ANALYSIS:Government estimates of Nama valuations appear implausible, are out of line with other property collapses and may impose massive losses on the taxpayer
WHAT HAS been dismaying about the recent acrimonious exchanges over Nama is that neither side seems to feel it necessary to produce any evidence to support its assertions about its likely cost to the taxpayer. Like most discussions in Irish public life, the Nama debate seems set to generate more heat than light.
If we want to make sensible predictions on the likely course of Irish property prices over the next decade, we need to see what has happened historically in the aftermath of similar booms. In other words, we need to find property booms where sharp increases in bank lending caused real prices to more than double.
In Ireland, between 1995 and the peak of the boom in 2007, the average price of housing and commercial property roughly tripled, adjusting for inflation, while disposable incomes increased by one half.
Two previous booms fit this pattern closely: Japanese urban land in the 1980s, and Irish agricultural land in the late 1970s.
In Japan between 1985 and 1990, the real price of commercial land in major cities tripled, while the price of residential land doubled. What makes the Japanese case particularly relevant to Ireland, as I pointed out here two years ago, is that at the peak of their bubble, Japanese banks had the same extreme exposure to development and construction loans – 30 per cent of their lending – as Irish banks did in 2007.
As Japanese banks buckled under bad property debts, lending fell sharply and prices with it. By 2005 – 15 years after the peak – residential land had fallen back to its pre-bubble level, while commercial land had fallen by nearly 90 per cent. Given that many people are claiming that Irish property prices will recover once the economy starts to grow again, it is interesting to note that Japanese property prices collapsed while the economy continued slowly to expand: real output in Japan rose 20 per cent between 1990 and 2007 and did not fall in any year during this period.
The next case is much closer to home but almost forgotten: the boom and bust in Irish farmland prices in the late 1970s. After joining the EEC in 1973, Irish banks began to lend heavily to farmers. As a result, the inflation adjusted price of agricultural land tripled between 1975 and 1977, reaching a peak equivalent to €14,000 per acre in 2009 prices. Real Irish GNP in 1977 was about one third of its present level, so this price is roughly equivalent to €50,000 per acre in current purchasing power for land with no development potential. For comparison, during the recent boom, when agricultural land prices were driven by demand for potential development, prices peaked in 2006 at an average of €21,000 per acre nationally.
The bubble quickly burst as farmers ran into difficulties servicing loans: between 1977 and 1980 real prices fell by around 75 per cent, and remained at this level, more or less where it had started in 1973, until 1995, 18 years after the peak.
These examples illustrate a general principle: property bubbles are the consequence of abnormal levels of bank lending. Once the bank lending that fuelled the boom returns to its usual levels, prices return roughly to where they started before the boom.
In ordinary times, property prices grow at the same rate as national income: people in industrialised economies spend much the same fraction of their income on housing as they did a century ago.
However, a surge in prosperity, which drives property prices higher and encourages banks to lend more on appreciating assets, can lead to a self-reinforcing cycle of rising prices and rising lending.
Eventually, banks get a fright and return to levels of lending they used to regard as prudent, causing prices to fall back to where they were before the bubble. Just like Irish farmland in the 1970s, and Japanese property in the 1980s, our recent property boom was the product of unsustainable bank lending.
Between 2000 and 2007, while nominal GNP rose by 77 per cent, mortgage lending rose from €24 billion to €115 billion, lending to builders from €2.4 billion to to €25 billion, and to developers from €5 billion to €80 billion. Should the usual post-bubble correction occur in Ireland, it would suggest that real prices of residential and commercial property would return to their levels of the mid-to-late 1990s, two thirds below peak values.
Already the Irish property market has seen unusually sharp falls by international historical standards. The Sherry FitzGerald house price index is down 35 per cent nationally, and 42 per cent for Dublin; while the Society of Chartered Surveyors estimate that commercial property prices have fallen 48.6 per cent from their peak; and Knight Frank estimate that farmland prices, which were driven by their development potential, are down 45 per cent from their peak but are still twice those of comparable UK land.
Despite these large falls, which already exceed the one third haircut on Nama assets rumoured to be proposed by the Government, the property market remains moribund. Property transactions, measured by stamp duty receipts, are two thirds down on this time last year, and 80 per cent lower than two years ago.
In other words, if nobody is buying despite large falls in price, then price needs to fall considerably further to reach its long-run equilibrium.
The impression that Irish property prices are still considerably above long-term value is reinforced by rental yields: the ratio of the rent you get from a property to the price you paid for it. As many of you have discovered to your cost, property is a risky asset that performs particularly badly during economic downturns. To compensate for this fundamental risk, property should earn a long run rental return of at least 8 per cent.
Despite some of the highest rents in the world at the peak of the bubble (according to Lisney, Dublin ranked as the second most expensive location for industrial property and ninth for offices, with Grafton Street coming in as the fifth most expensive retail street on earth), new residential and commercial property was earning a paltry rental yield of 3-4 per cent.
This means that, to restore long-run equilibrium, prices needed to halve from peak levels, or rents to double.
Suppose for a moment that the Government’s assertions are correct, and the long-run value of Irish property is two thirds of its peak value. In order for rental yields to rise from an unsustainable 4 per cent to a long-run equilibrium of 8 per cent, the Government needs rents to rise one third from their already extreme peak values.
In fact, instead of rising, rents have fallen, and nearly as sharply as prices. The Irish Property Watch website estimates that residential rents have fallen by 32 per cent since May 2008; while Lisney estimate that commercial rents have fallen 24 per cent from peak, with office rents down 35 per cent and now lower than they were a decade ago.
Again, these large falls have not been sufficient to restore equilibrium. The number of rental properties listed on Daft.ie has risen from 5,000 at the start of 2007 to nearly 25,000 now, while the average time to rent a property is now 76 days.
For offices, HWBC estimate that lettings are running at one fifth of their rate last year; while Lisney calculates that one fifth of Dublin offices are now empty (something they describe as “startling”) and one third in west Dublin.
The usual post-bubble correction in property prices is likely to be aggravated in Ireland’s case by large falls in national income, and the dislocation in the banking system and Government finances, caused by the collapse of our unusually large construction boom.
The effective ending of new construction activity, collapsing consumption, rising taxes and cuts in Government spending all make the 15 per cent contraction in GNP forecast by the ESRI and others look optimistic. The fall in national competitiveness and likely continuing difficulties in the banking sector make the prospect of a swift national recovery seem problematic.
What we have seen then is that as the abnormal lending that fuelled the property boom returns to its normal level, Irish property prices should fall back to their pre-bubble values, at around one third of their peak values.
In the absence of evidence to support it, the Government’s claim that €90 billion in developer loans are backed by €120 billion in assets appears implausible. While five-year developer loans were the norm, properties were usually flipped on after two years, meaning that existing loans were mostly taken out at peak prices.
In addition, while loans were supposedly 70 per cent of property value, the collateral supplied was usually equity in other property or personal guarantees, both now worthless.
It appears, therefore, that, by paying an average of two thirds of the face value for Nama assets, the Government is likely to impose severe losses on taxpayers of the order of €30 billion, or one fifth of national income.
Morgan Kelly is professor of economics at University College Dublin. During the recent High Court case involving the Zoe group of companies and ACCBank, he gave property valuation estimate evidence on behalf of the bank