BASKET CASE OF A BASKING SHARK: ANGLO-IRISH BANK

If investment banks were fishes they'd be sharks, never stopping in their search for profitable prey, ready to eat anything that looks or smells remotely tasty. But, if traditional banks were fishes they'd be toothless basking sharks, constantly hoovering up tiny payments and passing out loans through their gills. Anglo-Irish Bank, in these terms, called itself a basking shark, very common off the west coast of Ireland, but was it?
On the 9th September last, where else but at Lehman Brothers Financial Service Conference, Willie McAteer, Finance Director and Chief Risk Officer of Anglo-Irish Bank, presented a glowing account of his bank, watch-words, "balance sheet: old fashioned banking", "we lend against cash-flows not asset values", and with ratio statements like "customer loans €69bn, permanent funding €80bn" (Euro billions), customer and term funding = 116% of loans, "no requirement for external equity capital", core equity = 7%, "resilient and strong business model". McAteer also vouchsafed, "Significant market shift to “Balance Sheet lenders”, the upside from "Changed competitive landscape –exit of non-bank lenders", "25 years of performance through cycles –strong asset quality culture", a "Franchise strength with huge organic potential in existing markets" and a "Long established management team", that gives "Consistent delivery for shareholders". And the bank's strapline motto: "There is a difference!" The difference was that all the above statements were outright lies, and had been lies for years, and not just because of what happened later. Days later, the bank was staring at a run of Eur5bn in deposits heading for Eur12bn, when Lehmans went bankrupt, 3 days after its splendid conference, and Anglo-Irish had to be nationalised (of all banks in trouble after Lehman, only this one was 100% nationalised, and Fortis, except for parts to be sold to BNPP), after a short agonising debate within the Irish Government of whether to let the bank collapse or not. And, of course, all the shareholders were wiped out and the top executives and board terminated. A law was passed to allow criminal prosecutions and the Gardai (police) went in a few weeks ago.
Those of you who know about regulation will have spotted the obvious breach of CRD law that McAteer was both Finance Director and Chief Risk Officer! Pity the IFSRA didn't spot that one?
Let's now look at just why all the above statements by McAteer were blatent lies. The answers are to be found in a strictly private and confidential report for the IFSRA that has come into my hands today via the web-site of Anglo-Irish bank International. I've rarely had to read something so oddball as this PwC report in its reluctance to call a hook a hook, but it manages to show what the truth was, if you know how to read between the lines. Also, what is said of the bank in my opinion also clearly applies to HBOS's corporate loans property portfolio. The report implies Anglo was not risk-averse enough because the bank's business model seemed to perform well for years. It may have survived downturns before but was swimming into this one with foolish self-confidence.
Any risk-diversification in the bank's loans was merely by types of property, and property by business sector. Collateral was personal assurances & guarantees, but fallen in value. €2bn of treasury assets for sale alone needed writing-down to an extent that would wipe out at least 2 yrs profit. These Banking Book Assets (Available-for-sale) of RMBSs, ABSs, CDOs were €1.9bn difficult to value i.e. illiquid in current market. Impairments were absurdly recorded at below 1% of loans - easy to cover - probably because collateral had not been risk-checked and continued to be booked as of 'senior' high quality only. At the same time it was clear that the bulk of borrowers of big ticket loans are highly leveraged property developers, hence why the bank's covered bonds require 57-100% over- collateralization to get a AAA rating (400 loans at an average of €15m). By the 27 Sept. '08, Anglo forecast (a straight line 4 week projection) net negative cash-flow of €12bn by 17 Oct.'08. The reason was €10bn fall in corporate and retail deposits. PWC also says that in the last full week of September, customers withdrew €5.44 billion from the bank and the bank’s management was expecting a further €6.6 billion to be withdrawn by mid-October. This projection was based on €5bn fall in corporate deposits and €440m in retail deposits in the week Lehman Bros. failed? All the bank could do was hope to sell a securitisation of €2.2bn and to bid successfully in swapping assets for ECB treasury bills. Out of €70bn deposits, €20bn could be expected to be lost at a rate of €5bn per week for a month. The inter-bank and debt capital markets were effectively closed and projections had to assume they stayed closed, ushering in what we risk experts call "tail risk". So, on 28th Sept. Anglo phoned the central bank to say we're insolvent; you've got to help us and PwC was sent in for a 9 day high-level look-see. Pro-forma profit of €1.5bn was to be cut by half for impairments. Much of property development loans were non-performing currently. This was because interest was "rolled up" in many big loans to be paid from "future expected cash-flows". Is this what McAteer meant by lending on cash-flows, not assets? If so, he had taken a prudential risk mitigator and turned it into an extreme-risk speculative gamble. Anglo also lent on an interest-only basis against cash generative investment properties that depended on risk assessment of capital repayments from asset sales, or otherwise on roll-over refinancing. Hence Anglo's model was attractive to developers because they could roll debt from one project onto the next. In fact the model encouraged or even forced them to do so. Developers were forced to become sharks on Anglo's behalf. Anglo was a shark of sharks involved in direct commercial risk-sharing with its property developer customers.
This means Anglo would need long term capital sources.
The Annual Report for 2008, revealed that €451 million was advanced to 10 customers to buy shares amounting to 10% of the equity, which became available from the unwinding of a so-called contracts for difference stake, which was built up by billionaire businessman Seán Quinn. The bank said €83 million has been repaid but €300 million will be charged as bad debts in the interim accounts to March 2009, including a portion in respect of directors' loans. The "Golden Circle" of 10 investors, lent with less than 20% security and without recourse. Investment lending accounted for almost two-thirds of the bank’s loan book at 30 September 2008 of €74 billion with a geographical breakdown of €43.6bn in Ireland; €21.3bn in the UK (including N.Ireland) and €9.4bn in the US. Development & Land accounted for approximately a quarter of the total. The total amount lent to directors during the 2008 financial year was €255 million, of which €122 million was money lent to then chairman Seán FitzPatrick, who at the end of September 2008, for the eight time, hid his loans via an arrangement with Irish Nationwide Building Society.
In February, senior managers of Irish Life & Permanent, were forced to resign following a disclosure, that more than €7 billion had been placed on deposit at Anglo Irish in September, including €4 billion, which was lodged with Anglo Irish on September 30th, the last day of Anglo's financial year and the first day of the State bank guarantee. The deposit was treated by Anglo as a customer deposit to mislead investors and market analysts.
The report of extracts from reports produced by PwC, following the issue of the State guarantee to six Irish financial institutions on September 30, 2008, shows that Anglo Irish Bank had loans outstanding to about 15 customers in excess of €500 million each.
Apart from these outlandish relationship-banking deals, the margins it charged and factored in should have been high enough to cover the riskiness of its own model, not just the riskiness of the borrowers and their projects. The bank should have had extremely sophisticated property market analysis & forecasting. Did it? We know that the Irish central bank wrote papers in '05/'06 to show how extremely vulnerable the whole economy was to small rises in interest rates - these alone would trigger recession - had Anglo considered that context - had it any plan in place for cutting back its exposure, to become a lot less risky if it needed to - probably not. If anything it seems truly to have believed the next 10 years could and would repeat the past 10 years - hubris is not enough of a word to cover that? Development loans (no interest in the construction period, only a fat fee for the bonus money) are converted into investment loans when a property development is either sold or let and becomes revenue earning. The loans are retained on the Bank’s books post-completion of the buildings as not all get sold or tenanted and not all balloon-payments paid off, because then the outstanding loans are rolled over from one speculative project to the next project. This is an effective way whereby banks who lend like this are actually forcing up property prices and rents faster than the market would do so if the financing costs were not allowed to become cumulative from one project to the next and the next after that.
Banks who are thereby directly in property development business are effectively using the borrowers as high-earning front-men, and using the financing structure to push up property values! Equity-release is used to roll-over outstanding debt from project D to Project E to Project F so in time each subsequent project becomes burdened with more and more insupportable debt inherited from past chain of projects. This relies increasingly on property values and rents rising (even accelerating). Equity release loans include lending on the difference between past 'book-value' of land (price of bought with 'option') and current market values pre-development. Thereby, substantial collateral for loans can be based on an 'option' to buy, not on actual purchased land. This is equivalent to lending on derivative contracts and is a form of extreme leverage. In reality, land requires investment to justify its development land price, and at the very least the options should have been exercised. And of course the cost of putting infrastructure services into greenfield sites is usually enough to eat away most of that equity margin. Anglo built up strong relationships with its key customers. Anglo's now ex-management said that their strategy is deliberate one of developing deep relationships with who it deemed to be the strongest property operators. From PwC's review of the larger loans in the portfolio it is evident that a small number of key customers were involved in a large number of transactions and represented a "significant proportion of the loan portfolio" meaning too large, meaning "concentration risk". Anglo considered itself "able to attain a thorough understanding of its client’s business, finances and relevant risks, which are continually reassessed in face to face client meetings often held weekly", for which read great lunches somewhere nice, maybe at the Curragh or a yacht down at Dun Laoghaire or somewhere foreign much more jet-setting. These friends did not get put on the watch-lists. PwC noted that there are a number of customers which are not currently on the Bank’s watch0list, 'notable' or 'impairment' lists, "all of whom exhibit potentially serious short term liquidity issues".
This has the strong whiff of corruption! These 'relationships' were not subject to standard risk management, but privileged (an exclusive risk grade outside of any normal grading) by virtue of a mutual-interest collusion between bank and developers. But who were the Mr Bigs, the bankers or the developers? The watchlist was updated eventually only just before Anglo went to beg for help and mercy from the central bank. i quote, "...the Anglo model is dependent on customers’ ability to successfully refinance significant development and investment loan portfolios in the short to medium term. This is exacerbated where (i) Anglo is the lead bank in a wider syndicated loan (other banks can sue it for its underwriting assurance?) or (ii) significant additional debt funding would be required for the successful build completion to derive value from development land banks held by key customers. While the stress scenarios applied by management assume no new net lending for 2009 and 2010, it is assumed Anglo will successfully re-finance its own short term borrowings in that period." Note, by the way, that Anglo did participate in syndicated loans when McAteer implied it did not. It did, but as lead-bank, it sold syndicated loans. And, why, just to get in more to lend, not to spread the risk. "...carrying cost of assets may not exceed their economic value even if they are realisable and as a result the option of interest roll up may not be available to customers to the same extent as to date. In addition, while demographics remain favourable, the continued unavailability of mortgage funding and increasing unemployment may exacerbate already reduced demand for residential property resulting in a fall in price for units already built and less demand for new developments on the land banks held by Anglo’s clients. The retail trade is also struggling and if the difficult economic conditions continue into the medium term shopping centres and retail developments may begin to experience trading difficulties or not be developed. A number of Anglo’s customers have significant exposures to the retail sector ...accounting for (20%) ...of the Bank’s loan book ...may not lead to impairment in Anglo’s 2008 results, a further deterioration in market conditions could lead to a reduction in the discounted cash flow attributable to certain assets ... the Bank adopts impairment recognition policies as prescribed by ...(IFRS). The Bank has ...15 relationships in excess of €500 million. ...Bank considers that in all cases they are supported by diverse portfolios of assets underpinned by material contractual cash-flows and with significant personal/ corporate recourse."
These are not suddenly new risks - the bank should and probably did know. What this is trying to say is that Anglo was afraid to foreclose on borrowers for fear of a domino effect that would come back on the bank and expose its relationship-banking model. Even if LtVs are way above 100% of the net property portfolios, and or way above whatever the bank can get a hold off after tax authorities and others, it also means that Anglo could not just seize the cash-generating assets for itself in the event of default-risk and therefore the collateral. The security contracts were either not well-enough written and were therefore not worth the collateral value claimed? And these were borrowers who are almost always short-term cash-flow insolvent and in recent year or two already desperately trying to sell their properties and failing, into an increasingly illiquid market. relationship- - if a bank's officers and its customers are friends, but turn out also to be relatives or even same persons, then what? At least in Ireland when a bank collapses the regular depositors and household account holders are 100% protected. If the security on which some loans are secured was marked to market significant shortfalls would occur in line with the rest of the Irish and international banking sectors in current market conditions. Whilst the bank's lending model/underwriting standards rely in the first instance on contractual cash flows, collateral values are key in the event of default leading to a forced/ distressed sale. In accordance with IFRS, Anglo 'amortises' these cash-flow securities as IFRS does not permit a mark to market approach for such lending assets. The Bank has exposures to a number of customers who also have significant exposures to other domestic and foreign banks. There is a risk that (i) if other banks call in their loans customers may not have the resources to pay these and Anglo’s loans or (ii) these other banks may not participate in new rounds of refinancing. Anglo had a portfolio of available for sale financial assets with a fair value of €8.2 billion as at 30 September 2008 which reflected impairment and mark to market losses of €288 million in FY08. There was an AFS reserve of a negative €589 million held on balance sheet at 30 September 2008. The Bank use a number of sources to derive the fair value of the assets carried in the Available For Sale (AFS) balance sheet category. 75% of the prices for these portfolios come from Bloomberg with the remainder coming from external service providers and broker quotes. There is some uncertainty in the market about pricing for certain asset backed securities given market conditions. At 7 November 2008 the Bank was still seeking prices in respect of €978 million of AFS assets and expects to obtain third party prices for approximately 50% of these. Currently prices are based on the most recent counterparty prices.
But, the report is not explicit about Anglo's roughly €50bn 'funding gap'. Its roughly €50bn deposits may have evaporated by over half, hence it has a €75bn funding gap and only €8bn available for sale assets! So much for the "permanent funding! that McAteer claimed, a term that no-one would or should respect. Assume €50bn commercial development assets that have fallen 50% in market value the bank has a decade's worth of profit wiped out. The bank's assets were €101bn but RWA as high as €86bn! Top 70 borrowers = 17% or €12.6bn much of which are mothballed projects.
The state has nationalised something with a book of say €75bn that needs 100% refinancing and will cost €25bn to repay deposits. One strategy could be to hand over the deposits to other banks and give them €25bn of long term bonds secured on the rump of the bank, now deemed a 'bad bank' to be 'worked out' over 10 years, but nationalisation is only permitted by the EU for two years - that is a short time to redeem this bank's solvency?
Like fishes all banks swim underwater mostly all of the time. Solvency requires confidence. Can confidence be restored in 2 years for the bank to be wholly privatised - no. At best it will be sold off to another bank or only partly privatised at the end of 2010.