The Irish Crisis and the Latest Silver Bullet “Solutions"

This column is the second in a series of discussions of the developing EU crisis that was the subject of the recent Kilkenomics Festival in Kilkenny, Ireland.  This installment addresses four recurrent myths that are the enemy of effective financial regulation.

·    Assets aren't bad, only misunderstood
·    There are silver bullet solutions that will resolve bad assets at minimal cost
·    Control fraud by financial institutions can be ignored
·    Bank examiners and bank underwriting add little value and can largely be displaced by software

Thorough bank examinations and a sophisticated understanding of fraud schemes are prerequisites for reducing the frequency and severity of financial crises.  The recurrent efforts to substitute software for underwriting and bank examination have allowed hundreds of billions of dollars in losses.  Robert Field has offered the latest silver bullet solution to the Irish crisis – and cheered the BOE's desupervision as implementing “21st Century Oversight.”  

Field's basic idea is unassailable: transparency is desirable.  The problems are that he ignores the limitations of his proprietary software and overpromises what it could accomplish.  The Irish crisis and the BOE's desupervision illustrate the dangers of the resultant complacency and analytical errors that come from silver bullet “solutions.”        

The Irish crisis has nothing to do with structured financial products, which is Mr. Field's field.  This is important because securitization has been blamed for the current U.S. banking crisis.  Epidemics of accounting control fraud drove the financial crises in Ireland, Iceland, the former Soviet states (“tunneling”), the Enron era scandals, and the S&L debacle.  Securitization did not play a primary role in any of these crises.  Securitization (and Fannie and Freddie) are not necessary conditions for banking or economic crises.  That does not mean that securitization cannot exacerbate a crisis, but it does mean that it does not explain why we suffer recurrent, intensifying financial crises.  

The Irish crisis shows that several other aspects sometimes asserted by Americans to be essential to a financial crisis are not.  For example, some claim that the fact that some U.S. states do not allow deficiency judgments against homeowners (i.e., the loans are non-recourse) played a decisive role in the crisis.  Ireland, however, allows deficiency judgments.  Ireland's crisis, like the S&L debacle in the Southwest, was driven by a hyper-inflated bubble in commercial real estate (CRE).  The U.S. does not protect commercial developers from deficiency judgments.  It is revealing that it is very unusual for creditors to require U.S. CRE developers to provide meaningful personal guarantees.  Non-recourse CRE loans are the norm.  If recourse is a vital protection to creditors, why do they not consistently require it on far larger projects (CRE) for which the developers often have substantial personal wealth?   

Ireland's CRE bubble was the worst in the developed world in terms of relative magnitude.  The CRE bubble helped drive a bubble in residential housing by causing land prices to rise.  Ireland's record bubble produced epic destructions of wealth.  Control frauds increase their lending into the teeth of the bubble.  This misallocates assets to unproductive purposes (carpeting Ireland with “ghost” shopping centers).  The result was catastrophic losses to the banks, developers, and homeowners.  The losses drove a severe recession.  The recession drove a wave of migration.  

The Irish government then turned the banking crisis into a budgetary and sovereign debt crisis by guaranteeing virtually all of the banks' liabilities.  The government bought a pig in a poke because it ignored all the warning signs of massive fraud and monumental losses in the assets.  The Irish guarantee led to (1) huge debts for every citizen (roughly $15,000 per capita), (2) a budgetary crisis because Ireland's adoption of the Euro means that it no longer has a currency under sovereign control and is subject to the EU treaty limits on deficits, (3) a deeper recession because Ireland adopted austerity in the teeth of a severe recession, (4) substantial migration, which made the recession and the collapse of the bubble more harmful.  The banks' losses were so large relative to the shrinking Irish economy that the bailout of the banks was smoke and mirrors.  The bailout relied on promissory notes from the Irish government.  The notes did not guarantee any cash payments (of interest or the repayment of principal) to the banks prior to maturity.  The Irish government had guaranteed bank losses that were over 25% of its falling GDP.  

Mr. Field has never been a regulator or (he worked relatively briefly for the Federal Reserve as a research assistant and for a bank as an AVP).  He argues, however, that his software provides the ultimate in silver bullet solutions.  The title of his article says it all:  “Ireland's Choice: Inexpensive Cure or Expensive Bailout.”  We lived for 20 years in Silicon Valley, so I admit to skepticism about software claims.  Here's how Field explains his Irish cure:

“The Irish government faces a choice in how to handle its financial crisis. It can implement an inexpensive cure and resolve the cause of the crisis, or it can implement an expensive bailout and treat the symptoms of the crisis.

At its core, the cause of the Irish financial crisis is the same as the 2007/2008 global credit crisis. The root cause of the crisis is the inability of investors and other market participants to independently analyze and value commercial and residential real-estate loans.”

The premise is that investors and creditors could not determine the value of the assets that would determine whether their investments or loans would be sensible.  Neo-classical economics (and the efficient markets and contracts hypotheses) claims that this is impossible.  Lenders and investors would suffer such severe losses in such an inefficient market that they would not lend.  If one assumes inefficient markets (lemons markets) and applies Akerlof's findings then investors would only purchase at very large statistical discounts from the original lenders' book values (in reality, they often purchased them at a premium) and lenders would only loan if they could charge extreme yield premia (in reality, spreads between nonprime and prime loans fell in the U.S.).  Note that if multi trillion dollar securities and debt markets are lemons markets, then all of the models used to price securities and debt instruments will dramatically overvalue the securities and debt instruments.  Criminologists and regulators (and traders) have long known that markets and contracts are frequently inefficient, so I am not troubled by Field's recognition that markets are inefficient.  My difficulty is with his assumption that his software will make markets and contracts efficient.  

Field's assertions rely on three inaccurate premises.  First, Ireland and the U.S. did not suffer severe, real losses due to bad assets or bubbles.  Both crises had ready “inexpensive” resolutions available through his software.  Second, his software allows nearly real time valuations by outside investors of the true asset values of bank loans.  Third, his software allows investors to reach a reliable asset valuation that “reflects the reality of the situation and not fear of how bad the situation could be.”

With regard to the Irish crisis, Field argues:

“Like the sub-prime mortgage backed securities, the devaluation cycle on the Irish 'bad' assets appears to have no logical stopping point other than zero. As a result, investors are questioning the solvency of Ireland.

Naturally, if Ireland accepts a bailout from the EU and the IMF to close the difference in interest rates, it faces both the stigma associated with applying for aid and restrictions on the government and its fiscal policies. This is a very high price to pay for treating the symptoms of the problem like high interest rate differentials.

There is an inexpensive alternative available to the Irish government which treats and cures the cause of the devaluation cycle of the Irish 'bad' assets.”

Field does not understand that Ireland already faces crippling “restrictions on the government and its fiscal policies” because it is a member of the EU and adopted the Euro.  It cannot devalue.  It cannot run the type of deficits it should employ to reduce the severity of its recession.  It also faces constraints because of bank guarantee and the size of the bank losses relative to Ireland's GDP.  It cannot borrow except at very high interest rates because of these facts and its deepening recession.  Field does not propose any change in the guarantee.  He says his solution is premised on the currently projected Irish bank losses, and there being no future losses.  

“This success of this alternative is based on the Irish government having identified all the 'bad' assets and properly discounted them.”

He also states that he is assuming that the Irish “good bank” has only good assets.  The reality is that the Irish government has stated that it will need to create another good bank/bad bank division (out of the supposed good bank) to resolve Anglo Irish Bank.  His first premise means that his “cure” isn't a cure and isn't “inexpensive.”  Remember, the Irish government has not paid to resolve its banks' bad assets.  It has, primarily, given promissory notes with no assurance of cash payments of interest or principal until the notes mature in 10 years.  The bank losses that Field assumes for purposes of his proposal, given the (insane) Irish guarantee, are crippling, not “inexpensive.”  His software will not make the bank losses or the guarantee disappear.  The losses are so large that the guarantee creates a fiscal crisis and in conjunction with austerity imperil the Irish recovery from the recession.

Mr. Field's second premise is that his software would allow investors to independently determine accurate market values for Irish bank assets.  He promises that his software would provide sufficient information to investors to allow them to determine the true economic value of Irish bank assets.

“The proposed solution of providing observable event based reporting puts the loans into the equivalent of a clear plastic bag. With observable event based reporting, all the changes, like payments or defaults, to the loans are reported on the day that the changes occur. Investors can see what they are buying and can value the securities or the good or bad banks the loans support. If investors can value the loans, they can also determine that Ireland is solvent without a bailout.”

 The quotation requires us to return briefly to the problems with Field's first premise.  It is unclear what Field means by “solvent” when he refers to Ireland.  A nation with a sovereign currency never needs to default on sovereign debt, but Ireland has adopted the Euro.  Irish bank losses are so large that, given the guarantee, Ireland either has to push its economy deeper into recession through increased austerity or face promptly greatly increased interest rates.  Of course, they can face both simultaneously because the combination of the deepening recession, emigration, budgetary crisis, austerity, and increasing interest rate expenses can create a self-reinforcing downward spiral.  Ireland also faces the danger that another nation will trigger a “flight to quality” and cause a dramatic increase in spreads on Irish debt at some future date.  Absent a bailout, leaving the Euro, or terminating the guarantee of bank losses, Ireland faces a crisis.

Turning now to Field's second premise, “observable event based reporting” cannot deliver the information necessary to determine correct asset valuations.  Instead, it is likely to worsen one of the worst features of existing models.  There is a famous saying that captures the problem:  It's not the things you don't know that cause disaster; it's the things you do know – but aren't true.  The financial models were absurd primarily because the controlling officers created perverse incentives to create models that inflated asset values.  The alleged sophistication and precision of the models made them particularly dangerous to those that did not know the models were bent.  

Field's model is fatally flawed by its claim that “observable” events are sufficient to value assets.  The Irish and U.S. crises show how severe this flaw is.  Nonprime loans, particularly liar's loans, were endemically fraudulent.  Sophisticated frauds are designed not be readily “observable.”  The only way to detect them in time to minimize losses is to look for fraud “markers” in the individual loan applications files as part of a prudent underwriting process.  Where the lender is an accounting control fraud the lender will not conduct reliable underwriting, it will not keep accurate records, and it will not make accurate reports.  Only vigorous banking examiners, reviewing adequate samples of loan files are likely to have the independence and expertise necessary to detect accounting control frauds prior to catastrophic losses.  Field is naïve if he really believes his metaphor – that his software will turn banking into a “clear plastic bag” – perfectly transparent.    

Reporting on loan payments and loan terms (the two key deliverables that Field's software is supposed to provide) is inadequate to identify accounting control frauds until they are near collapse.  Fraud occurs at multiple levels in a fraudulent bank.  U.S. nonprime lenders and Irish banks provided false information about the borrowers' creditworthiness and the value of the collateral.  Epidemics of accounting control fraud cause bubbles to hyper-inflate and make it simple to hide losses for years on massive portfolios, but even individual accounting control frauds can hide losses for year by refinancing bad loans.  The loan delinquencies and defaults will eventually occur in profusion, but that is far too late for investors or regulators to act to prevent massive losses or bubbles.  Epidemics of accounting control fraud, particularly when they hyper-inflate a bubble, can cause catastrophic losses (Anglo's bad assets have experienced losses of over 70 percent).  A software system that misses fraud losses will grossly overestimate asset values in a situation like Ireland.  

Field also assumes that if investors had more information on loans they would impose effective private market discipline.  There was no effective private market discipline even when nonprime specialty lenders and the worst Irish banks publicly informed investors that the banks had provided grossly inadequate loss reserves and it was public that there was an epidemic of mortgage fraud.      

Field also makes the common mistake of assuming that knowing current loan delinquency frequencies is adequate to determine the risk of future losses.

 “Based on their own analysis, investors can then set a differential between Irish and German government bonds that reflects the reality of the situation and not fear of how bad the situation could be."

Field thinks the “reality of situation” (current information on loan payments) should remove investors' “fear of how bad the situation could be.”  An accurate snapshot of loan performance inherently cannot tell an investor “how bad the situation could be.”  (I have explained why his software will not provide an accurate snapshot in the situations where it is most important to have accuracy.)  For example, the fact that CRE loans are not currently delinquent may be an aspect of the “reality” of the “situation,” but the fact that the interest is being “paid” out of an interest reserve may be a far more important aspect of the “reality” and the fact that the appraised value of the shopping center being constructed was deliberately inflated by the lender is by far the most important aspect of the relevant “reality.”      

When a hyper-inflated bubble collapses formerly good assets can also suffer serious losses.  This is particularly true in CRE bubbles.  Again, the “observable” events will come far too late to protect the investors.  Note also that asset values change continuously as the bubble hyper-inflates and over the course of its collapses.  Similarly, asset values change as a recession deepens or eases.  Even if we could assume, contrary to reality, that Field's software delivered accurate snapshots of “reality” the software could never remove the investor's need to consider risk – a concept that includes the “fear” “how bad the situation could be[come].”  Field does not understand risk or he is not being candid about the limitations of his software.