Economics 16/02/2010: Daft.ie and rental markets

So Daft.ie numbers for rents for January are out and there is a bit of a hoopla going on in the blogosphere and the media about how things are improving. Well, they might be. 1% rise on December sounds like good news. The first rise in 24 months sounds like fantastic news. Falling net surplus of properties for rent on the market sounds like it is time to rush out to H&MacD office near you and buy-buy-buy those apartments in Dublin 78 for 250K before they are all rented out to… Who, I would wonder?

In my humble opinion, the hype is being overdone. Here is why:

1. There is seasonality issue – explained below – which suggests that January rise might be just a dead cat bounce;
2. There is demand issue – also elaborated upon below – which suggests that there is no fundamentals-based explanation for January rise; and
3. There is a momentum issue – again, more below – which implies that after 24 months of straight downward trajectory, a small correction is long over due and that this will not necessarily establish an upward trend.

So let us take a look at the 3 possible factors listed above.

Seasonality. A chart might help, or two.
The first chart shows Daft rental index, marking in red circles all the cases where January posted an improvement on December (table below brings this out in numbers in terms of m-o-m changes in the index). Only one occasion – end of 2005 – was the case where this local peaking took place one month before the normal January peak. So this January is no exception here.

Some have argued that this January is different because it is the first reversal of the established trend. True, if we take the trend to mean 2007 peak to today. But if you look at the chart above, you notice that January shows exactly the same performance relative to preceding months and following months on the upward trend and on the downward trend. So I do not buy this argument that because we were falling before, a bounce today means a change in trend.

Now take a look at daft own chart (I have no data for transactions from them, so can’t really do any analysis).
Notice the V-shaped segments? Aha, they too take place on end of 2007 to the beginning of 2008, end of 2008 to the beginning of 2009, and end of 2009 to, you’ve guessed it the ‘Great Improvement Month’ of January 2010. But here is more worrying thing: take a look at within-year trend lines for 2007, 2008 and 2009:

• Numbers of properties inflowing and exiting (rented, withdrawn, sold, demolished and soon also Namacised) trend up in 2007 and 2008 almost at the same trend line and intercept.
• Number of properties inflowing and exiting trend still up in 2009, but with higher intercept than before and flattening slope.
• Number of properties listed overall is down, true, but this simply means we have soaked up some of the overbuild into rentals. How much of it? 2007-2010 differential is about 15,000 units. Surely this is about 1/6th of the supply out there in terms of new-built, plus another 20,000 units vacated by the leaving immigrants and emigrants. Good luck if anyone thinks that we are bottoming out in terms of supply. We are just pausing.
  

So what does this tell us about 2010? Little, but… if this continues, numbers of transactions will flatten more, with greater overall average volume (higher intercept and positive slope) in 2010 than in 2009. Does this mean we are out of the woods and that supply is finally catching up (downward) with demand? I don’t know this. Why? Because I do not know anything about the drivers of supply and I know something about the drivers of demand.

On supply side, 2009 saw no new built properties hitting the market.

And it saw some reductions in supply as banks took possessions of some properties that might have been on the market for renting, but never rented. Absent actual contracts for rent, banks have no incentive to go into the expensive rental market themselves. They would rather rent wholesale to the local authorities and may be sign up with rental agents. Rental agents will list in bulk, so one listing on daft might mean a large number of actual apartments behind it. Statistics show improved (reduced) supply, but reality shows increased supply.

Other contractions took place in estates that are now completely frozen. In anticipation of continued work, half-finished estates might have seen developers listing some properties there for rent. Now that estates are abandoned – in court proceedings or simply frozen by cash-strapped developers – the listings ‘exited’ (green line went up in the chart above). Happy times? I doubt it.

But what is even more concerning in my view is the demand side. We know that there is no growth in demand out there – demographics is slow moving, so expectations based on kids finishing college and renting their first apartment are static. Foreigners are not flooding into Ireland and net emigration is now a reality. So what is happening on demand side to keep things from going bust? People move, given falling rents, to better accommodations. This leads to hollowing out of the cheaper apartments and rise in demand for more expensive (still deflating, though) better quality properties.

Daft really should do some analysis here to see if this is true. But it looks plausible. If this is happening, then we can expect to see: number of exits improving, while number of listings growing slower (lags in re-listing cheaper properties, etc). This is why the green line above is trending up faster than the blue line.

But the implication of this being true – if it is true, that is – is that within a month or so, once contracts are shifted to new and better quality properties, the cheaper, smaller apartments market will implode. And it will also drag down the more expensive market with a lag of, say 3-6 months.

In short, I simply do not buy the idea that the rental markets are signaling improvement. It will take 3-4 months of continued up-trending for me to buy the story.