Reggie Middleton is an entrepreneurial investor who guides a small team of independent analysts, engineers & developers to usher in the era of peer-to-peer capital markets.
1-212-300-5600
reggie@veritaseum.com
This is part 2 of the Municipal Bond Market and the Asset Securitization Crisis, continuing the primer listed as number 5 in the series below. It is to provide a background for the increase in stress and pressure in the monoline industry, which I believe has a strong chance of creating a CDS domino effect throughout the investment banks and other insurers. For more info on the risks and threats of the CDS market see Counterparty risk analyses – counterparty failure will open up another Pandora’s box. See "I know who's holding the $119 billion dollar bag" for a listing of the most likely candidates to suffer, as well as Banks, Brokers, & Bullsh1+ part and Banks, Brokers, & Bullsh1+ part 2 for my take on the general risks in the investment banking industry today. Reggie Middleton on the Street's Riskiest Bank - Update drills down on one of my short positions to reveal why I am bearish on Morgan Stanley - way before the sell side started yelling sell may I add, very similar to the contrarian position taken in Bear Stearns late last year.
The following municipal bond portion of the asset securitization crisis is also a tie-in to the prospects of the monoline insurance industry. The latest of my monoline analyses is the Assured Guaranty Report. You can also peruse the work I did on MBIA and Ambac starting from the inception of my short position in these companies last year, which turned to be nearly as profitable as the Bear Stearns short (see Is this the Breaking of the Bear?) instituted late last year as well, and based on the same investment thesis. A quick background of my older musings on the monoline industry:
Thus far in the Asset Securitization Crisis Series, we have:
Further building on the municipal bond default analysis Part 1, we have calculated the likely default amount on the municipal bonds issued in the last four years (2004 to 2007). We have assigned default rates on the municipal bonds for various states on the basis of property price decline and the decline in the building permits witnessed in each state. In this analysis, we have also considered defaults on the general obligation bonds (GO bonds) as the macroeconomic conditions have deteriorated and could result in increased stress on municipalities. Although historically, the GO bonds have defaulted rarely (the contribution to total default by Municipal bonds is 3.54% for GO bonds and the remaining 96.46% defaults is on Revenue bonds), declining property prices and rising foreclosures are likely to have a negative impact on municipalities’ revenues in the form of taxes.
Since we have maintained from the beginning that this crisis is far worse than any crisis that the US economy has witnessed for close to half a century, our underlying assumption while calculating the default probabilities by GO and Revenue bonds has been a premium over historical default rates on the munis for the period 1979-97. This premium is dependent on the degree of decline in housing prices, building permits and the broader infrastructure investment. In the case of Revenue bonds, the multiple has been considered higher as compared to GO bonds since historically; Revenue bonds have defaulted more than the GO bonds.
House price decline |
Building permits decline |
Premium over historical defaults forRevenue bonds |
Premium over historical defaults forGO bonds |
-5% |
-10% |
1x |
1x |
-10% |
-20% |
2x |
1.5x |
-15% |
-30% |
3x |
2.0x |
> -15% |
> -30% |
4x |
2.5x |
We have calculated the likely defaults on municipal bonds issued since the year 2004 since this is the period where most US state and local governments had prepared budgets based on the existing real estate boom. In addition, the prevailing low interest rate environment was very conducive for muni bond issuance. However, with the collapse of the housing market, property values went down and increasing numbers of homeowners applied for the property revaluation to reduce their property tax burdens. This increased the burden on the respective municipalities, as homeowners, in an attempt to mitigate the increase of their financial obligations obtained during the housing boom equity spending spree, cut corners by any means necessary. Construction permits and the associated fee income dropped precipitously, further constricting the bloated budgets of municipalities who, like the fabled subprime refinancing, SUV driving 1st time homeowner binged on easy equity-sourced cash.
Additional strains in the revenue sourcing for municipalities are the rampant foreclosure rate increases and the actual volumes of foreclosures. Up until the event of actual foreclosure, property taxes are usually not paid, further hampering the cash flows of municipalities that relied on these funds. It gets worse. Even after foreclosure, and even on behalf of the municipality, the back taxes cannot be monetized and actually paid until the property is sold. Many auctions in high foreclosure areas are seeing properties with no bid at the upset price. This portends very bad things for the banks, the municipalities and the insurers who wrote insurance to cover them!
These developments are likely to have a severe negative impact on the tax inflow for the state and local governments which forms the basis of our underlying assumptions. According to our estimates, on the total municipal bond issuance of US$1.6 trillion in the year 2004-07, the potential losses due to defaults will be US$22.8 billion or a default rate of 1.44% with Revenue bonds contributing majority of the default amount of US$22.5 billion while GO bonds account for US$304 million. This indicates a default rate of 2.12% for the Revenue bonds and 0.06% for GO bonds.
Multi family housing and healthcare led the defaults in municipal bonds
Historically, housing and healthcare sectors have been the biggest contributor to municipal defaults (after the industrial development bonds which account for 24% of defaults). In housing, the multifamily segment has recorded maximum losses accounting for 19.3% of total defaults while single family accounts for 1.1%. In the healthcare sector, hospitals and nursing homes accounted for majority of defaults of 7.5% and 7.1%, respectively.
Default rates in municipal bonds have varied significantly across the subsectors. The defaults in the tax-backed, water/sewer, and other plain vanilla municipal bonds has been significantly low. According to Fitch Ratings (the only of the big ratings agencies that can garner even the slightest modicum of respect these days), the cumulative default rates on such bonds have been less than 0.26%. However, default rates in municipal bonds issued on behalf of corporations or municipal entities were significantly higher. Historically, the cumulative default rates were 14.9% for industrial development bonds, 4.9% for multifamily housing, and 2.6% for health care.
Industrial development bonds, multifamily housing and healthcare sector’s accounted for 8% of total bond issuance and 56% of total defaults while education and general purpose sectors accounted for 46% of issuance and 13% of defaults.
Multifamily housing defaults peaked in 1991 as the US economy went into recession and then again in 1998 the defaults soared as the economy faced stress during that period with the collapse of LTCM and the subsequent dotcom bubble burst. This denotes difficult market conditions manifested as defaults, as witnessed in 1991 when multifamily defaults soared. The current housing market scenario featuring historically severe declining home prices and home sales and rising foreclosure rates portends an equally historic hit to the multifamily housing segment that could witness a surge in defaults that this country has not seen thus far.
Multi family housing defaults (by year of default)
Year |
No. of Defaults |
DefaultedAmount (US$ million) |
Avg. Time to Default |
Rated entity |
Non-Rated entity |
1990 |
13 |
94 |
60 |
2 |
11 |
1991 |
33 |
1,359 |
55 |
24 |
9 |
1992 |
17 |
200 |
66 |
11 |
6 |
1993 |
17 |
85 |
67 |
3 |
14 |
1994 |
5 |
41 |
82 |
2 |
3 |
1995 |
5 |
33 |
88 |
1 |
4 |
1996 |
9 |
39 |
97 |
2 |
7 |
1997 |
10 |
44 |
64 |
0 |
10 |
1998 |
26 |
102 |
55 |
6 |
20 |
1999 |
18 |
52 |
56 |
0 |
18 |
Totals |
153 |
2,050 |
63 |
51 |
102 |
In the multifamily housing segment, default rates increased significantly and were extremely high for the period 1987-90, i.e. at the time of the S&L crisis when real estate lending was reckless due to declining lending standards by banks and other financial institutions. The default rate peaked in 1988 in the eleven year period reviewed to 4.31%, followed by 3.41% in 1989. However, the overall default rate for multifamily housing sector is 1.11% for the 11 year period (1987-1997).
Multifamily housing default rates by year of issuance (1987-1997) |
||||||
Issuance year |
No. of defaults |
Total issues |
Default rate (by no of issues) |
Defaulted amount (US$ million) |
Total amount issued (US$ million |
Default rate by amount |
1987 |
7 |
182 |
3.85% |
50 |
2,961 |
1.70% |
1988 |
14 |
202 |
6.93% |
137 |
3,180 |
4.31% |
1989 |
16 |
237 |
6.75% |
106 |
3,110 |
3.41% |
1990 |
8 |
240 |
3.33% |
64 |
3,062 |
2.09% |
1991 |
2 |
277 |
0.72% |
23 |
3,561 |
0.63% |
1992 |
6 |
400 |
1.50% |
33 |
5,733 |
0.58% |
1993 |
13 |
517 |
2.51% |
68 |
6,614 |
1.03% |
1994 |
7 |
501 |
1.40% |
16 |
4,930 |
0.33% |
1995 |
9 |
603 |
1.49% |
27 |
6,132 |
0.44% |
1996 |
10 |
607 |
1.65% |
35 |
6,638 |
0.52% |
1997 |
5 |
606 |
0.83% |
9 |
5,412 |
0.17% |
Total |
97 |
4372 |
2.22% |
568 |
51,333 |
1.11% |
In the healthcare sector nursing homes accounted for majority of defaults at approximately 49%, followed by the retirement sector (29%) and hospitals (12%) based on the number of defaulted issues. However, on the basis of default dollar amount, retirement and hospitals take the lead over the nursing subsector.
Healthcare sector default break up |
||||
Subsector |
number of defaults |
default amount |
Default rate(%) based on number |
Default rate(%) based on amount |
Nursing |
116 |
454 |
48.5% |
22.8% |
Retirement |
70 |
568 |
29.3% |
28.5% |
Hospitals |
29 |
546 |
12.1% |
27.4% |
Other |
24 |
426 |
10.0% |
21.4% |
Total |
239 |
1,994 |
100.0% |
100.0% |
Total Health care sector number of defaults and default amounts
Year |
No. of Defaults |
Defaulted Amounts US$ million |
Avg. Time to Default |
Rated entity |
Non-rated entity |
1990 |
40 |
299 |
51 |
3 |
37 |
1991 |
31 |
128 |
67 |
4 |
27 |
1992 |
18 |
224 |
56 |
5 |
13 |
1993 |
18 |
220 |
57 |
0 |
18 |
1994 |
8 |
33 |
52 |
1 |
7 |
1995 |
13 |
40 |
70 |
0 |
13 |
1996 |
17 |
218 |
77 |
3 |
14 |
1997 |
17 |
112 |
77 |
0 |
17 |
1998 |
18 |
152 |
47 |
0 |
18 |
1999 |
59 |
567 |
48 |
8 |
51 |
Total |
239 |
1,994 |
58 |
24 |
215 |
The healthcare sector has witnessed significant stress during the 1990-91 period resulting in increased number of defaults owing to recession in the US. The bond defaults of the Graduate Health System accounts and the Michigan Healthcare Corporation resulted in higher defaults in 1990 period.The healthcare sector again witnessed increased strain in 1999 driven by increased failures on the part of nursing homes and hospitals. In May 1999, Greater Southeast Healthcare System filed for bankruptcy protection and suspended payments on its $46 million of outstanding bonds. In addition, the bonds issued for Graduate Health System in the Philadelphia-area accounted for $155.94 million of the defaulted amount in 1999. The healthcare sector witnessed significant strain across all its subsectors which saw the default amount rose to very high levels of US$567 million.
Significant decline in building permits has been witnessed in California, Florida, Alaska, Arizona, Georgia, Nevada, Michigan, Minnesota, Ohio and Oklahoma in the last few years. Most of the US states are witnessing the decline in building permits as the construction sector strives to stay afloat in the midst of what predict to be the worst US housing crisis, period. Up until March 2008, the building permits have declined at a rate significantly higher than that witnessed in earlier years with states such as Arizona, California, District of Columbia, Florida, Georgia, Idaho, Illinois, Kentucky, Massachusetts, Michigan and Minnesota witnessing 50% or more declines in building permits. The building permits have declined at a CAGR of 19.5% in the US in the last two years (2005-2007). In the state of California, Riverside, San Bernardino, Ontario, Sacramento and Yuba City are witnessing significant decline in building permits. In Arizona, Phoenix, Mesa Scottsdale and Tucson are the areas with significant decline in building permits. In Florida, Orlando, Punta Gorda, Sarasota, Brandenton, Venice and Cape Coral are the areas witnessing decline in construction activities. To make matters worse, declining property prices in these states are likely to hit the states’ finances. According to S&P Case-Shiller index, property prices declined 2.6% in February 2008 from a month earlier, after a 2.4% decline in January 2008.
* 2008 (A) building permits represent the YTD March 2008 annualized figure.
Building permits change |
|||||
States |
YTD2008 |
2007 |
2006 |
2005 |
2004 |
Alabama |
-28% |
-19% |
5% |
12% |
23% |
Alaska |
-50% |
-38% |
-5% |
-8% |
-11% |
Arizona |
-55% |
-24% |
-28% |
0% |
21% |
Arkansas |
-24% |
-21% |
-23% |
13% |
7% |
California |
-49% |
-31% |
-22% |
-1% |
8% |
Colorado |
-30% |
-23% |
-16% |
-1% |
18% |
Connecticut |
-10% |
-16% |
-22% |
0% |
13% |
Delaware |
-30% |
-19% |
-21% |
4% |
1% |
District of Columbia |
-82% |
-9% |
-26% |
48% |
36% |
Florida |
-44% |
-50% |
-29% |
12% |
20% |
Georgia |
-52% |
-30% |
-5% |
1% |
12% |
Guam |
|||||
Hawaii |
-51% |
-7% |
-23% |
9% |
24% |
Idaho |
-51% |
-29% |
-21% |
19% |
20% |
Illinois |
-55% |
-27% |
-12% |
12% |
-4% |
Indiana |
-40% |
-18% |
-24% |
-2% |
0% |
Iowa |
-32% |
-16% |
-20% |
3% |
2% |
Kansas |
-6% |
-22% |
4% |
6% |
-12% |
Kentucky |
-47% |
-10% |
-21% |
-6% |
11% |
Louisiana |
-6% |
-18% |
26% |
-1% |
3% |
Maine |
-39% |
-19% |
-19% |
2% |
11% |
Maryland |
-24% |
-20% |
-23% |
10% |
-8% |
Massachusetts |
-50% |
-22% |
-20% |
9% |
11% |
Michigan |
-44% |
-39% |
-36% |
-17% |
1% |
Minnesota |
-48% |
-32% |
-28% |
-13% |
0% |
Mississippi |
-28% |
1% |
24% |
-8% |
21% |
Missouri |
-36% |
-26% |
-12% |
1% |
12% |
Montana |
-39% |
-9% |
-5% |
-3% |
32% |
Nebraska |
6% |
-8% |
-17% |
-9% |
6% |
Nevada |
-49% |
-31% |
-17% |
7% |
3% |
New Hampshire |
-23% |
-20% |
-25% |
-12% |
0% |
New Jersey |
-8% |
-26% |
-11% |
7% |
9% |
New Mexico |
-27% |
-32% |
-4% |
13% |
-9% |
New York |
-39% |
-1% |
-12% |
16% |
8% |
North Carolina |
-30% |
-14% |
2% |
5% |
17% |
North Dakota |
3% |
-5% |
-13% |
0% |
8% |
Ohio |
-37% |
-21% |
-28% |
-8% |
-3% |
Oklahoma |
-26% |
-7% |
-14% |
8% |
14% |
Oregon |
-49% |
-21% |
-14% |
14% |
9% |
Pennsylvania |
-25% |
-14% |
-12% |
-10% |
5% |
Puerto Rico |
|||||
Rhode Island |
-56% |
-18% |
-16% |
12% |
11% |
South Carolina |
-36% |
-20% |
-6% |
25% |
13% |
South Dakota |
-38% |
-4% |
-7% |
-3% |
17% |
Tennessee |
-37% |
-19% |
-1% |
4% |
19% |
Texas |
-19% |
-18% |
3% |
12% |
7% |
Utah |
-54% |
-22% |
-7% |
15% |
8% |
Vermont |
-38% |
-22% |
-10% |
-19% |
26% |
Virginia |
-12% |
-20% |
-22% |
-3% |
13% |
Virgin Islands |
|||||
Washington |
-41% |
-5% |
-6% |
6% |
17% |
West Virginia |
1% |
-15% |
-8% |
7% |
11% |
Wisconsin |
-33% |
-20% |
-23% |
-12% |
-2% |
Wyoming |
-29% |
29% |
-12% |
21% |
18% |
Looking at this through a graphic representation puts things into perspective. Notice how far below the zero line this graph is in just one year!
The property prices in the US continue to witness significant decline since reaching record levels in 2005. The distressed home owner, having leveraged his or her appreciated home value during the peak of a bubble combined with deteriorating mortgage credit quality (increased Alt-A and subprime mortgages), has increased the intensity of the current crisis. The decline in US home sales has dwarfed the previous housing declines as conditions continue to worsen, depicted by declining building permits and home prices. Rising housing inventories, increased foreclosure, declining home prices, and the correction in home sales volumes are likely to make matters much worse.
US home prices continued thier downward foray as reported by Office of Federal Housing Enterprise Oversight (OFHEO) through a reported decline of 1.7% in 1Q 2008 as compared to 4Q 2007. The fall surpassed the 1.4% sequential decline witnessed in 4Q 2007. The OFHEO’s Index recorded its largest y-o-y decline in the last 17 years with a drop of 3.1% in 1Q 2008. California, Nevada and Florida witnessed the steepest y-o-y declines in home prices. The states of California, Nevada, Florida, California led the fall depreciating by 10.6% followed by Nevada 10.3%, Florida 8.1%. The metropolitan areas witnessing steepest decline are Merced (down 24.7%), followed by Stockton (down 21.5%) and Modesto (down 21.0%), all in California. According to the OFHEO, 43 states witnessed price decline in 1Q 2008 wherein in the prices fell by more than 3% in eight states and more than 8% in two states, namely California and Florida. California, which derives 22% of state revenue, including both state and local government from property taxes continues to witnessing significant decline in building permits and housing prices. It is worth noting that California was one of the states that created significant budget bloat during the boom times of the RE bubble, which serves to further exacerbate the problems at hand.
Housing starts as well as the existing and new home sales have declined significantly since the beginning of 2005 across the country with the West and Midwest states the most severely affected.
New home sales (% change) |
||||
Year |
North EAST |
MID WEST |
SOUTH |
WEST |
2004 |
||||
2005 |
-2% |
-2% |
14% |
3% |
2006 |
-22% |
-21% |
-12% |
-25% |
2007 |
3% |
-27% |
-26% |
-32% |
Existing home sales (% change) |
||||
Year |
North EAST |
MID WEST |
SOUTH |
WEST |
2004 |
||||
2005 |
2% |
2% |
6% |
3% |
2006 |
-6% |
-7% |
-4% |
-17% |
2007 |
-8% |
-10% |
-13% |
-20% |
In addition the annual housing starts have declined significantly across the country. In 2007, the overall US housing starts have declined 24.8%, followed by a 12.9% fall in 2006.
Total Annual Starts |
2002 |
2003 |
2004 |
2005 |
2006 |
2007 |
United States |
1705 |
1848 |
1956 |
2068 |
1801 |
1355 |
% increase/(decrease) |
6.4% |
8.4% |
5.8% |
5.7% |
-12.9% |
-24.8% |
Northeast |
159 |
163 |
176 |
190 |
167 |
143 |
% increase/(decrease) |
6.7% |
2.5% |
8.0% |
8.0% |
-12.1% |
-14.4% |
Midwest |
350 |
374 |
355 |
358 |
279 |
210 |
% increase/(decrease) |
6.1% |
6.9% |
-5.1% |
0.8% |
-22.1% |
-24.7% |
South |
781 |
838 |
908 |
996 |
911 |
682 |
% increase/(decrease) |
6.7% |
7.3% |
8.4% |
9.7% |
-8.5% |
-25.1% |
East |
416 |
473 |
516 |
525 |
444 |
320 |
% increase/(decrease) |
6.4% |
13.7% |
9.1% |
1.7% |
-15.4% |
-27.9% |
These statistics fully support my assertion that as the current market situation continues to worsen, the potential losses and defaults on municipal bonds could be – and probably will be, much higher than estimated by the majority. This portends considerably more stress than anticipated by the stress testing models employed by the monolines and the rating agencies that rate them. It is not as if I have had much confidence in their predictive ability to gauge risk, considering how bad they have performed in the mortgage sector over the last two years and the public admittance of software glitches and algorithms that assumed perpetual house price appreciation (that is housing prices that go up forever without even one downturn, not to mention a protracted downturn). This analysis should be read against the backdrop of:
The Asset Securitization Crisis Series
The Commercial Real Estate Crash Series
and the Riskiest Bank on the Street Update.
Next in this series is an overview of leverage and risk of the 32 Banks in Deep Doo-Doo.
Reggie Middleton is an entrepreneurial investor who guides a small team of independent analysts, engineers & developers to usher in the era of peer-to-peer capital markets.
1-212-300-5600
reggie@veritaseum.com