places and on several continents and consider the Flagstaff area as my all-time favorite. We'll probably pull the trigger next year, regardless. I would be grateful for any more information in this regard.

Some of the readers here seem to be very gloomy. This is surprising to me, given the macroeconomic back-drop. Apart from the languishing residential housing market, the economy seems very strong; the stock market has been extremely bullish and the raising interest rates show that there is demand for loans even at higher rates, thus there are projects and developments economically viable at higher rates. This is unheard of since the 1970s. In conclusion, this seems the beginning of a long lasting and sustained uptrend in US and world economy.

Let's look together at how we can participate on this trend.

Thanks again, Joe

****WARNING****

Stay out of PA.!!!

"Restrooms in the Capitol Complex and all state-owned buildings will not be cleaned, for example, and a limited amount of maintenance personnel will only provide core services, such as chiller plant operations. The Capitol fountain will be turned off, as well as exterior lighting of the Capitol dome."

MarketWatch.com

In conclusion, this seems the beginning of a long lasting and sustained uptrend in US and world economy.

Wow, sounds just like the late 1920's!

For the record, I think the rating agencies aren't any worse than anyone else when it comes to model construction and statistical analysis.

gasp

u mean when people claim the value of their residuals is objective and accurate because they valued it using LEVELS it might not be so????

i don't know what to believe in anymore if i can't believe in LEVELS...

hesitates, then takes a long pull from a bottle of whiskey and stares forlornley out the window.

Russ Winter wrote:

Posted on 06-Jul-07 at 11:01 am

Winter (Economic and Market) Watch » Winston Wolfe Shows Up Again For Brain Tissue Clean Up

Tanta-

just when you thought you had seen it all-

Court freezes Amerifirst's assets - Dallas Business Journal:

Joe,
Alec has it right- I call Flagstaff "Poverty with a view", which is exactly what it is. We used to say you could make a small fortune in Northern Arizona, especially if you started with a large one. Major employers are the University (Lowest wages for a state school in the state.) The City/County governments, and hotel rooms. WL Gore has a bunch of plants, but they only hire friends and family. Great to retire to with money and the desire to leave for the winter- hard to make a living at with all of the retirees/students pushing down wages due to starvation.

I still have some family interests up there so I go up a couple of times a year- blech in terms of profits at times.

U guys need to remember the bigger picture...

It's not just the models and the levels, but also the leverage and the fraud. When we're not as smart as we think we are, when we're brutally dishonest with each other and when -15% = -150%... joo gotta pro-lem baby.

Excessive leverage especially, is the hauntingly similar to 1929. Well above the others, which have been more or less contant over time.

Levers can be explosive when they getted cranked too hard.

Any economist here?

http://chevallier.turgot.org/a633-The_variations_of_the_monetary_aggregates_are_only_and_single_reliable_indicators_of_the_economic_activity_.html

This guy is predicting a 2nd Q US GDP to be 0% and 1.26% YOY, based on the chnages in monetary aggregates, M2-M1 etc.

Anyone care to comment about the soundness of his methodology?

Tanta,
This article shows that as UK homes move towards foreclosure, the banks are agreeing to buy back the homes for a low price and rent them to the previous "owners" so that they don't have to move. Can this be done in the US? I would have expected laws against it as it is easy for the banking system to goose up prices and it would be advantage to do so and then rent back. What is the law here. How can it be capitalism if they don't have to sell at market prices and can instead hold up the market by hanging onto the homes? Where did mark to market go?

Central banks take aim at inflation in Europe - The New York Times

Great information for a rube like me, thanks!

I just wanted to thank everybody for their comments on Northern Arizona.

I would also urge to reconsider the bearish stance.

Thanks, Joe