Sunday 18 May 2008

Comparing bubbles, predicting the crash

During the last 20 years, the UK had the misfortune of living through two huge housing bubbles. The first was carefully nurtured under a Conservative government, while second happened under New Labour. Thus, it would seem that housing bubbles are non-party political beasts.

We see the current house bubble through the prism of the last one. Therefore, it might be useful to compare these two extraordinary periods. Before comparing them, it is important to date them. I will use a simple rule. Over the long term, there appears to be a steady relationship between house prices and average incomes. Historically, this ratio fluctuates between 2 and 3. When the ratio rises above 3.5, I assume that house prices have departed from long-run fundamentals and the bubble has begun.

In the case of the Conservative bubble, this happened in April 1984, and in October 2001 for the Labour bubble. Once I have the starting date, I set the price on that date to equal 100 and then superimpose one price series on top of the other. If I do this, I get the chart below.

Superficially, both bubbles seem similar. In terms of housing inflation, prices increased by 118 percent under the Conservatives, and 113 percent under Labour. The length of both Bubbles is also similar; the Conservative one took 63 months to reach its peak, while the Labour bubble took 70 months.

However, behind the superficiality lie some profound differences.

The Conservative bubble is an easier story to tell. After several years of steady price appreciation, prices accelerated in one last mad burst to the top. Once it reached the peak, inflation was raging and the government ordered the Bank of England to raise rates. House prices then began to tumble for the following 80 months.

The New Labour bubble was more front-end loaded. During 2001-05 price appreciation was extraordinary. By mid 2005, inflation picked up a little, and the Bank of England cut rates. House prices began to slip, and the Bank of England quickly began reducing rates. The policy reversal gave the bubble a second lease of life, and prices increased for a further 2 years.

The differences become much more apparent when one looks at price to income ratios. In the case of the Conservative bubble, the ratio only reached 4 in December 1987, and it peaked at 4.8. Therefore, for long periods, one could argue that house prices were only marginally detached from incomes. At the time bankers and estate agents used this argument extensively.

In contrast, the price to income ratio in the Labour bubble shot up rapidly, passing 4 within a few months, and almost reached six at the peak.

The reason for the different price to income ratios is straightforward. Nominal interest rates during the Labour bubble were much lower, which allowed people to take on much higher mortgages. Interest rates were low, because inflation worldwide was under control.

Real house prices point to another major difference. In real terms, the Labour bubble was larger. During the 1980s, real house prices increase by 60 percent, whereas the most recent bubble, prices increased by over 80 percent.

We can also see how house prices returned to fundamentals once the Conservative bubble finally burst. There was a modest reduction in nominal prices, but most of the adjustment came through inflation. Since inflation and average earnings move together, house prices became affordable because eventually nominal wages caught up. The real adjustment, during the 1990s, was dramatic. From the peak to trough, prices fell in real terms by 34 percent. From the beginning of the bubble in 1984, to the bottom in the late 1990s, house prices increased in real terms by just 10 percent.

What does this tell us about the likely adjustment facing today’s housing market? Today’s prices will need a much higher nominal fall to return the market to a more reasonable price-to-income ratio.

To understand why, it is important to recognize that in the 1980s both inflation and interest rates were well above 10 percent as the bubble burst. The higher levels of inflation ate away at the real value of housing and put a floor under nominal prices. Furthermore, as inflation subsided, interest rates came down reasonably quickly. This also helped cushion the nominal fall because housing affordability improved. Although the market still crashed dramatically, these two effects limited the size of the crash in nominal terms.

Neither of these two effects is present in quite the same way right now. Although inflation is rising, it is likely to fluctuate between 3-5 percent for the next two years. It is not enough to seriously dent the real value of housing or reduce the price-to-earnings ratio over the medium term. Interest rates, and especially mortgage spreads, are now adjusting upwards, reflecting higher levels of risk. As inflation picks up by 1-2 percent, it is likely to put a floor under interest rates and may even contribute to higher rates, and therefore keep housing affordability high.

The 1980s adjustment had other things going for it; a recession, a high repossessions rate, and the sterling crisis. However, I doubt there is much comfort here. In the fullness of time, the UK economy may also produce its own complementary recession, which will push up repossessions and unemployment.

Three things make the current housing market much more frightening than the one in the 1980s. First, it is incredible just how unaffordable house prices are today. The UK average price-to-income ratio is well over 5. In some parts of the country, like London and Northern Ireland, it is approaching 8. Second, personal sector balance sheets are in appalling shape, with many households holding crushing levels of debt. Finally, banks have lent out too much and they know it. They have abandoned the housing market and they are unlikely to return.

It all points in one direction. The UK housing market is facing a massive nominal price correction. As hard as it is to believe, this correction could be much worse than the one in the 1990s.

20 comments:

Anonymous said...

A well researched article.

Anonymous said...

I am amazed at how high inflation was during the 1980s-1990s. Was it really so high?????

SD said...

firstly i liked the article, interesting read. but i want to bring forth a few points.

There are stark contrasts between the two periods being compared here.
1.Globalisation
2.Real Estate surge this time around is driven by global cues than local factors.
3.It is an unwarranted to say banks won't return to housing market, its their bread. They might exercise control in the way they would enter the housing market. Might avoid structured and complex financial products.

Anonymous said...

A lot of useful and diligent research here... I especially like the supposition that house price inflation is not party-political.

With this in mind, maybe we're looking at too short a time-frame? There is no doubt that house prices took a tumble in the early 1990s, but, without the complication of the ERM debacle, maybe that wasn't a crash? Might it not be more appropriate to look at the current credit-crisis as the culmination of 24 years of credit expansion from 1984 to 2008 with a brief lull 1990-1996?

Mark Wadsworth said...

Alice, excellent stuff, but why not include the 1948 or the 1973 bubbles as well?

Graph here, these bubbles have far more similarities than differences. You get a self-perpetuating asset-price bubble and credit bubble (you can't have one without the other) then it goes *pop*

Anonymous said...

Yeah, a great article. I imagine it was alot more difficult to draw the threads together than it looks in the final piece.

Things I'd add:

- The Tory bubble happened with high interest rates and the bursting was cushioned with lower rates. It's the opposite this time round, which is ominous;
- Incomes grew into house prices following the Tory bust. This time house prices will fall to income;
- The impact is that under the Tory bubble the homedebtors were bailed out by inflation and reduced interest rates. Basically their debt was devalued. This time around homedebtors are more screwed because rising interest rates will make the debt more, not less, crushing.

Nick

Anonymous said...

I think inflation might go well above 5%.

We are in the midst of a Sterling crisis right now.

How much have house prices already fallen in Euro terms?

Anonymous said...

Interesting conclusion re: inflation and interest rates.

Mark Rainey said...

You have missed out an important factor in the UK of supply and demand, with the population growing and supply reducing house prices will not crash.
House prices have trebled in the last 10 years so even if they went down 20-30% is this a crash ? No

powerman said...

Prices have scope to go down about 40% to restore the historic average ratio between house prices and earnings. This would certainly feel like a crash to a highly-leveraged buy to let investor or somebody with less than 40% equity in their home who needed to sell up and move with their career.

Population and housing supply are two factors in house prices. The third is the supply of credit people use to bid on the houses.

The price increases we saw over the last decade were mainly in response to a massive expansion of this supply of credit. It is now being withdrawn.

Alice Cook said...

Simon,

A wonderful question. Just posted something on it.

Alice

Anonymous said...

House price crash? 40 percent in two years.

Anonymous said...

Mark... Firstly a point of definition. A crash is widely understood to be a fall of 15% or more, so a fall of 20-30% would definitely be a crash.

While the population is growing (slowly) emigration counteracts this effect, as does the fact that we've been building significantly more quickly than the population has been growing.

When you talk about "supply and demand" you make an elemental error. You equate desire (or perceived need) with demand. They are not the same... not at all. Demand, in the economic sense - that would not be a nonsense in the context of your "supply and demand" argument for housing - has already evaporated. The crash in value has already happened... we're now waiting for owners to come to accept the new prices.

Anonymous said...

Mark, you need to check out the archive in this blog. There are posts that deal with housing supply and demographics.

Anonymous said...

Agreed. "effective demand" is not the same as "demand". Everybody wants to live in a big mansion by the sea, but unless you've got the cash or loan to pay for it, you won't affect the price.

The UK population is declining except for four sectors:
1) Old people, who don't work
2) Welfare scroungers, who don't work
3) Low income immigrants, who don't make much money and tend not to buy and will flatshare
4) Professional immigrants, who do have money.

The latter two categories are drying up as the economy tanks.

Nick

Mark Wadsworth said...

@ mark, if price trebles from £100 to £300 and then drops by 30% from £300 to £210, the 30% drop has eaten away nearly half the original gain!!

Anonymous said...

The nonlinearity of percentages, up is not equal to down.

DTR said...

Mark, when prices are rising fundamentals going out of the window, because people cannot see what will stop it rising, so along the way they make all kinds of things up. Supply and demand, and we are joining the euro so rates will be 2% were two of them. Demand was part speculative we all know that, but debt is the bigger problem here. Look at Japan.. UK and other countries are not much different.

When prices fall, then fundamentals kick in as that's the only method you can use to determine the floor.

I predict prices will fall back to late 2001 price levels which creates a 35% fall.

Here are some samples.

Prices will fall 35% I believe and they will fall back to at least 2002 levels.

This means even if you say the average house was 70k in 1998 which is now 160k for example, -35%, they will be worth 100k, which is well above 70k 1998 level, because over the last 10 years if we say inflation has been 5% a year then a 50% increase is ok making the price 105k...

They are my predictions

325 detached fall to 212k..
294 detached fall to 192k
220k detached needs to fall to approx 150k
160k semi needs to fall to 105K
120k terraced house to fall to 80k

Anonymous said...

Yes. Those figures look plausible to me.

£350K => £230K is my target... from peek.

It seems reasonable to expect, too...

DTR said...

I agree Steve. It will take a bit of time, but I'd say after the summer period, say August, real falls will shown on the nation figures which the the banks have been trying to hide falls on, and by DEC it will show average falls up to 10% with more to come in 2009, though we know locally falls will have been between 10-15%...