Elephant in the room: NAMA’s yield problem hasn’t gone away

Six months on from the National Asset Management Agency Bill, outlining how NAMA will work, we are getting to the business end of things, with the first few loans expect to transfer over shortly. Many commentators are still doing their best to point out shortcomings with the NAMA concept and viable alternatives to it – see for example David McWilliams’ “Anglo is our Stalingrad” and Brian Lucey’s “Micawbernomics“. At this point, though, I have to confess I’m resigned to NAMA going ahead and, if that’s the case, the focus has to be damage limitation and to the elephant in the room that is the yield on Irish property.

In a way, this isn’t as hopeless as it sounds. NAMA is based on the concept of long-term economic value, i.e. relating the value of properties, in a time when markets are thin/broken, to what they could realistically fetch in rents, when out the other side of the Great Recession.

However, as I pointed out in the Irish Times in September, this relies on those running NAMA getting its core assumptions right. In particular:

• It has to get the fall from peak to time-of-transfer right.
• It has to get the yield right.
• It has to get how yields might correct right.

What’s the loan book worth now?

On the first, no harm reminding ourselves that NAMA is scheduled to take over the loans of properties that were, when the loans were issued, worth €88bn. The loans themselves were worth €68bn, although there’s about another €10bn in accruing interest, by this stage. As I pointed out six months ago, while there is some international diversification, the 6% of the loan book that’s not in Ireland or the UK just about offsets the 6% that’s up in Northern Ireland, let alone the 66% in the Republic.

In terms of getting the fall right, the big problem with NAMA is that, for all the detail we have – €110m in Czech developments compared to €90m in Italian associated loans – we just don’t know what’s in about €64bn of the original €88bn. All we know is that these are broken down into €32bn of land and €32bn of “associated loans”. It seems likely that about €21bn  is in Irish land. This is the same land that, for what it’s worth, Judge Peter Kelly of the Commerical Court estimates has fallen in value by about 75%. Indeed, if some land is down 95% or more – like the land in Athlone over which he was judging – while other land is down by (only!) 60%, the original €21bn of land is probably now worth less than €5bn.

The focus then turns to the €16bn or so in Irish commercial and residential developments. It’d be a brave person who’d put €10,000 (or, if you like, €10,000 for every citizen in the State) on the average peak-to-trough fall in prices in both segments being less than 50%. (Asking prices – more than likely above closing prices – were 43% down from peak levels in Dublin city centre in late 2009.) So, looking purely at the Republic of Ireland, it’s hard to see how how falls of 50%-75% could average at 47%.

The vast bulk of the remaining €21bn of foreign land and developments is in the UK, either in Britain, where property prices are falling again after a supply drought had help them up, or Northern Ireland, where prices are over one third below the peak.

In short, even a 20% average fall in NAMA’s non-Irish loan book means that a 47% is already looking optimistic. The figures outlined above would suggest an average fall of 51%. (And remember, from this graph, every percentage point more on the way down makes the 10%-in-10-years argument increasingly risible. A 51% fall means NAMA needs a 25% rebound in property prices by 2020 to break even, give or take the token amount the banks would have to pay back to the state in such an occurrence.)

What’s the problem with yields?

The real problem for NAMA, however, is in relation to the second and third bullet points above, i.e. in relation to the yield on property. The following paragraph is from the supplementary documentation published with the NAMA bill, with my additions in italics:

A property yield is derived by expressing a year’s rental income as a percentage of the property cost. There is an inverse relationship between yields and property prices; as property prices rise, yields fall [assuming rents stay constant]. The prime yields broadly represent the market’s evaluation of the attractiveness of investment in property. It is clear from the table that follows the majority of European cities current prime yields [on commercial property] are higher than their long term 20 year average, particularly Dublin where the Prime Yield [on commercial property] is at 7.25% in comparison to a 5.56% long term average. As yields move towards their long term average this would indicate an increase in [commercial] property prices [or a decrease in commercial rents].

The good news is that by looking at yields, NAMA is looking in the right direction for long-term economic value. The bad news…

1. This is only the yield for commercial property in Dublin. This is a subset (commercial) of a subset (development) of a subset (Dublin) of a subset (Ireland) of NAMA’s loan book – perhaps 5% of NAMA’s loan book. Surely we as the shareholders in NAMA deserve a better prospectus than this? A somewhat more representative set of yields, on residential property around the country, is presented in the graph below.
2. By portraying prices as the only mechanism through which a yield can correct, the paragraph above shows either a stunning lack of understanding about the fundamentals of the property market or a conscious or unconscious willingness to direct attention to prices not rents. Yields can correct through either property prices changing or through rents changing. A  glut of supply on to the market would mean rents would have to fall to balance the market.

This is exactly what we’ve seen in the residential rental market in the last two years as a result of boom-time over-construction. What is the prospect for commercial rents in Ireland? A February 2010 report by Savills on the Dublin industrial market contained the following eyebrow-raising statistics: During 2009, uptake of industrial space in Dublin fell by 33% to 110,000 sq/m. Meanwhile, about 1 million sq/m (nine years supply based on 2009 take-up) is sitting available on the market, up from 700,000 sq/m a year ago. As the report authors themselves say:

This has and will continue to put downward pressure on rents and force landlords to deliver more flexible terms than in the past.

The overwhelming evidence from Ireland’s commercial and residental markets is that rents are responding to boom-time over-construction by shifting downwards. (No bad thing, either, as Ireland seeks to get its property costs out of the global top 20.) This means that any positive “yield gap” could easily be corrected by changing rents, not increasing property prices. Not only that, the graph below shows that, for residential property at least, the gap between current yields and what NAMA assumes as a long-run yield actually goes the other way, and by some amount.

If NAMA could show that it’s aware of these very basic facts about Ireland’s property market before it starts taking over any loans, I would be a lot less worried about giving it access to tens of billions of euro worth of current and future taxpayers’ earnings. If it doesn’t, the elephant is very much still in the room. If it does, and taxpayers pay a fair price, cash-starved banks will mean a new elephant will take its place – nationalisation.