Archive for the ‘housing mess’ Category

Why is the municipal market selling at such a premium?  I have been asked this question several times in the last few weeks — not by career municipal analysts at mutual funds or rating agencies, but sophisticated investors who are trying to make sense of the asset class.  The counterpoint is coming from writers forecasting the collapse of municipal bonds.  How do we reconcile this disconnect?  For one, the supply/demand dynamics have changed significantly since a year ago.  Aside from fear of tax increases propelling more buyers into the tax exempt shelter, there is another factor: BABs.  The Build America Bond program has grown to $97 billion, as of the end of April, according to Treasury’s May 12 release — or 20% of the municipal market since the program’s inception in April 2009.  (BABs are taxable municipal bonds; the borrower receives a 35% reimbursement from the federal government – on the theory that this creates a neutral rate compared with tax exempt bonds.)  But this 20% understates the recent trend.   A closer look shows that BABs issuance in recent months hovered around 25% of the new issue market.  All taxable municipal bonds issued in 2010 through April amounted to a whopping 33% of the market according to the Bond Buyer market statistics.  These changes have expanded the buyer base for municipals to those wanting long term public sector debt, but not affected by the US tax code, such as foreign buyers, pension funds and certain corporate investors — limiting supply of paper for the traditional  tax exempt investor.  This is not to mention that most BABs are structured at the long end of the curve, squeezing tax exempts into the shorter end — since most issuers will structure both into a single offering.  (Except of course for Illinois, which prohibits a mixed structure –hunh?)   The pattern is likely to continue in the foreseeable future as Congress extends the taxable BAB structure, although market dynamics may shift modestly when the subsidy is reduced to a more logical 28%.

———————

So how’s the economy?  The recent Bureau of Labor Statistics jobs report was encouraging.  We created 290,000 jobs last month.   Unemployment notched up to 9.9%, but this is to be expected at the early stages of recovery.  As signs of recovery emerge, more people enter the job market, and if that’s proportionately more than new job creation, the unemployment rate will go up.  (The unemployment rate expresses the ratio between the number of jobs and number of people counted as part of the labor market.  So if everyone stopped looking for work, unemployment would go down.)  In fact, 805,000 people entered the labor market last month far more than the number of jobs created. 

Not to be a downer, but many of the new jobs are temporary hires for the decennial Census count.  According to the BLS, the federal government employed 154,000 people for the census count as of April, an increase of 66,000 over the prior month.  This increase followed hiring of 48,000 in February.  Don’t get me wrong, these jobs will infuse spending into the economy, ease unemployment for many and get an important count accomplished.  But they are temporary jobs that will peel off around the same time that other federal stimulus programs wind down. 

Another factor slowing down the recovery is the lack of migration.  The housing mess is a contributing factor.  We noted in a previous post the first-time reversal of migration patterns since those statistics were collected.  We have always been a nation of restless movers – opportunity seekers since the founding of the U.S.     In large part this has contributed to the active municipal bond market for infrastructure growth and development.  As jobs moved from north to south, east to west, city to suburb, people and development followed.   In the current economy, this trend has reversed in many places. 

William Frey, the noted demographer, stated in a recent report for the Brookings Institution:

The detailed 2010 census results won’t be available for another year. But this week (back in March, ed.) the Census Bureau unveiled its latest population estimates for metropolitan areas and counties for the year ending last July. What they show is a country that is demographically standing still.

Last week, the Census bureau reported an uptick in the migration rate in 2009, from 11.9% to 12.5%.  But the majority of movers went from one county to another within the same state while job moves are typically inter-state.  Further, renters moved at five times the rate of homeowners.   Homeowners, already battered by the housing downturn are finding it difficult to move to better jobs (or jobs at all) when they cannot sell their homes.

We have two opposing camps in the muni-market at the moment: those who say it is the next systemic shoe to drop and the rating agencies that are systemically raising ratings.

Which is right?

We have moved from a market that has had heavy intermediation from the bond insurance companies to one where investors are on their own. Bond insurers historically performed heavy analysis and monitoring of the credits they insure and actively affect the legal protections embedded in a security. Bank credit enhancers have done the same, although in some cases their agreements allow them out of the risk under certain circumstances. Rating agencies assign their rating on sale and their business model has never been designed to provide robust monitoring of issuers in the secondary market. The Moodys and other agency default studies covered this time period – a relatively prosperous period from 1970 – 2007 when pending defaults on the investment grade level were mostly managed by bond insurers. Further, the default studies did not include insured bonds.

Municipal bonds do default and have throughout this period –not at all in large numbers by any comparison to the corporate world. Knowledgeable high yield investors are aware the risks in hospital, senior care, land development, and corporate-style tax exempts such as industrial revenue bonds. Project finance, such as the Harrisburg incinerator, Las Vegas monorail and Connector 2000 also carry high risk. (I should add WPPSS and other nuclear power projects to this list.)  These should not be lumped together with the other 30,000 or so municipal bond issuers. Like the stock market, there are many, many nuances among borrowers. Only now the investor is left to grapple with his own resources – and the loss of granularity among this diverse and deep market from ratings “re-calibration”.

(Concerning re-calibration of ratings:  As I have said before, lowering corporate ratings to their appropriate probability of default relative to municipals – would achieve the same effect, and be truer to their own research. Instead, the agencies are compressing ratings into fewer categories and eliminating much granularity for investors. But of course the other approach would have been business suicide and harmed many corporate portfolios.)

Within this, market prophets are now generalizing greatly, mostly those who have not spent their careers in this non-standard, messy, sometimes corrupt and poorly understood state and local financial world.

This brings me to some comments on Rick Bookstaber’s recent post. Many of his points are strong and need to be actively debated, but others generalize to the detriment of this very varied marketplace.

Leverage and Opacity. Leverage in the municipal market comes from making future obligations to employees in order to pay them less now. This is borrowing in the form of high pension benefits and post-retirement health care, but borrowing nonetheless. Put another way, in taking lower pay today, the employees have lent money to the municipality, with that money to be repaid via their retirement benefits. The opaqueness comes from the methods of reporting. For example, municipalities are not held to the same standards as corporations in their disclosure.

Agreed. The Tower Amendment should be abolished and municipalities that want to participate in public capital markets and be rated like corporates should be held to the same reporting standard. Period. Would you buy a corporate bond whose disclosure was two years old?

The argument that it’s too expensive and too burdensome for smaller municipalities to achieve transparency can and must be resolved. Technology and know-how exist to achieve this in a cost effective manner. Those municipalities that cannot produce timely and accurate reporting – should borrow locally and not expect to participate in national markets. Or they should not be given (potentially misleading) ratings. Borrowing locally will likely end up costing more than paying the auditor to get the report done in a timely manner.  Hiding behind the “municipals don’t default, so what difference does it make?” argument is so yesterday.

The rating agencies still do not fully incorporate pension and other benefits into the debt statement that produces their debt ratios and medians. Taxpayers certainly do consider these factors. All obligations that are paid from the public purse should be clearly disclosed. Understandably, rating agencies argue that they serve investors not taxpayers. But political risk – unwillingness to pay – clearly reaches back into the investor’s pocket.

Size and potential systemic effects. That this is a big market in the credit space goes without saying.

This is a big market, but is not uniformly systemic like housing.

Diversification. Geographic diversification would give a lot more comfort for municipals if it hadn’t just failed for the housing market. Think of why housing breached the regional barriers. It was because similar methods of leveraging were being employed through the country. So the question to ask is: Are there common sorts of strategies being applied in municipalities across the nation?

In some cases. Given our federal system, each state has its own set of rules for local municipal finance. For example financing infrastructure for housing and economic development is vastly different in structure and credit quality in California (Mello Roos for example) vs. Florida (community development districts) vs. Texas (municipal utility districts).

Bookstaber’s point does apply to the municipal market in several ways. Investment portfolios and cash management are prey to marketers of the security “du jour”. As one of his commenters suggests, Fannies and Freddies could put municipal investments systemically at risk of a federal policy change — affecting many municipal government investors at one time. I would add to this list: securities lending practices among pension plans and other large public funds  (which some practice but others do not).

Ratings triggers on counterparties that create terminations, unwinds and a change in interest rates, basically legal provisions that automatically change the terms of the deal — also create across the board risk for those borrowers involved. Examples include variable rate securities, swap transactions, LOCs and GICs. These structures effectively embed corporate and market risk into municipal credit.

At the traditional fixed income municipal debt level however, there is significant diversification among security types and legal structures – given the 10th amendment and individual state peculiarities.

Gross versus net exposure. The leverage for municipals is not easy to see. It might appear to be lower than it really is because many, including rating agencies, look at the unfunded portion of these liabilities. They ignore the fact that these promised payments are covered using risky portfolios. And not just risky — the portfolio might apply hefty (a.k.a. unrealistic) actuarial assumptions of asset growth.

Agreed. Analysts and investors should incorporate a haircut for unfunded pension and benefit liabilities that use overly ambitious earnings as a discount rate. The debate about “risk free” vs. a historical earnings rates is an important one. Stated another way, this debate points to the fact that public pensions (defined benefit) are irrevocable, guaranteed and must be paid.  How would you invest funds when you could not afford to lose a penny vs. money you are willing to take on more risk? Since pension obligations are long-lived, the counter-argument that earnings over time should be used has some merit as well.   But the inequities between public retirees and the rest have piqued the taxpayer whose 401K and life savings have been subject to market losses.  Also, the “defined benefit” retirees are facing off against the “defined contribution” beneficiaries.  Fairness is an important public good, as difficult as this debate may be.   

Rating agencies. In terms of the work of the rating agencies, here are two questions to ask. First, list the last time they did an on-site exam of the municipalities they are rating. Second, are they looking at the potential mismatch between assets and liabilities, or simply at the net – the underfunded portion of the portfolio.

Absolutely. Maybe they don’t have to do a site visit on each review, but they should disclose the date of their latest full review for each rating, no question. How many of the re-calibrated ratings have been freshly reviewed?

Defaults. Municipalities are not quite as numerous as homeowners, but there certainly are a lot of them. And they have the same issues as homeowners. Granted, they will not pour cement down the toilet before walking away. But they have a potentially equally irrational group – the local taxpayers – to deal with.

Disagree. See my comments above and other blog posts on this site. Municipalities do not behave the same as homeowners. Clearly there are scoundrels and irresponsible politicians who just want out of obligations they did not understand or were misled from the get-go.

Even with debt repudiation talk in the air, it is difficult to file bankruptcy in most states, and default is not a good option for a municipality that needs capital markets access. Plus, as covered elsewhere, given the 10th amendment, bankruptcy does not give the federal courts the same kinds of control over a municipality as a corporate entity.

Keep in mind that the majority of municipal borrowers (like the majority of homeowners) want to do the right thing.

Neither are taxpayers irrational. Protests are explainable. There are only certain states where initiatives and referenda are permissible (and this too has historical roots). There are correlations between rapid growth in taxable assets, public spending increases and taxpayer protest. (I have written about this on my blog and elsewhere.) 

Taxpayers are right to be angry at politicians who are spending their money mainly to buy votes and for poor policy reasons. We need to look at the spending patterns in government and have this debate. How many say, “It’s a pre-election year, of course nothing will get done.” In the private sector, someone with that approach would be fired.

Two important worries Bookstaber hasn’t addressed directly:

Liquidity: There are major liquidity issues at some of the states, quickly filtering down to the local level. At least three states have held back making payments to localities and vendors – California, New York and New Jersey. At some point liquidity markets may stop lending – or the cost will become prohibitive — which will have immediate impact on the economy in those states. Where enough taxes roll in the next month to cover payments, these states will limp along until they are able to cure their structural imbalance or they will hit the wall. While some look for a federal bailout in this situation, a failure to tackle the cure for structural deficits is bad policy. As we learned in recent banking crisis, institutions that require regular market funding for survival will fail in a liquidity freeze.

State solvency: tied to the first point, there is no mechanism in our society to address state insolvency. States, as sovereign entities, do not go bankrupt. There simply does not exist any legal or political institution with the authority to facilitate an orderly reorganization of state obligations. We need to think up objective, non-political structures that put the brakes on spending increases when times are good and that facilitate reorganization when times are bad.

A particularly toxic form of adjustable rate mortgage is going to hit the headlines in the spring and summer of 2010 with defaults, foreclosures and workout discussions extending into 2012.  “Option ARMs” also known as “Pick-a-pay” allow the borrower to choose how much to pay each month and reports indicate 94% of all borrowers paid the minimum.  What’s the catch?  The difference between the minimum and the actual loan rate gets added to the loan balance creating “negative amortization”. 

These loans were originated in the 2004-2007 bubble period with a majority sold during the peak year, 2006.  They were popular in high cost areas, typically more exurban and inland places where houses were slightly more affordable.  A Moody’s report mentions that 54% of these loans were made in California and another 13% in Florida. The next largest origination states were Arizona and Nevada.  About 28% of all home loans originated in the 2004-2008 period in Vallejo-Fairfield (Solano County) California were option ARMs.  One quarter of the home loans originated in Santa-Rosa-Petaluma in Sonoma County were Option ARMs.  These figures are according to the San Francisco Chronicle

The following chart shows the top-10 MSA’s by size of outstanding option ARM balance, including negative amortization, thanks to Alla Sirotic of Fitch Ratings. Note that these are as of August, 2009 so the balances may be different today.  Also, these represent Non-agency RMBS (residential mortgage backed securities).

CLICK TWICE TO ENLARGE

CLICK TWICE TO ENLARGE

The payment option mortgage lured buyers into thinking a big, new house could be affordable.  The pitch?  The borrower thought he/she could re-finance into a more sensible structure or in the worst case, sell the house and pocket some additional equity.  Many of these loans were also used by speculators who hoped to make a profit on the sale and by holders whose files were poorly or fraudulently documented.  Articles citing homeowners who were told the loans had no risk abound.  The housing bust is driving down values and some of the 2006 vintage loans now have a loan-to-value (LTV) of 163%, making the refinance or resale prospect out of the question.

Many of the loan structures had a period of five years before the loans re-cast to fully amortizing and fully loaded interest – hence the toxic brew hitting in 2010-2012.  Around 78% of the option ARMs have yet to recast, and of those loans, 93% have negative amortization, according to a recent Standard and Poor’s teleconference.  The structures also typically carry an amortization trigger when the loan exceeds 110%-125% of the original balance, which could happen before the five year period.  The recast payments could be significantly higher than the monthly amount the homeowner deemed affordable at origination. 

A popular chart that has been circulating among the housing bubble blogs comes from Credit Suisse and shows the timeline for adjustable rate mortgage resets and option ARM recasts.  While the bulk of sub-prime adjustable interest rate mortgages have been re-structured, or are making their way through the foreclosure process, the chart illustrate how option ARMs recasts have yet to hit the fan.  (This version from the “Mortgage Insider, Matthew Padilla of the Orange County Register, updated through April 2009.)

CLICK TWICE TO ENLARGE

CLICK TWICE TO ENLARGE

The solution?  Moody’s concludes that loan adjustment is really the only feasible solution, short of major additional dislocation for homeowners and additional heavy mark-to-market write-downs for the institutions holding this paper.  Wells Fargo is reportedly doing just this on mortgages it owns on the books (accounting for some of the discrepancy in the numbers).  On the political front, the prospect of loan modification is highly controversial.   Some elected officials do not want to look like they favor one group of troubled homeowners over another.  Hmmm. 

Finally, the size of this coming problem is also controversial.  Credit Suisse had to make some assumptions regarding recast dates.  Since payment amounts are, well, optional, the level of negative amortization must be estimated and the time when these loans hit the trigger ahead of the five-year period is also estimated.  Some loans were 40-year, rather than 30-year, complicating the calculations further.  Moody’s and Fitch’s numbers show about $200 billion will be recast in the 2010-2012 period.  The Wall Street Journal cited $750 billion originated between 2004 and 2007 (according to Inside Mortgage Finance).  Some of the discrepancy is between publicly held RMB (residential mortgage-backed) securitizations, loans held on bank balance sheets and agency securities. Sirotic points out that Non-Agency RMBS volume is approximately 17% of the overall outstanding US mortgage market and securitized Option ARMs represent about 13% of the Non-Agency RMBS.  There are definitely more Option Arms loans that are held in bank portfolios, she commented.  So the rest are in federal agency securities, FNMA, FHLMC and GNMA, now a federal balance sheet problem.

A Barclay’s report mentions that many of these loans are already in default ahead of the recast and so should be removed from the numbers. (So is this good news or bad news?)  Housing Wire further reinforces this point:  “The prospect of option ARMs swollen with negative amortization coming to recast and defaulting on a massive scale inspires bond investor fear — and hysterical headlines — but the real fallout might be dimmed by the share of loans that have already foreclosed and no longer remain in the pool.”  Stephen Berg, MD of Lender Processing Services told Housing Wire that they only had about 800,000 active option ARMS.   Whether you think 800,000 loans and $750 billion are small or large numbers, the option ARM will punch more holes in household, government and institutional balance sheets in the coming months.