Irish Fish, and Banks, Rot from the Head

By William K. Black
(Cross posted at

I’m back from my annual purifying rite – participating in Kilkenomics IV in Kilkenny, Ireland.  It was a big success again this year, with dozens of sold out events in which economists (and the odd criminologist) are paired with superb professional comedians who serve to keep things lively, understandable, and blunt.  Regular folks stop you on the street and talk to you for hours in the bars about economics.  This year, the organizers (Richard Cook and David McWilliams) added “young economist” awards for the equivalent of high school and junior high school awards.  The brilliant Dan Ariely and I had the privilege to help judge the final round of the competition.  Should you ever find yourself near Ireland in early November during the festival I encourage you to join us in Kilkenny.

Jean Claude Juncker, PM, Luxembourg, and the head of the Eurogroup council of eurozone finance ministers infamously said: “When it becomes serious, you have to lie.”  Kilkenomics says: When the “serious” people feel that they “have to lie” it is vital that we resurrect the comedic “fool” who speaks truth to power.

Here and here are two articles in the Irish papers that try to capture the weird charm of the event.

The “note of banking reality” I sought to inject in Kilkenomics IV was entitled “Fish Rot from the Head.”  A number of diverse cultures use this metaphor to emphasize that society’s most important problems come primarily from society’s VIPs.  In the Irish case, this might seem to be obvious given the decisive role that criminal banking enterprises played in driving the Irish financial crisis.  The Irish people are distinct, however, in blaming themselves for the crisis.  This is one of the contributing barriers to the Irish holding accountable the prominent politicians and bankers who are vastly more culpable for causing the crisis.

The biggest obstacle to reform is that the Irish political parties are the most unusual in Europe.  The two largest parties are still sister parties many decades over they split over (Irish) Republican tactics.  Neither party has stood for anything distinctive for many decades.  The “Labour” party is an anti-laborer party.  The “greens” elected to parliament so disgraced the movement that the voters wiped them out of parliamentary representation.  Sinn Fein has been the only significant Irish party with a coherent economic policy, but it still struggles to overcome the discredited romance of violence exemplified by its older leaders.

Ireland doesn’t have U.S.-style gridlock in its parliament, but nothing fundamental changes even in response to an economic crisis that threatened Ireland’s existence as a modern nation state.  Had Lehman Brothers’ failure been delayed by two years Ireland would still be spiraling downward today.  Lehman’s failure in no way “caused” the Irish crisis, but it did precipitate it.  There were no effective regulatory steps being taken – none – by Ireland in 2008 that would have placed even the equivalent of a significant speed bump in the path of Ireland’s criminal enterprises.  In late 2008 – roughly 30 months after the residential real estate bubble ceased expanding – the Irish banking regulatory leadership universally claimed that Ireland’s criminal enterprises were “well-capitalized” and had no solvency concerns.  By 2007, the criminal enterprises, realizing that the jig was up in trying to hyper-inflate further the residential real estate boom, had switched massively to hyper-inflating the commercial real estate bubble.

Commercial real estate, as the U.S. saw in the savings and loan debacle, serves as superior “ammunition” for the “weapon” of accounting control fraud relative to residential real estate.  It is far easier for the officers who control a bank engaged in accounting control fraud to grow extremely rapidly through commercial real estate lending.  The bank can use a literal double-handful of commercial real estate borrowers to make tens of billons of euros of loans.  The borrowers will soon be joined at the hip with the controlling bank officers because the commercial projects will cease to cash flow as the glut of commercial real estate reduces rents.  The commercial real estate developers’ primary means to stave off default is to borrow ever greater amounts from the banks and use the new loan proceeds to pay interest on the old debts.  (The banks can also fund purchases of the older commercial real estate developments at grossly inflated prices.)  I told the Irish audience that the saying in the trade in the U.S. is that “a rolling loan gathers no loss.”  They liked it and chuckled at the industry using such a poetic twist.  I explained that the poetry is to be expected because America is a nation of poets (they got it).

My “fish rot from the head” talk was aimed in large part at cutting through the myths foisted on the Irish people through a troika of reports that purport to explain the Irish crisis.  In practice, the Irish people got charged millions of euros for reports that read like defense briefs for the bank CEOs who caused the financial crisis.  The primary insanity is reveal in two paragraphs of the primary report, by Peter Nyberg.  Nyberg is an economist who exemplifies the IMF/EU apparatchiks’ view in which accounting control fraud epidemics cannot exist.

“The mandate of the Commission did not include investigating possible criminal activities of institutions or their staff, for which there are other, more appropriate channels. Under the Act, evidence received by the Commission may not be used in any criminal or other legal proceedings.”

The Commission has not investigated any issues already under investigation elsewhere. Instead, the Commission used its limited time and resources to investigate, as its Terms of Reference specified, why the Irish financial crisis occurred” (Nyberg 2011: 11).

Yes, you read that correctly.  Nyberg ignored accounting control fraud, the leading cause of modern financial crises, but he knows without studying the role of fraud that it could not have been an important contributor to “why the Irish financial crisis occurred.”

Here’s one arcane but critical example of the analytical incoherence that became dominant as soon as Nyberg excluded any consideration of fraud.  Readers will recall that the fourth “ingredient” of the accounting fraud “recipe” for a lender is to provide only trivial “allowances for loan and lease losses” (ALLL).  The international accounting rules have come close to making this a perfect crime through an obscene provision that is being interpreted in a manner that greatly aids accounting fraud.  This is a doubly perverse irony because the international accounting standard is being “sold” as vastly superior to U.S. generally accepted accounting principles (GAAP) because the international standard is a “principles-based” rule that is supposed to prevent evasion and ensure voluntary compliance by (presumed) highly ethical audit partners by being interpreted to achieve “the purpose of the rule.”  The purpose of the particular rule (IAS 39) on the ALLL is to prevent accounting control fraud (through the creation and manipulation of “cookie jar” reserves for losses).  The reality is that IAS 39 is being interpreted to make lawful the vastly larger and more damaging form of accounting control fraud exemplified by “recipe.”

The ALLL is supposed to force a bank to reflect currently the losses inherent under its current lending standards.  This is not only logical, but essential if one is to prevent the three “sure things” provided by the fraud recipe.  By making bad loans today that will cause severe losses several years later the officers controlling an accounting control fraud guarantee that (1) the bank will report record profits in the near term (and for many years if they hyper-inflate a bubble), (2) the controlling officers will promptly be made wealthy, and (3) the bank will eventually suffer severe losses.  The title of George Akerlof and Paul Romer’s famous 1993 article (“Looting: The Economic Underworld of Bankruptcy for Profit”) captures these three “sure things.”  The reality is that the bank suffers large losses when it guts its underwriting standards to make billions of dollars in poor quality loans.  If the bank has to recognize this fact at the time it makes the loans the controlling officers’ fraud scheme will fail because the ALLL will be so large that the bank will report losses at the time it makes the bad loans.

The controlling officers of banks, however, cause them to commit multiple forms of accounting scams.  The “cookie jar” reserve scam is another means to inflate the controlling officers’ compensation.  If the bank has an exceptionally good year (or reports through accounting fraud that it has done so), the CEO may decide to “bank” the profits in excess of those needed to be recognized currently to maximize the CEO’s bonuses.  The CEO does so by increasing the ALLL.  When the bank’s (real or inflated) future profits fall to a level that would reduce the CEO’s bonus he can reduce the ALLL and the bank’s reported income will increase and max out his bonus.  The CEO uses the ALLL like a “cookie jar” that he can stick his hand into whenever he desires and pull out sweet treats for him.

Unfortunately, the drafters of IAS 39 focused solely on the cookie jar fraud scheme and ignored the far more destructive fraud “recipe” scheme.  The drafters created a rule that auditors for the control frauds were delighted to interpret as banning allowances for future loan losses in excess of losses currently acknowledged by the bank.  Given the well-known fact that “a rolling loan gathers no loss” this was either madness certain to encourage accounting control fraud epidemics and hyper-inflated bubbles or willful perversion of an accounting standard purportedly designed to prevent fraud into a “free pass” for accounting control fraud.  Contrary to the vaunted alleged advantages of the international accounting standards over GAAP, auditors for the banks engaged in accounting control fraud never interpreted IAS 39 in a “principled” manner to require loss reserves consistent with lending risks and the drafters of IAS 39 failed during the crisis – and to this day – failed to clarify the rule in a principled manner or to revise the rule to remove the invitation to fraud.  The international accounting standards setters have made it clear that they recognize that IAS 39 is being used to commit massive fraud and that they are troubled by this fact and would like, some decade, to fix the problem.

Here’s how IAS 39 aided the Irish accounting fraud – and how the Nyberg report bizarrely interpreted the CEOs as the wannabe heroes of the story.

“In the benign economic environment before 2007, the banks reduced their loan loss provisions, reported higher profits and gained additional lending capacity. The banks could no longer make more prudent through-the-cycle general provisions, or anticipate future losses in their loan books, particularly in relation to (secured) property lending in a rising property market” (Nyberg 2011: 42).

First, the Irish economic environment “before 2007” was not “benign” – it was the largest bubble proportional to GDP of any developed nation.  The economic environment was sure to be catastrophic for Ireland’s economy, the banks, and the Irish public – but exceptionally profitable to the senior officers leading the banks.  This was the classic period in which “a rolling loan gathers no loss.”  Nyberg is well aware of this, as one can see in his phrase about the lag in recognizing real estate lending losses that fraudulent CEOs cause in “a rising property market.”  Note that by mid-2006 the Irish residential real estate bubble was no longer growing and by late 2006 residential real estate prices were falling.

Second, the Irish banks that were accounting control frauds never made “prudent through-the-[business]-cycle provisions [ALLL] during the run up to the crisis – even before Ireland junked GAAP and adopted IAS 39.  They didn’t because their CEOs did not want them to recognize that their bad loans were (net) losses at the time they were made in 2005-2007.

Third, we can test Nyberg’s implicit claim that the bank CEOs wanted to make “prudent” (far higher) loss reserves but were stymied by their mean auditors who forced them to make imprudent loans.  If you were running an honest, “prudent” bank and you were told that IAS 39 was being interpreted in this perverse manner you would have had two obvious means of making only prudent loans.  You could have stopped making loans that required much higher ALLLs to avoid “through-the-cycle” losses.  Indeed, you could have stopped making the grotesquely imprudent 100% loan-to-value (LTV) and non-amortizing home loans with at least ten years of “interest-only” payments that were sure to cause catastrophic default rates as soon as the bubble stalled.  By making prudent loans the banks could have dramatically reduced the required ALLL.  Instead, the worst Irish banks not only vastly increased the percentage of loans they made at (and beyond) 100% LTVs on interest-only terms, but also increased enormously their far more imprudent commercial real estate loans to a tiny (ultra-concentrated) number of borrowers whose properties rarely cash-flowed.  The required ALLL for the new form of lending was certain to make the already farcically inadequate ALLL under IAS 39 ever more grotesquely inadequate.  “Inadequate” is an inadequate word to describe what a bad joke the worst banks’ ALLLs became.

The other obvious option if one wished to test the logic of Nyberg’s implicit assumption that the bank CEOs wished to make “prudent” loans with prudent ALLL but were stymied by the crazed auditors is that the bank CEOs could have greatly increased their capital.  The reality is that the CEOs of the worst banks: (i) were virulent opponents of the Irish financial regulators’ weak rule that increased capital requirements, (ii) created fake capital by lending money to straw purchasers of stock, and (iii) reduced their capital by paying dividends even in 2008 and through huge compensation payments to senior officers.

There is an obvious alternative explanation for why the senior bank officers would have welcomed grossly inadequate ALLLs.  When the ALLL is reduced the bank’s (fictional) reported income is reduced by the same amount.  As Nyberg conceded (but ignored):  “The higher reported profits also enabled increased dividend and remuneration distributions during the Period. All of this led to reduced provisioning buffers….”  Nyberg has, of course, reversed the order of the story.  The “reduced provisioning buffers” (ALLL) led dollar-for-dollar to “higher reported profits.”  The CEOs would have retained those (fictional) reported profits and increased their (fictional) reported capital if they were seeking to establish (faux) “prudent” capacity to prepare their banks to bear the massive losses they knew were inherent in their loan portfolios.  Instead, as Nyberg concedes, the CEOs continued to pay substantial dividends and pay themselves exceptional compensation.

Nyberg, however, presents the issue in a fashion designed to make it appear that the CEOs and toothless outside auditors were helpless champions of their banks’ best interests.

“From 2005 the banks’ profits, capital and lending capacity were enhanced by lower loan loss provisioning while the benign economic conditions continued. As the global crisis developed from mid-2007, the banks were constrained by these incurred-loss rules from making more prudent loan-loss provisions earlier, and the auditors were restricted from insisting on such earlier provisioning” (Nyberg: 55).

Note that the inflated reported profits flow through to higher reported (albeit fictional) capital which allows them to grow more rapidly by making more lousy loans (the first two ingredients of the accounting control fraud “recipe”), which produces higher reported (albeit fictional) income and increased dividends for the shareholders and compensation for the controlling officers.  And the cycle continued, as Nyberg implicitly concedes.

“As a consequence of not making this level of loan loss provisions, increased accounting profits effectively provided additional capital of up to €3.5bn to the covered banks. This, in turn, increased their capacity to lend by over €30bn” (Nyberg 2011: 43).

Nyberg documented how criminogenic IAS 39 is.

“The composite provisioning level for the covered banks at end 2000 was 1.2% of loans…. If this 1.2% provisioning level had been applied at the 2007 year end by the covered banks, aggregate provisions would have increased by approximately €3.5bn (i.e. from the €1.8bn actual to €5.3bn)” (Nyberg 2011: 43).

To make this understandable, we need to consider that Irish banks’ average ALLL of 1.2% as of 2000 would have been equivalent to the U.S. ALLL in 2006.  That was the lowest the ALLL had been in the U.S. since the Savings and Loan debacle – a less than propitious goalpost.  Nyberg presents 1.2% as if it were evidence of a prudent ALLL instead of a grossly inadequate ALLL.  This correction also allows the reader to understand that the ALLL level became farcical under the interpretation of IAS 39 that the Irish bank CEOs and auditors applied under a supposedly “principles-based” accounting standard that was supposed to be interpreted in a manner that prevented rather than aided accounting control fraud.  The Irish ALLL fell to roughly 0.4 percent – actual losses at the worst Irish banks ran about 100 times larger than the ALLL.

At this juncture, Nyberg goes completely off the track and displays even broader analytical failures that arise from his failure to understand accounting control fraud or logic.

“In the competitive market, many property loans were made at margins of less than 1% per annum. A composite year end provisioning level at the 2000 level of 1.2% might have caused the banks to reconsider the amount of low margin property lending and might have led to more appropriate pricing for risk” (Nyberg 2011: 43).

In actual “market[s]” as opposed to faux “competitive markets” property loan “margins” are well above “1%” because lenders would lose money on any portfolio of even extremely low-risk property loans with such low margins.  The worst Irish banks, however, were making exceptionally risky property loans.  Their ALLLs and margins for such extremely high-risk property loans should have been at least ten times higher than those Nyberg reports.  Consider Nyberg’s (implicit) claim that the margins fell to less than 1% (presumably by 2007) while the (grossly inadequate) ALLL provision typical in 2000 was 1.2 percent.  Remember, the Irish banks’ loans in 2000 were vastly lower risk than the loans they made in 2005-2008.  If we take Nyberg’s assumption that the bank CEOs were seeking to make prudent loans with prudent reserves and margins then Nyberg’s description of what the bank CEOs actually did refutes every aspect of Nyberg’s efforts to shill for the CEOs.  “Banks” do not “reconsider,” bankers make decisions.  Nyberg concedes that the worst Irish bank CEOs caused their banks to make incredibly risky loans, for spreads of “less than 1%” with ALLLs of 0.4% – three characteristics that, individually, were certain to produce enormous losses, but were in fact committed simultaneously.  This only makes sense if the CEOs of the worst Irish banks were running control frauds.

Nyberg concedes six pages later that what was really driving CEO behavior at the failed Irish banks (though he gets the description dead wrong in a telling fashion).

“High profit growth was the primary strategic focus of the covered banks…. Since the potential for high growth (in assets) and resultant profitability in Ireland were to be found primarily in the property market, bank lending became increasingly concentrated there” (Nyberg 2011: 49).

What Nyberg actually means, of course is that “high [reported, albeit fictional] profit growth was the strategic focus” and that the “resultant [reported, albeit fictional] profitability [was] primarily in the property market.”  Recall that six pages earlier he conceded that the banks were lending on terms that were certain to cause catastrophic losses and that only the hyper-inflation of the residential and then the commercial real estate bubbles postponed the recognition of these facts.  Nyberg seems to understand this, for his next passage reads: “The associated risks appeared relevant to management and boards only to the extent that growth targets were not seriously compromised” (Nyberg 2011: 49).  To concede that the “risk” of the loan was “relevant” to the bank CEOs “only to the extent that growth targets were not seriously compromised” is to admit that the CEOs were leading an accounting control fraud rather than an honest bank.

Nyberg also concedes that it was this pattern of lending without regard to risk (which is to say without regard to actual profitability) that hyper-inflated the Irish property bubbles and provided the pretext for continued growth of the Irish bank frauds.

“Thus, banks accumulated large portfolios of increasingly risky loan assets in the property development sector. This was the riskiest but also (temporarily) the easiest and quickest route to achieve profit growth.

Credit, in turn, drove property prices higher and the value of property offered as collateral by households, investors and developers also” (Nyberg 2011: 50).

Nyberg does not cite and shows no understanding of Akerlof and Romer’s point that accounting control produces a “sure thing,” but the necessary implications of Nyberg’s findings are that the fraudulent CEOs exploited that “sure things” and hyper-inflated the bubble by making ever crappier loans.


One response to “Irish Fish, and Banks, Rot from the Head

  1. Pingback: Irish Fish, and Banks, Rot from the Head | New Economic Perspectives - brought to you by Ephi Ullmann