Un articulito interpretando series de datos inmobiliarios europeosde la OCDE y el FMI, extraído del boletín económico semanal de Lehman Brothers (banco de inversión americano). El enlace es de pago, así que no lo puedo poner directamente. Faltan los gráficos, que muestran el efecto de un "bust" definido por el FMI como el cuartil inferior de la serie histórica de bajadas de precios) sobre el consumo y el PIB: pasan de una media de 3% a -2%. O sea, recesión. La probabilidad de que a un "boom" le siga un "bust", según estos datos, es del 40%.
Housing and consumption risks
Most mild uplifts in house prices are followed by mild downturns. But about 40% of
“booms” end in “busts”. The implications for the aggregate economy are usually severe.
In recent months we have warned of what might be called “close-to-home” downside
risks to the euro-area economy emanating from its own housing markets and its own
consumers. The first reflects the fact that formerly “hot” housing markets – such as
Spain’s and France’s – are now showing signs of cooling.1 The second reflects the fact
that our recent research suggests that German consumers have been saving more than
usual because of an increased perception of job insecurity: we worry that these fears may
be fanned next year by the combination of continued structural reform efforts and a
Now, a third risk appears to be building – that the US housing market slows more sharply
than generally expected, hurting the US consumer and, via weakened export demand and
confidence, the European economy too. (Indeed, the minutes of the Bank of England’s
recent MPC meeting, published on Wednesday, revealed that this increased risk, for the
UK, is one that the Bank too is watching, and evaluating, closely.) In order to better
gauge the risks of housing markets going belly up, we have taken a close look at recent
OECD and IMF research into housing booms and busts, and the consequences thereof.
Mild upswings appear to be fairly benign events…
Recent OECD research has highlighted that, at a global level, real house prices appear to
be trended higher, but with quite long cycles around that trend (Figure 1).3 Moreover, its
analysis shows that most regions have experienced “booms” (Figure 2). Using the same
sorts of techniques typically applied to business cycle dating research, the OECD
researchers split each of the 18 member countries’ real house price series into up- and
down-turns, ignoring the latest (protracted) upswings, as it is not yet possible to identify
peaks for them with any great confidence.
The research finds that the average upswing lasts 23 quarters, and the average
downswing 19. In other words, they are typically long drawn-out affairs, rather than
short, sharp, tempestuous events (which appears to be the general perception). A second
point to note is that the average decline in real house prices, peak to trough, is just shy of
25% – and just shy of half the rise that occurs from trough to peak. As for what the
OECD terms “major” price moves – of magnitude greater than or equal to 15% in real
terms – it turns out that the chances of a major move up being followed by a major move
down are high, at roughly 75%. So, without knowing anything much about the specifics
of recent country experiences, it might seem reasonable to make one’s modal forecast
that of a significant slowdown in real house prices – consistent with nominal prices
broadly stabilising and inflation eroding houses’ real value for three or four years. These
“softish” landings are precisely the sorts of stories that we have been telling for the
United States, the United Kingdom and the warmer bits of the euro area – i.e., the single
most likely outcomes. In other words, our modal forecasts are for outcomes that entail
pain for the consumer but more on the headache scale, as opposed to a migraine.
…but big booms are often followed by big busts
The IMF has also carried out some event-study style research into past housing cycles,
again starting out by trying to identify peaks and troughs and thus define downturns, in
order to see what the typical cycle looks like and what a “bad” one does.4 Having
ordered all the past real house price downturns in the major economies, it then defines a
“bust” as the lower quartile of outturns. Perhaps not too surprisingly, on average these
entail greater declines in house prices than in the case of the OECD’s “major
downturns”. (After all, they are one of the tails of the distribution, as it were.) In
addition, they engender bigger hits to the macro-economy too. Figures 3 and 4 illustrate
– showing what the median bust looks like, along with the upper and lower quartiles, in
terms of consumer spending and GDP up to three years either side of the bust beginning.
The bottom line of the Fund analysis is clear. During “busts”, the hit to the consumer,
and to the economy at large, is of a different order of magnitude than the sorts of
slowdowns incorporated in the consensus, or for that matter our own, forecasts for the
US, UK and euro-area economies. Rather, they imply that outright recessions – in which
real GDP contracts – are the norm when “busts” occur. As a first stab at thinking about
the probability of such a scenario developing, it is perhaps worth asking the data what
the probability is of a “bust” occurring after a period of “boom” (as has happened in
recent years). The answer to that question turns out to be 40%. So, the tail-risk might
perhaps be greater than what might be imagined from looking at the survey of
professional forecasters (SPF). After all, the SPF for the euro area reckons on the
probability of GDP growth being under 1% at just 4.5% in 2007 and 3.3% in 2008. And
the equivalent surveys for the US and the UK do not put the numbers much higher, at
7.6% (2007) and 8% (2008 Q3) respectively.■
1 See “The importance of housing”, Global Weekly Economic Monitor, 23 June 2006.
2 See “A sense of insecurity”, Global Weekly Economic Monitor, 8 September 2006.
3 See “Recent house price developments: the role of fundamentals”, by Girouard, N., Kennedy, M., van den
Noord, P., and Andre, C., OECD Economics Department working paper no. 475, January 2006.
4 See chapter II of the IMF’s World Economic Outlook, April 2003.