NAMA was required to deliver to the Department of Finance its quarterly and annual management accounts for 2011 by 31st March, 2012 so it is likely that they are now resting on Minister for Finance Michael Noonan’s desk. Last year the annual management accounts were published on 4th May – it is unclear on here why there is any delay at all in the Department of Finance, but they have other tasks occupying their time – for example, we’re still waiting for an explanation for the €3.6bn error in the national accounts uncovered at the start of November 2011. In addition to NAMA’s management accounts, which will be published shortly, we will get an annual report but we must wait until the summer for that. Both the latest management accounts and the annual report will, for the first time in 2011, make provision for what NAMA calls its “impairment charge” – the management accounts will have a preliminary estimate and the annual report will have the final figure.
First up, what is an “impairment charge”? This is what the NAMA 2010 Annual Report says:
And although the above makes reference to two other sections of the Annual Report, there is in fact precious little justification for the impairment charge.
To help explain the “impairment charge”, NAMA bought €74bn of loans from the banks and paid €32bn for them. The €32bn was NAMA’s valuation of the loans by reference to 30th November 2009 and included that controversial premium for so-called “Long Term Economic Value”. Normally when valuing a loan, you would look at the borrowers capacity to repay the loan. If you think the borrower will be unable to repay the loan, then you need look at the underlying collateral or security, which in the case of NAMA is the property. You may also look at other factors such as the borrower’s income, guarantees and other collateral, but if when you take all the factors into account, you estimate that you will not be able to recover the full value of the loan, then you are required under accounting rules to estimate the shortfall, and that is the definition on here of “impairment charge”
Now consider the following two illustrative examples for valuing a loan. If you’re a bank and you loan Mary €100,000 to buy a house today that is valued at €150,000, and then a couple of months later after Mary has paid €2,000 off the loan principal, she loses her job, has no savings and can’t pay another penny to the bank; how much is the original €100,000 loan worth to you as the Banl? In this case, it will depend on the value of the property and because the bank insisted on a €50,000 deposit meaning the original loan had a “Loan to Value” of 67%, and because only a couple of months have expired, it is likely the loan is still worth its full value because if Mary sells the property, there is likely to be more than €98,000 of proceeds on which the bank will have first dibs.
But take the second example where John borrows €100,000 for a €150,000 home and in two years John repays €20,000 off the principal but then loses his job and is unable to pay any more, but this time the property has halved in value to €75,000. So this time, the bank doesn’t value the loan at €80,000 – John having repaid €20,000 off the original loan – but values it at €75,000 because that’s all the banks thinks it can recover. And the bank takes an impairment charge of €5,000 being the difference between the amount outstanding on the loan, €80,000 and what the bank now expects to recover from selling the property, €75,000.
In the case of NAMA, we have a general split of NAMA’s loans by property type and location. We also have indices from various organisations that tell us how property performed since November 2009. So we can have a stab at how much NAMA’s loans are worth today, though without access to the detail of the loans and other information which is available to NAMA, we have to make a large number of assumptions, but the hope on here is that the results give a sufficiently accurate ball-park estimate, that we can have a basic discussion on the implications of the impairment charge for NAMA.
Property price changes since November 2009
(1) Residential for Dublin and rest of Republic of Ireland, both come the monthly CSO Residential Property Index
(2) Residential for Northern Ireland and London, both come from the quarterly Nationwide Building Society house prices report. There is a problem with this because the reports are quarterly ending on the usual quarter days and we need get an index which is at 30th November 2009. So it is assumed that 30th November 2009 was priced at an even linear point between Q3, 2009 and Q4,2009 – in other words, you start out with the Q3,2009 prices and the Q4,2009 prices and assume that 30th November 2009 is two thirds of the way along the trend between the two prices. Nationwide doesn’t provide a “rest of Great Britain” price and there is considerable disparity between regions. Because London was the strongest performer over the 25 months and Northern Ireland was the weakest, it is assumed that rest of GB is half-way between the two
(3) Commercial – retail, office and “other investment” – Ireland– Jones Lang Lasalle quarterly commercial indices and as with Residential for theUK, assume that November 2009 is two thirds along the trend between Sept 2009 and December 2009. “Other investment” is assumed to be industrial property. JLL doesn’t provide a sector index for hotels so the assumption is hotels performed in line with the average of retail,office and industrial.
(4) Commercial – retail, office and “other investment” – London, rest of Great Britain and Northern Ireland – IPD monthly index, this is not analysed by region so starting assumption is that all UK regions performed in line with the average in the IPD reports. IPD doesn’t provide a sector index for hotels so the assumption is hotels performed in line with the average of retail,office and industrial.
(5) Development Ireland- NAMA apparently didn’t acquire development property which had a negative value in November 2009. That’s important because if you have a part-complete estate and the only viable option is to demolish existing property and return the land to agricultural use, the average demolition/disposal cost in Ireland for a demolition is north of €20,000 per property. So you can see how an unviable development could have a negative value when the resultant agricultural land is only worth €8-12,000 an acre on average right now. There is no Irish index which tracks the value of development property which is assumed to be partly developed property, rather than a field. Savills said in February 2010 that development land had fallen 75-90% in value from peak – the Savills report no longer appears to be online but a subsequent report in June 2011 (registration required) said “estimated values have slipped further.While most of the drop in land values was evidenced quickly between 2008 and 2010-with values falling between 75-90%-lack of activity has pushed even prime values back closer to minus 90%.” I am interpreting this to mean a 10% decline in prices from November 2009.
(6) Development UK– there isn’t a development land index in the UKeither, and it is not clear what the development land comprises, whether it is residential or commercial. Knight Frank publish quarterly research on residential development property in London and the rest of the UK. For 2011, it says prime London rose by 20.3% and the UK overall rose 1.3% and for 2010 says prime London rose 15% and “England Urban Land” rose 8%. It is assumed that NAMA’s London development land is “prime” and that the rest of UK/England Urban Land is appropriate for Northern Ireland and “rest of GB”, so London rose 40% and rest of UK rose by 9%. I know that Northern Ireland development land has fallen by as much as the Republic’s, but am not aware of any index that captures this by period.
(7) Land Ireland– NAMA says that in some cases in Ireland, it valued land it was acquiring at agricultural values. According to the Farmers Journal, agricultural values have dropped 4.3% in 2010 to €8,420 but the average price paid had bounced back 20% in 2011 to €10,024. The assumption is that ALL land was acquired by NAMA at agricultural values, though this is not the case, and undoubtedly NAMA will have paid a development-potential premium in some cases on the basic agricultural value. So the assumption is that NAMA’s land has risen 11% since November 2009 (€8,420 plus 4.3% compared with €10,024.
(8) Land UK– Although NAMA may have paid agricultural prices in Ireland, it is not clear how land was being valued in the UK. However the Knight Frank UK quarterly farmland review shows prices at GBP 4,973 in September 2009, GBP 5,123 in December 2009 and GBP 6,049 in December 2011. Assuming November 2009 prices were GBP 5,073, then there was an 19% increase to the end of December 2011, and this is what is used.
(9) Rest of World property – NAMA said in October 2011 “In terms of our overseas assets, we have approximately $600 million worth of assets in theUS. The majority of them, by value, are in New York. …Our main assets in continental Europe, by value, are in Germany, France and Portugal and we have smaller pockets of assets, ranging from €10 million to €20 million, in Malta, the Czech Republic and Poland. Collectively, these add up to €120 million.” I also believe NAMA has some other national exposures to Bulgaria,Cape Verde and South Africa. I am not proposing to delve into any of these countries individually though I know Poland has performed well, there are pockets in the US that have performed well, though there are pockets which have declined. Because NAMA can’t recognise gains – see below – the markets which have performed strongly are not of immediate concern to this exercise. For the time being, I have assumed 0% change since November 2009 for all property types in all territories that comprise “Rest of World”
Assumptions about loans in November 2009
(1) That developers had no equity remaining in property subject to NAMA loans. “Equity” means an excess between the value of a property and the loan on the property, so a €150,000 house subject to a €100,000 loan has €50,000 of equity. We know that banks were lending at high LTVs during the boom, and given that developers borrowed the maximum available and tended to top up loans when property prices increased, the perception is that NAMA developers have no equity left in their loans – that is, they are like John above.
(2) Unlike John, these developers may have other businesses, cash deposits, other property and personal guarantees that might have been given when the loan was obtained, so although the underlying property might not pay off the loan, there may be other assets pledged by the borrower which might pay off any residual balance owing after the property is sold, or which might mitigate the loss. This is ignored in the analysis below because (1) we don’t know how to quantify the value of additional assets and guarantees and (2) in the context of tens or hundreds of millions of euro loans, these assets are not understood to be substantial – there’s “no pot of gold” according to the NAMA CEO. However NAMA claims to have reversed €160m of asset transfers so there might be some mitigation there, though it remains unclear if these reversals were agreed in the context of NAMA making additional advances to the developers.
(3) That the price paid by NAMA for the loans was the November 2009 value of the underlying property PLUS a 9% premium for so-called “Long Term Economic Value” (LTEV). We know that NAMA’s average LTEV premium on the first five tranches of loans was 9% – 8.8% to be precise. There is a caveat here – NAMA has failed to provide clear details of how it valued loans, and what we have is a mish-mash of partial NAMA statements and an unhelpful Comptroller and Auditor General report on NAMA’s acquisition process which included a sample valuation which was unfathomable to valuers and academics alike. But it does appear to be the case that NAMA deducted from the valuation of any given loan, a provision for legal and other costs which NAMA incurred in acquiring the loans. These provisions are being ignored here, not least because NAMA recharged certain legal, valuation and due diligence expenses to the banks.
(4) NAMA paid for the loans it acquired from the banks using two completely different IOUs. The first IOU is called the NAMA bond and it represented 95% of what was paid for the loans. The NAMA bond is guaranteed by the State and requires NAMA to pay interest of about 1% per annum to the holder of the bond and the NAMA bond must be redeemed by NAMA by 2020. The second IOU is called the NAMA Subordinated Bond and this represented the remaining 5% of what NAMA paid for the loans. Unlike the NAMA bond, the Subordinated Bond doesn’t have to be redeemed or honoured at all if NAMA makes a loss by the end of its life.
Accounting assumptions about the impairment charge
(1) NAMA is required to recognise impairment losses immediately but is not allowed recognise gains. For example, lets say NAMA paid €100 for a loan which had a face value of €200 and which is now worth €85, then NAMA must recognise the €15 loss now even though it will not be until the property is sold and NAMA has finished with its recovery action, that the true loss will be known – the actual loss might turn out to be more or less than €15. On the other hand, if the €200 loan which NAMA bought for €100 is now worth €120, then NAMA is not allowed under the accounting precept of “prudence” to recognise the gain until it is fully realised. The layman might say that this was unfair but the accounting principle of prudence requires you to take losses now, but not to count your chickens until they’re hatched. Remember also that as far as NAMA is concerned, any change in value to the loan is to be examined by reference to the €100 which NAMA paid, and not the €200 original loan value.
(2) NAMA is overstating the actual interest income being received in cash, for example for the first nine months of 2011, NAMA received €537m in cash in interest from developers, but the Agency booked €786m in its management accounts because it thinks it will recover some interest when the underlying asset is sold. In other words, NAMA assumed that €249m of interest that it told us was income would be received in future. What makes this deeply unsettling is the fact that NAMA assumes in its calculations that it will at least receive what it paid for the loans by reference to November 2009 values PLUS the LTEV premium. What has all of this to do with the impairment charge? It has been suggested that the impairment charge represents BOTH the decline in value of the loan AND a provision to reflect the fact that NAMA’s interest calculation overstates profit. The calculation of impairment here only addresses the first of these, in other words there is no provision in the impairment charge calculated here for overstated interest income.
(3) There have been no NAMA disposals. NAMA says that it has approved nearly €7bn in disposals though it has conceded that to the end of September 2011, it has only actually booked €2.7bn of disposals. NAMA says that 80% of its disposals have been overseas and it is understood most have been in London. NAMA has booked €132m of profit on the disposals. Because London prices had typically increased, and because gains are not recognised in the impairment charge, the effect of not deducting disposals from the total NAMA acquisitions should not be material.
The calculation of impairment
(1) NAMA acquisition value which is assumed to be November 2009 price plus 9% for the Long Term Economic Value
(2) Less 9% to deduct average LTEV
(3) Plus the change to property prices November 2009 to December 2011
(4) Equals the value of the underlying property in December 2011
(5) If property/location has risen in value, the difference is ignored as accounting rules say gains shall not be recognised in impairment unless the gains have been realised
(6) If property/location has fallen in value, then impairment for property/location is (1) less (4)
(7) Total impairment is the sum of property/location declines in (6) less the 5% of NAMA acquisition price – which totals €1.6bn – which was paid in Subordinated Bonds which don’t need be paid if NAMA makes an overall loss by 2020
So putting figures against these stages in the calculation of the impairment charge
(1) The NAMA acquisition price comes to a total of €31.81bn
(2) Stripping out (2) the LTEV and adjusting the November 2009 value to December 2011 value gives €27.98bn. This is what the original loans had come to be worth in December 2011, recognising both gains and losses.
(3) Stripping out (2) the LTEV and adjusting the November 2009 value to December 2011 value but ignoring any property/location gains gives €25.69bn. This is what the original loans had come to be worth in December 2011, recognising losses only -remember the accounting rules say you recognise losses only.
(4) The impairment is therefore €31.81bn less €25.69bn which equals €6.12bn but we add back the €1.6bn of NAMA’s subordinated bonds which won’t need be honoured if NAMA makes a loss, and that gives us €4.52bn.
(5) NAMA recognised an impairment charge in 2010 of €1.485bn, so the estimated impairment in 2011 will be €3.035bn which gives a total impairment across both years of €4.52bn.
The results and the implications
At the end of 2011, NAMA had nearly €600m of remaining capital – being the €100m of original capital PLUS NAMA’s classification of its €1.6bn of Subordinated Bonds as capital less its net loss of €1.1bn in 2010. The NAMA CEO said a fortnight ago that NAMA was set to make a profit before impairment in 2011 of “at least €750m” and remember this includes over-inflated interest income but even that will not be enough to save NAMA if it must book a €3.4bn impairment charge. The Agency will end up with at least €1.8bn of balance-sheet insolvency.
Now the question has been posed whether NAMA can operate in circumstances where it has a negative balance sheet or is balance-sheet insolvent. Although NAMA has the moniker of “bad bank”, it is not a bank in the sense of having an Irish banking licence. Which is interesting when you think about it – NAMA is advancing €1bn+ of new loans to developers for working capital and to finish out projects, which probably makes NAMA the most active lender in the State for property development. But if NAMA doesn’t have a banking licence, then it is not thought to be subject to bank capital adequacy rules. I’m sure Bank of Ireland and Ulster Bank which compete with NAMA would love to have such concessions! This is likely to become an issue and it is quite possible there will be a competition complaint to the European Commission.
Companies can lawfully operate in the Republic of Ireland even if they are balance-sheet insolvent. And it is understood that the ECB will continue to accept NAMA bonds as collateral for lending to banks even if NAMA is balance-sheet insolvent, as long as the Irish state guarantees the bonds, but this becomes very messy because the Irish state is presently avoiding classifying NAMA bonds as part of the national debt.
So NAMA’s balance-sheet insolvency may not be such an issue from a regulatory viewpoint. From a political viewpoint, will the Government countenance a balance-sheet insolvent NAMA? Maybe, remember this is the Government which continues to say that NAMA will recover any ultimate loss from the banks in the proportions of the originally acquired loans, in other words IBRC in 2020 will pick up over 50% of NAMA’s losses! But the Government will look like a bunch of clowns if NAMA does indeed have a €1.8bn negative balance sheet, and NAMA bonds are kept off the national debt.
2012 and beyond
None of us has a crystal ball that can forecast Irish property prices with absolute certainty. But I would suggest it would be a brave soul that would forecast an increase in the next year. And remember flat is not good enough for NAMA which needs to cover nearly €200m a year of operating costs plus its interest obligations, and the number of performing loans is now understood to be slightly below 20%. So I don’t see how NAMA or its political sponsors can accept a balance-sheet insolvent NAMA but claim that this is a temporary phenomenon because 2012 looks shaky also with the CSO saying that residential property has declined by 4.1% in Q1, 2012. We must wait another couple of weeks for the quarterly commercial property indices but the betting on here is that they will be negative.
Beyond 2012 becomes even more difficult to forecast. Economic forecasters say that the economy should return to 3% annual growth in a couple of years, but then again that’s always what long term forecasts look like – no-one forecasts for long term failure generally! We have issues with overhang, oversupply, vacancy, credit supply, weak demand, falling costs but as the economy eventually recovers these should be lessened, but what brave soul will quantify the future NAMA loss or profit?
So for now, I would say that NAMA faces a challenging year in 2012 and that it probably won’t see a recovery in prices, and whatever balance sheet insolvency it has at the end of 2011 is exacerbated in 2012. The implication is that balance-sheet insolvency can’t be dismissed as a temporary aberration, and there will be pressure to address the matter, probably with the injection of additional capital into NAMA, which will not be a politically-easy decision.