Kash, the first graph is the same as the 2nd.
Best Wishes.
Oops. Thanks, CR.
But luckily we are so "resilient" (see if that little bump for 9/11 doesn't say as much) and our risks are so remediated, repatriated, restructured, and altogether deriskified that economics is on the verge of extinction.
I refuse to worry with such lousy risk premiums, you?
Now if I could obtain those Greenspan speaking engagement fees, my worry oratory performance could reach operatic heights.
If the RE loans that are on the bank's books were originated by the bank, wouldn't they be the "good" risks? The bank would certainly be able to differentiate.
With credit spreads giving very narrow risk premimums, the banks could keep the good and sell the bad for almost good prices. So why would they keep the bad loans on the books?
But hasn't a good percentage of the 6.1 Trillion in loans been sold into the market. Won't this provide some cushion?
Regarding comments about banks selling off loans: Certainly they've sold off some loans. But the data shown here is what remains on their books - i.e. what they have not sold off. It's possible that banks have sold off more loans of lower quality and kept more of higher quality, and they would certainly argue that they have. But (a) I would like some evidence before I agree that banks are smarter than other players in the credit markets (and thus better able to nicely separate out the bad loans from good), and (b) those bad loans are still out there somewhere, so the credit markets as a whole will still have to absorb the write-offs.
If those figures are only on balance sheet, then the true situation could be worse, not better. Selling off a loan doesn't always eliminate exposure - some loans are sold with recourse, most loans are sold with exposure to early payment defaults, many loans are sold as tranched CDOs with the originator retaining the original exposure, and all loan sales are subject to the risk of being put back for fraud. Any idea what the off balance sheet exposure of these outfits might be?
Most of the real estate loans on banks' books are NOT home mortgages. They are typically securitized and sold. Mort_fin is right about recourse issues, but my impression is that the independent mortgage brokers have been more active in the subprime world than the banks. The res mortgages held by banks are most often non-conforming ARMs, which is bad news because of the delinquency problems with ARMs.
So what else is in real estate? Lots of development loans, in which a developer buys dirt, gets a bank loan, then will pay off the loan when the house/office/store is sold. There's certainly risk here if prices fall, but it's a very different situation than the subprime problem.
Banks will also take real estate as collateral for what is really an operating line for a business, under the "abundance of caution" idea, and I believe these are classified as real estate loans, though they are really C&I.
Bill: I disagree with you about one of your points. Looking at the Fed data, it looks like about 2/3 of banks' real estate loan portfolios are indeed residential, and only about 1/3 commercial. Furthermore, banks have another several hundred billion dollars of mortgage-backed securities on their books, so even though they've sold off individual mortgages, they've bought back bundles of them. So I'm not persuaded that their exposure to bad residential mortgages is so small.
You might have a look at this post at Naked Capitalism, which looks at Kash's post, as well as a WSJ article, and comes to a different conclusion.
http://www.nakedcapitalism.com/2...nt-of-
real.html
Whose portfolio is holding all these securitized mortgages that the banks and brokers have off-loaded from their balance sheets? If it pension plans, then how will they be recapitalized? What is the funding status of the Pension Benefits Guaranty Corp (PBGC)? How big of a hole will the Federal taxpayer be asked to fill? If it is state and local governments, how high will taxes have to be raised to make good on underfunding?
The second plot shows a pretty wide variation in the fraction of non-performing loan percentage that gets written off. For instance, in mar-94 it looks like about 33% of the nonperforming loan balance was written off; in early '06 it looks more like 70%.
A question born of pure ignorance from a non-finance person. Is there a simple explanation for why it varies so widely? For instance is there a lag for write-offs to catch up with nonperforming? (It doesn't look like it from the plot.)
And whatever that explanation is, does it say anything about what that write-off fraction is likely to be this time around?
Kash, Credit Suisse had a report that should give a preliminary picture of who bought collaterized securities. It's Exhibit 5 in the March 12 report by Ivy Zelman et al. The link, and credit for the Roubini-poster who provided it is on my site.
The exhibit is not dollar-weighted or weighted according to subprimes and other exotics, and there may have been resales, repackaging etc. However, it suggests that foreign governments and investors are holding ca. a quarter of mortgages. Fannie and Freddie are going to have to eat huge losses, as will commercial banks and maybe insurance agencies. There may be spillover into "safe" instruments, like CDs and money market accounts, but it should be minor.
Exhibit 16 shows outsized purchases of subprime in RI, MS, and HI for 2005, which could give a clue as to who is holding the hot potato. Exhibit 21 shows Fannie and Freddie holding ca. $150B in very risky subprimes. There's another ca. $170 B in interest only non-agency subprimes and slightly more in alt-As.
I have finally revisited all figures on mortgage. Only 1/3 of 9500 billions home mortage is held directly by banks. that is aboout 3500 bn. suppose a loss on this amount of 3 % that represents 105 bn. A lot of money compare to the net profit of the industry (150 bn in 2006) but nothing compared to their base capita (above 1000 bn). The balance is securitized among private investors (2300 bns) and agencies ( 3752 bns). Suppose a loss of 5 % on the value of these securties on average that is about 305 bn. a big amount but nothing compared to the loss of value of the Wall sreet market cap since February. Economically speaking it is more an investor problem than a banking problem.
Having said that they are 2 points to be made. 1) Agencies who hold these MBS are not very solid, and then comes the question "who will bail out them ?" chinese again ? 2) banks who have transferred the risk to the investors have re-created risks by lending to hedge funds investing in CDO's who are investing in MBS instead of increasing their treasury bond holdings. but here opacity is the rule. my conclusion : it is not going to be Japan made in USA. it is going to be quite different.
The changes in what could be deducted from your taxes, removing consumer interest as a deductible, has encouraged the middle class and working class to borrow against their biggest asset, their homes, and thus got the creditors around the problem of state laws that protected the homes of borrowers when they declared bankruptcy and allowing them to foreclose on this property to satisfy outstanding debts. Now these changes are becoming a threat to our economy. The Republican {big business} constant call for "deregulation" has returned us to the wild west conditions that prevailed during the 1920s. Economists should be required to also be historians.
Hey, none of this is a problem. When the crunch comes, the Fed'll just open the floodgates and cut the interest rates down into Japanese central bank territory. Right? Right?
The other line you need on the graph is total bank reserves for loan losses. After the '91 CRE problems, many of the banks which survived carried a large reserve position. The OCC eventually thought too high and felt banks were using the reserves to smooth earnings (what else are they for in good times!). As a result, the anounts the banks have in reserves for loan losses needs to be a line on the graph.
If the reserves are relatively high as a ratio to total NPLs, the banks will need to provision less. As the ratio declines, more provision will hurt earnings and capital which can be recession inducing as banks can make fewer loans.
CRE is a particular problem because it tends to have a long half-life. The bank will recoup some of the value shortfall, but it tends to take a lot longer than wringing value out of C&I loans.
agree with kash on this one: banks have another several hundred billion dollars of mortgage-backed securities on their books, so even though they've sold off individual mortgages, they've bought back bundles of them.
This site contains a very good information, and here is another similar site which is having related information,for more log on to the website.
Bad Credit
This site contains a very good information, and here is another similar site which is having related information,for more log on to the website.Bad Credit
wow thats a nice info about business loans.
loans
Not all of that $170 bn will be lost to banks, since they should be able to foreclose and sell some of the underlying properties, but it seems safe to guess that banks will end up writing off some fraction of that total - probably many tens of billions of dollars worth of loans, or several tenths of a percent of all bank loans.
Thanks for the info...
Commenting by HaloScan
