Bill Gross and Moody's US Ratings

Last week, Bill Gross sent a tweet suggesting that investors should not trust Moody’s US sovereign rating. Given my own concerns about biases in sovereign ratings generally and a review of recent Moody’s pronouncements on US debt, I think Gross has a valid point.

On July 18, the agency affirmed America’s Aaa rating and raised its outlook from negative to stable. The timing of Moody’s action looks a bit odd in an environment of budget gridlock and threats of default if Congress fails to raise the debt ceiling. In 2011, US credit downgrades were sometimes justified in terms of political dysfunction. Since ongoing deficits will necessitate further debt ceiling hikes in coming years and we continue to face a Cold War style domestic political environment, future drama is all but inevitable. Stable, triple-A sovereigns are not supposed to be a source of drama.

Back on July 13, 2011, Moody’s placed the US rating on review for downgrade “given the rising possibility that the statutory debt limit will not be raised on a timely basis, leading to a default on US Treasury debt obligations.” We are now in the same situation, yet Moody’s has not initiated any sort of review. Worse, on September 11, 2012, Moody’s wrote the following about its plans for the 2013 debt ceiling debate: “the government's rating would likely be placed under review after the debt limit is reached but several weeks before the exhaustion of the Treasury's resources.” This is where we are now, so what happened to the review?

After the 2011 debt ceiling increase, Moody’s affirmed its Aaa rating but assigned a negative outlook to US sovereign credit. It gave four conditions that could trigger an eventual downgrade, one of which was the failure to adopt further fiscal consolidation measures in 2013. No such measures have been adopted this year, nor are any feasible in the current political climate. We did see a resolution to the fiscal cliff debate back in January, but that was not a fiscal consolidation measure. Had nothing been done in January, all of the Bush era tax cuts would have expired. Instead, these cuts were made permanent for 99% of Americans at an estimated ten year cost of $3.6 trillion.

In its September 11, 2012 update, Moody’s conditioned the country’s Aaa rating on the adoption of “specific policies that produce a stabilization and then downward trend in the ratio of federal debt to GDP over the medium term.” In lay terms, Moody’s was asking for a grand bargain which would address taxes and entitlement reform. But, as we all know, there has been no grand bargain nor is it reasonable to expect one until 2015 at the earliest.

Based on Moody’s 2011 and 2012 pronouncements, it is hard to justify the agency’s July 2013 action. Moody’s rationalized it on the grounds that “the US government's debt-to-GDP ratio through 2018 will demonstrate a more pronounced decline than Moody's had anticipated when it assigned the negative outlook”. 

Since Moody’s appears to rely on CBO numbers, it is worth checking this contention against changes in the CBO budget baseline. In August 2011, CBO’s baseline budget projection called for a 65.2% debt to GDP ratio in 2018. The latest CBO forecast estimates a 68.4% ratio in 2018. So things actually look worse in 2018 than originally anticipated, yet Moody’s reaction is an upgraded credit outlook.

So we now see the basis for Bill Gross’ tweet. Moody’s pronouncements on US debt are inconsistent and thus not useful to the investment community. Changes in the rating stance do not appear to have a basis in policy; instead they seem to portray a reluctance to offend the federal government. 

From the perspective of Moody’s shareholders, however, this may be a wise approach: given S&P’s claim that the federal lawsuit it faces was payback for its having downgraded the US debt, Moody’s shareholders may (rightly or wrongly) be fearful that their stock would lose value should Moody’s downgrade the US. Bill Gross may be saying that the presence of such a conflict can undermine any hopes for independence or integrity in the ratings process.

In Trust, We Trust

PIMCO's Bill Gross put out a curious statement on his Twitter account on Wednesday.

A company spokesperson reportedly told media outlets that Gross' remark was in regards sovereign credit ratings, not necessarily all credit ratings. 

No support is given for either claim - that Moody's and the US Treasury are in cahoots, or that we can trust S&P, Fitch and Egan Jones as alternative providers. 

Why So Curious? 

What's also got to be at least mildly interesting is that if his comment isn't investment advice, it must be close to it: he's suggesting whose credit opinions are trustworthy (or reliable?) and can be taken into account when considering an investment. Given there's no substantiation for his claims, different from a developed theory, he's saying: "trust me, you can trust these guys."  Is he putting his name or rep behind the future performance of sovereign ratings issued by these three companies? 

And of course, we're all clinching our seats in anticipation: Does PIMCO have proof that it hasn't yet shared, either of Moody's-Treasury collusion or that S&P's, Fitch's, and Egan-Jones' (sovereign) ratings are all trustworthy?  These claims should be "provable" after all, shouldn't they? 

We'll let you know if we find out. 

In the meantime, perhaps this speaks to the development of at least one positive trend: that investors will be encouraged to differentiate between and among the rating agencies - preferably not purely on Bill Gross' say-so, but maybe on performance. Rating agency ABC has a more formidable methodology over here, whereas rating agency DEF's ratings hold greater predictive content over there. 

If no differentiation is made by investors, rating agencies will have few reasons to spend moneys improving their systems, or turn away business in a fight for higher standards or increased accuracy.

Crystal-Clear Country Ratings

If you're one for ratings transparency, you'll be somewhat enthusiastic about Moody's changes to their methodology for rating sovereign debt.

Moody's previous methodology was more of a framework -- there were no "numbers" for the mathematicians among us.

Aa credits had – "Very high economic, institutional or government financial strength and no material medium-term repayment concern."
A credits exhibited – "High economic, financial or institutional strength and no material medium-term repayment concern."

The new methodology provides significant mathematical guidance for those looking to independently verify what a country's rating ought to be, either to prepare for an upgrade or downgrade, or to begin the ratings process for an unrated sovereignty.

It looks like Moody's has taken the stance that their ratings process should be somewhat visible, or "reversible."  The language, too, has changed from their methodology of 2008 to their September 2013 release.

The old methodology held that (emphasis added): 

"There is no adequate model for capturing the complex web of factors that lead a government to default on its debt. Rating sovereign entities involves an unusual combination of quantitative and qualitative factors whose interaction is often difficult to predict." ... "a mechanistic approach based on quantitative factors will be unable to capture the complexity of the interaction between political, economic, financial and social factors that define the degree of danger, for creditors, of a sovereign credit. ... This [ratings methodology's] step by step approach produces a narrow rating range. In some instances, however, the final rating may diverge from the range – in other words, the unusual characteristics of a sovereign credit may not be fully captured by the approach." 

Moody's new methodology offers to provide more than a road-map (emphasis added):

"The aim of this methodology is to enable issuers, investors and other interested market participants to understand how Moody’s assesses credit risk in this sector, and explain how key quantitative and qualitative risk factors map to specific rating outcomes. Our objective is for users to be able to estimate the likely credit rating for a sovereign within a three notch alpha-numeric rating range in most cases."

Importantly, the new methodology affords Moody's analysts some (possibly substantial) flexibility when applying its model, in the form of what they call in-model "adjustment factors."  If properly applied, the adjustment factors can allow analysts room to maneuver to the extent the pure mathematical model alone isn't capturing the risk they're identifying. (As an aside, we would recommend investors push for adjustment factors to be disclosed, so that they cannot be arbitrarily influenced to suit an analyst's opinion: if they can be changed to produce a pre-defined rating expectation, it becomes questionable what the point is of having the model!)

The implementation of the adjustment factors is, unfortunately, not well-defined in any sense.  If, how and when they will be enforced is somewhat unclear -- and the magnitude of their impact is only partially developed, with the implementation and effect of the "diversification" adjustment factor being especially vague:

"This ‘credit boom’ adjustment factor can only lower the overall assessment of the sovereign’s Economic Strength. For most countries, the ‘credit boom’ risk will be Very Low or Low; in these cases, the ‘credit boom’ adjustment factor will be neutral for the assessment of Economic Strength. However, when the combination of the probability of excessive credit growth and its severity lead to a Medium, High or Very High score, this can result in the assessment of Economic Strength being lowered by between one and six scores in the 15-notch Factor 1 score (which translates into a lowering by up to two rating notches). Additional adjustment factors may be considered in the assessment of Economic Strength if deemed appropriate.

Second, the ‘diversification’ adjustment factor allows for the shock absorption capacities afforded by a developed country’s degree of economic diversification and flexibility to lift its overall assessment by one score. We determine the potential for such an adjustment based on the distribution of different sectors’ gross value added in the economy’s annual output. The ‘diversification’ adjustment factor can also lower the overall assessment of a sovereign’s Economic Strength, if, for example, a country is significantly reliant on a single industry or commodity.

Additional adjustment factors may be considered in our assessment of Economic Strength over time if we find that another indicator can provide a universally high degree of explanatory value for Economic Strength." (emphasis added)

While the new methodology doesn't look likely to meet its objective of allowing a market participant to predict a rating "within a three notch alpha-numeric rating range," it must be considered a step in the right direction, at least for those seeking ratings transparency.

------------------------------------------------------------

Reference Documents (may require Moody's log-in):

2013 Methodology
Associated Document -- Refinements to the Sovereign Bond Rating Methodology
2008 Methodology

A Proliferating "Putback" Problemo

Several media outlets have focused of late on the ongoing litigation costs being incurred by JPMorgan (and suffered by its shareholders) pertaining to crisis-era loans and structured products, and post-crisis concerns, like investigations into the "London Whale" trade and allegations into potential manipulations of LIBOR and the energy market manipulations.  (See for example the FT's "JPMorgan pledges to clean up legal woes.")

We're closing in on 5 years since Lehman filed for bankruptcy protection but, of course, the problematic mortgage loans originated between 2005 and 2008 haven't completely left the system.  In some ways the problems they're causing the banks are increasing.  We'll explore the headaches one particular aspect – mortgage repurchases – could be causing at Deutsche Bank, and how a couple of recent court rulings could exacerbate the pain.

The WSJ recently reported "Deutsche Bank Net Profit Halves on Charge for Potential Legal Costs." After absorbing the charge, the Journal reports, Deutsche Bank had litigation reserves of EUR 3.0 billion; as of June-end, DB had cordoned off $534 million in provisions against outstanding mortgage repurchase demands - or "putbacks."

We went back and took a look at DB's reserves historical against put-backs.  Since year-end, they have increased their reserves roughly 20%, from EUR 341 mm to $534 mm (applying a 1.3x exchange rate from EUR to USD at both year-end '12 and mid-year '13).  These are provisions against outstanding demands that grew roughly 28% from $4.6 bn as of YE to $5.9 bn as of June-end.

Will repurchase demands continue to grow, requiring the posting of additional reserves?

First, let's discuss mortgage put-backs.

Mortgage Put-backs

According to Deutsche, from 2005 through 2008, as part of its U.S. residential mortgage loan business, it sold "approximately U.S. $ 84 billion of private label securities and U.S. $ 71 billion of loans through whole loan sales, including to U.S. government-sponsored entities such as the Federal Home Loan Mortgage Corporation and the Federal National Mortgage Association."

To keep this simple, Deutsche essentially sells or transfers or conveys residential mortgages into securitization vehicles, or trusts: it "puts" them into the trust.  Thousands of loans can be transferred or deposited into a single trust. If any of the loans is later found to be, let's say, problematic, or in violation of agreed-upon representations made, a trust's trustee would ideally be able to put 'em back to Deutsche at cost: the "put-back". (Keep in mind that DB may not have made representations on all of the loans transferred into trust vehicles for which DB acted as the securitization sponsor.)

In DB's words now, "Deutsche Bank has been presented with demands to repurchase loans or indemnify purchasers, other investors or financial insurers with respect to losses allegedly caused by material breaches of representations and warranties. Deutsche Bank’s general practice is to process valid repurchase demands that are presented in compliance with contractual rights. Where Deutsche Bank believes no such valid basis for repurchase demands exists, Deutsche Bank rejects them." (emphasis added)

Importantly, two parties don't always agree with the validity of a repurchase request.  The question becomes, then, whether DB's disclosed $5.9 billion of repurchase demands outstanding constitute those that are "agreeable" to Deutsche, or whether DB might reject some of them?  And if they are all agreeable, how large is DB's exposure to other repurchase demands that it has previously rejected, but that may become contentious or the subject of litigation.

We can't be sure, but we do know that DB's disclosure changed in this regard between Sept-end 2012 and YE 2012.

From: "As of September 30, 2012, Deutsche Bank has approximately U.S. $ 3.3 billion of outstanding mortgage repurchase demands (based on original principal balance of the loans and excluding demands rejected by Deutsche Bank)." (emphasis added)

To: "As of December 31, 2012, Deutsche Bank has approximately U.S. $ 4.6 billion of outstanding mortgage repurchase demands (based on original principal balance of the loans)."

Since then the description hasn't changed. In their YE earnings call transcript, they mention there having been an increase from $3.3bn to $4.6bn, "attributable to demands made by RMBS investors." This seems to suggest that the change in description is incidental to, rather than the cause of, the increase.

They also intimate on the call that the improving market conditions – probably home prices – understandably translate into lower losses on the repurchased loans.  Could this be one reason why the repurchase demands are growing or outpacing settlements – or that Deutsche is postponing, slowing, or opting against the repurchase or settlement of these loans, as indicated by the slow growth in the third row of the table?

Mortgage Put-back Litigation 

As mentioned above, parties don't always agree as to the viability of a put-back claim.  Deutsche Bank unit DB Structured Products is on the receiving end of a number of lawsuits in which indenture trustees, on behalf of the mortgage-backed securities trusts themselves, claim that DBSP has failed to repurchase some or all of the violating loans presented to them.

Last week, District Judge Harold Baer added to a recent string of decisions in favor of securitization trusts’ ability to “put-back” non-complying mortgage loans to the conveyor – in that instance UBS Real Estate Securities Inc. (“UBS”).  Like many related complaints, the argument was that UBS had breached its contractual obligation under the relevant Pooling and Servicing Agreements (“PSAs”) to repurchase certain mortgage loans that did not conform to its representations and warranties.

In denying UBS’ motion to dismiss the case, Judge Baer explicitly disagreed with an earlier decision, more favorable to the defendants, made by US District Judge John Tunheim in MASTR Asset Backed Sec. Trust 2006-HE3 ex rel. U.S. Bank Nat. Ass'n v. WMC Mortgage Corp. – taking rather the stance consistent with a 2002 First Circuit decision and referencing other recent decisions more favorable to trust investors, including that of Judge Rakoff in Assured Guar. Mun. Corp. v. Flagstar Bank, FSB, No. 11 Civ. 2375.

One such matter of Deutsche's, before Justice Kornreich of the Supreme Court of the State of New York, concerns a trust that originally contained 8,815 home loans, according to the complaint.  The trustee-plaintiff put forward in the complaint that "[in] total, 1,642 repurchase demands have been submitted to DBSP thus far," and contended that "[the] Trust has suffered over $330 million of losses on non-performing Mortgage Loans to date, and will continue to suffer damages as a result of DBSP’s failure and refusal to comply with its express contractual obligations."

According to the complaint, a significant portion of mortgage loans tested has failed to comply with the representations made:

"A loan-level investigation performed by an independent forensic mortgage loan review firm (the “Forensic Review Firm”) has, thus far, revealed breaches of the Representations and Warranties with respect to 696 Mortgage Loans in the Trust out of the 697 Mortgage Loan files that were analyzed. Furthermore, an analysis of publicly-available data regarding the Mortgage Loans in the Trust uncovered breaches of the Representations and Warranties with respect to 946 Mortgage Loans."

In mid-May 2013, Justice Kornreich decided to let the case continue, denying defendant DBSP's motion to dismiss the complaint.  ACE Securities Corp. v. DB Structured Products, 650980/2012, NYLJ 1202604071610, at *1 (Sup., NY, Decided May 13, 2013).

Our investigation came up with 17 such unique, current, put-back lawsuits filed against DB Structured Products since March of last year (and there may well be others) by trusts holding an aggregate of roughly $15bn in loans at origination. (In some of the cases, parties other than the trustee have sued DB Structured Products, purportedly on behalf of the trust or trustee.)

The trusts have since declined in size, but it is worth noting that even some loans backed by homes which have foreclosed upon, may still be put-back to the transferor: in this case DB Structured Products. Justice Kornreich maintained that DBSP ought still be responsible for repurchasing such non-complying loans, finding the defendant's argument to be, in her words, "unconvincing." 

"If DBSP were correct, it would be perversely incentivized to fill the Trust with junk mortgages that would expeditiously default so that they could be Released, Charged Off, or Liquidated before a repurchase claim is made. Indeed, if DBSP learned that loans were non-conforming and played a crafty game of accounting by moving them off the Trust's books to their own to evade their repurchase obligations, such actions would be a breach of the duty of good faith and fair dealing. Consequently, it is to no avail to contend that the nonconforming loans are long gone and the Trustee's repurchase rights have been extinguished by DBSP's actions."

Richmond's Million Dollar Eminent Domain Homes

Three of the mortgages Richmond, California is threatening to acquire through eminent domain have balances in excess of $880,000 - suggesting that the underlying properties were worth over $1 million at the peak of the housing boom. Public records show that two of the three properties did, in fact, change hands at prices in excess of $1 million several years ago. (On the 624 loans up for acquisition, the median and average outstanding balances are roughly $380,000.)

The three homes are all in the tony Point Richmond neighborhood. The accompanying picture, taken from Google maps, shows the house carrying the third largest mortgage in the eminent domain pool. The other two homes were apparently too far back from the road for Google’s van to photograph.

The biggest mortgage, with a balance of slightly over $1.1 million, is on a property that sold for $1.4 million in 2001 - well before the housing boom crested in 2006. The property currently has an assessed value of about $750,000 and Richmond is proposing to buy the mortgage for $680,000. The three bedroom waterfront home is 2500 square feet and sits on a 17,000 square foot lot - quite large by San Francisco Bay Area standards.

While this particular homeowner appears to be underwater, appraisals for such unique homes are prone to both error and volatility given the lack of recent comparables. Thus, the city’s “offer” to mortgage backed security holders may be significantly less than the homeowner could receive on the open market.

More importantly, the fact that such expensive homes are included in Richmond’s eminent domain initiative raises the question of what public interest is being served. Point Richmond is not a blighted neighborhood and any foreclosed home would likely sell very quickly. Further, anyone in a million dollar home is probably not poor - at least not by any conventional definition of the word poor.

Proponents of the use of eminent domain to resolve underwater mortgages see this as a way for the little guy to “take it” to Wall Street. While it is true that Wall Street made substantial profits packaging up pools of dodgy mortgages, Richmond’s action does not address that injustice. Those profits have already been taken: what remains are mortgage borrowers and MBS investors, many of whom are public employee pension funds.

So the question really is: Should the city of Richmond use eminent domain to transfer wealth from public employees (and the taxpayers that fund their pensions) to affluent homeowners who took on more mortgage debt than they should have?

------------------
For an update on this, visit the Wall Street Journal's coverage or the San Francisco Chronicle's coverage.

Are Moody's and S&P Growing Apart?

Leading up to the November 2012 announcement that McGraw Hill would sell its education division – and become more of a rating agency/financial institution like Moody's Corp. – the two companies' stocks moved in tandem, with a correlation of over 90%. Since then, the correlation has gone way down, into the 70-80% region.

AAAs still Junk -- in 2013!

Breaking news: Several securities, boasting ratings higher than France and Britain as of two weeks ago, are now thought quite likely to default.

Moody's changed its opinion on a number of residential mortgage-backed securities (RMBS), with about 13 of them being downgraded from Aaa to Caa1.

The explanation provided: "Today's rating action concludes the review actions announced in March 2013 relating to the existence of errors in the Structured Finance Workstation (SFW) cash flow models used in rating these transactions. The rating action also reflects recent performance of the underlying pools and Moody's updated expected losses on the pools."

In short: the model was wrong - oops!  ($1.5bn, yes that's billion, in securities were affected by the announced ratings change, with the vast majority being downgraded.)

Okay, so everybody gets it wrong some time or other.  What's the big deal?  The answer is there's no big deal.  You probably won't hear a squirmish about this - nothing in the papers.  Life will go on.  The collection of annual monitoring fees on the deal will continue unabated and no previously undeserved fees will be returned. Some investors may be a little annoyed at the sudden, shock movement, but so what, right?  They should have modeled this anyway, they might be told, and should not be relying on the rating.  (But why are they paying, out of the deal's proceeds, for rating agencies to monitor the deals' ratings?)

What is almost interesting (again no big deal) is that these erroneously modeled deals were rated between 2004 and 2007.  So roughly six or more years ago.  And for the most part, if not always, their rating has been verified or revisited at several junctures since the initial "mis-modeled" rating was provided.  How does that happen without the model being validated?

A little more interesting is that in many or most cases, Fitch and S&P had already downgraded these same securities to CCC or even C levels, years ago!  So the warning was out there.  One rating agency says triple A; the other(s) have it deep in "junk" territory.  Worth checking the model?  Sadly not - it's probably not "worth" checking.  This, finally, is our point: absent a reputational model to differentiate among the players in the ratings oligopoly, the existing raters have no incentive to check their work. There's no "payment" for checking, or for being accurate.

Rather, it pays to leave the skeletons buried for as long as possible.

---------------

For more on rating agencies disagreeing on credit ratings by wide differentials, click here.
For more on model risk or model error, click here.

---------------
Snapshot of certain affected securities (data from Bloomberg)

Outdated Ratings

Two Bloomberg reporters wrote a thought-provoking piece late last week (see Lost AAA Brings Falling Yields-to-Deficits on Downgrade) on the forthcoming ratings downgrade for the US.  In some ways they explore the age-old question of the (odd) relationship between a ratings change and the market price, or yield, of a bond.

"Yields on Treasuries are lower, the dollar is stronger and the S&P 500 Index (SPX) of stocks reached a record high since Aug. 5, 2011, when S&P said the U.S. was less creditworthy than Luxembourg and 17 other sovereigns."

But their article really tests the case of whether a US downgrade can be implemented despite an improving economic picture.

"... the unemployment rate has fallen, household wealth has reached a record and the budget deficit is shrinking. More downgrades may be coming, anyway."

Stepping back, it's fair to debate whether the economy is in better shape.  There are likely economists on both sides of the table on this one.  And certainly the statement that "the budget deficit is shrinking," if read on its own, can be misleading: the estimated budget deficit this year is expected to be lower than that of last year, but the overall federal deficit is expected to increase (or in more proper US fiscal parlance, the debt is expected to increase).

According to the Bloomberg article, "Moody’s Investors Service said it’s awaiting lawmakers’ budget decisions this year as it weighs reducing America’s Aaa."

But given the US GDP is growing, while Europe is in a recession, can it make any sense to downgrade the US?  Remember, the rating agencies claim their ratings are RELATIVE measures of risk.  In other words, the rating agencies are ranking each country relative to other countries.  So for the US to be downgraded, it would need to be getting worse relative to its competition.

Our guess is that what's happening here is the result of a fair share of awkwardness surrounding a situation in which many of the rating agencies' outstanding ratings may not reflect their current opinions.  If they delayed implementing the downgrade since the US first failed to meet the relevant criteria necessary to maintain the AAA rating, they would now look a little silly downgrading so long after the fact, now that the economy has stabilized, or turned the corner.

According to the article, "Fitch Ratings, which has a “negative” outlook on the U.S., said in February that the debt trajectory isn’t consistent with a AAA borrower."  Even if you agree with Fitch on this, okay, the US may still not be back at AAA if the rating agencies were to strictly adopt their criteria. But, as the Bloomberg article forces us to ask, how can a downgrade be appropriate now?

There's Always a Model

Opponents of quantitative credit models and their use in bond ratings contend that the decision to default is a human choice that does not lend itself to computer modeling. These critics often fail to realize that the vast majority of ratings decisions are already model-driven.

Recently, Illinois’ legislature failed to pass a pension reform measure. Two of the three largest rating agencies downgraded the state’s bonds citing leadership’s inability to shrink a large unfunded actuarial liability. The downgrades were a product of a mental model that connects political inaction on pensions to greater credit risk.

Indeed most rating actions and credit decisions are generally supported by some statement of a cause and effect relationship. Drawing a correlation between some independent driver – like greater debt, less revenue or political inaction – and the likelihood of default is an act of modeling. After all, a model is just a set of hypothesized relationships between independent and dependent variables. So ratings analysts use models – they just don’t always realize it.

The failure to make models explicit and commit them to computer code leads to imprecision. Going back to the Illinois situation, we can certainly agree that the legislature’s failure to reform public employee pensions is a (credit) negative, but how do we know that it is a negative sufficient to merit a downgrade?

In the absence of an explicit model, we cannot. Indeed, the ratings status quo has a couple of limitations that hinder analysts from properly evaluating the bad news.

First, ratings categories have no clear definition in terms of default probability or expected loss, so we don’t know what threshold needs to be surpassed to trigger a downgrade.

Second, in the absence of an explicit model, it is not clear how to weigh the bad news against other factors. For example, most states, including Illinois, have been experiencing rapid revenue growth in the current fiscal year. Does this partially or fully offset the political shortcomings? And, all other things equal, how much does the legislature’s inaction affect Illinois’ default risk.

In the absence of an explicit model, analysts make these calculations in an intuitive manner, creating opportunities for error and bias. For example, a politically conservative analyst who dislikes public pensions may overestimate their impact on the state’s willingness and ability to service its bonds. Likewise, a liberal sympathetic to pensions may underestimate it.

A news-driven downgrade like the one we saw last week may also be the result of recency bias, in which human observers place too much emphasis on recent events when predicting the future. Worst of all, an implicit mental model can easily be contaminated by commercial considerations unrelated to credit risk: what will my boss think of this proposed downgrade, or how will the issuer react? In an ideal world, these considerations would not impact the rating decision, but it is very difficult for us mortals to compartmentalize such information. Developing and implementing a computer rating model forces an analyst to explicitly list and weight all the independent variables that affect default risk.

Computer models are also subject to bias, but they provide mechanisms for minimizing it. First, the process of listing and weighting variables can lead the analyst to identify and correct her prejudices. Second, to the extent that the model is shared with other analysts, more eyes are available to find, debate and potentially eliminate biases.

Once the model is in place, news developments can be tackled and analyzed without recency effects. In the case of Illinois pensions, the legislative development would have to be translated into an annuity of expected future state pension costs (or a distribution of same) so that it can be analyzed with reference to the state’s other fiscal characteristics.

Computerized rating models – like any human construction – are subject to error. But the process of developing, debating and iterating explicit models tends to mitigate this error. We all apply models to credit anyway; why not just admit it and do it properly?

An Open Source Alternative to No Bid Contracts

At Muniland, Cate Long reports that the US Treasury Department’s Office of the Comptroller of the Currency (OCC) awarded Municipal Market Advisors (MMA) a contract to evaluate the risk of municipal bond holdings by banks it regulates.  OCC did not find any credible alternatives and thus is awarding the contract to MMA on a no-bid basis.  Quoting at length from Cate’s excellent blog post:

So federal regulators, who can no longer use credit ratings for evaluations of the municipal bond holdings of the commercial banks that they regulate, just gave a no bid contract to MMA, a relatively small firm with four principals. In essence the OCC will be substituting the opinions of MMA for those of the credit ratings agencies. Federally chartered banks held $363 billion of municipal securities as of 4th quarter 2012 according to the Federal Reserve ... The federal bank regulator will essentially be substituting the work of credit rating agencies, which issue over 1 million individual municipal ratings, with “research” from a small private shop. Is this wise? I think restricting themselves to such limited information is short-sighted given that muniland has over 80,000 issuers with $3.7 trillion of municipal debt outstanding. High quality credit analysis for even the debt of 50 states requires a shop bigger than MMA. Let alone all the other issuers. … Of course all the folks at MMA are nice, informed market professionals. But this process of hiring independent municipal research is ridiculous. No bid contracts have no place in our new, more transparent, post Dodd-Frank regulatory framework. The municipal bond market is facing its toughest challenges since the Great Depression and this BPD/OCC process needs more public input and openness.

As I will discuss at Tuesday’s SEC Credit Ratings Roundtable, there is a better alternative to this kind of arrangement. For almost 50 years, academics have been churning out corporate default models.  This modeling effort could be extended to structured and government bonds.  If the modeling data and software were fully open, these academic tools could undergo rapid, iterative improvements through a process of mass collaboration:  like Wikipedia or Linux.  If the SEC were to create a standards board for open source credit models, a group of experts would be empowered to separate the wheat from the chaff among these open source products. Regulators could further encourage the development of such tools by allowing results of certified models to be used in lieu of ratings as a credit-worthiness standard – meeting the spirit of Dodd-Frank Section 939A.  I make this argument at greater length here.

What should supplement or replace ratings?  Confidential, non-reproducible findings from a proprietary vendor, or transparent tools developed using academic research protocols and benefiting from peer review? I think the answer is clear.