Steven Keen – “I will be wrong on house prices”

How do I know? Because I looked it up on Wikipedia, which counts as research on such topics – though not on, say, the link between bushfires and climate change. Why did I bother? Because my call on house prices was shanghaied by the property lobby, and now that I am near certain I’ll be wrong even as I defined it, I want to un-Shanghai myself first.

For the record, what I said (from memory, in answer to a question from Kerry O’Brien on The 7.30 Report) was that after its Bubble Economy collapsed in 1990, Japanese house prices had fallen 40 per cent over 10-15 years, and I saw no reason why Australia would avoid the same fate. This turned into the famous bet when Rory Robertson sprang it on me at Parliament House on November 26th, 2008 (one month after the Rudd government had doubled and trebled the First Home Vendors Grant as part of its package to fight the global financial crisis — a move I described at the time as “Rescuing the Bubble”). The actual exchange was as follows (click here to hear it for yourself):

RR: I think you said a 40 per cent fall on average across Australia, is that correct?

SK: Yeah, but over a 10-15 year period, mate… it’s a long-term bet, but I’m willing to stick to it

RR: Well how about this? If Australian house prices, as measured by the Statistician, fall from peak to trough in nominal terms by 40 per cent, I will walk from Canberra to the top of Kosciuszko, and if in fact Australian house prices fall by less than 20 per cent, so it turns out you’re less than half right, you walk…”

We never negotiated the fine print of this bet — a failing I have since regretted. Three obvious points that needed clarification were:

– When was the peak?

– Was the fall in real or nominal terms?

– How do you define the end of the bet?

As I found out to my chagrin, Rory and the property lobby defined these three points as:

– The peak was March 2008, the date before the bet that house prices had peaked, which wasbefore Rudd’s First Home Vendors Boost was introduced;

– Nominal prices, not inflation-adjusted; and

– The bet was over as soon as house prices exceeded the March 2008 peak.

My definitions were:

– The peak might have been March 2008 had Rudd not restarted the housing bubble via the First Home Vendors Boost, but the peak would probably come sometime after the boost expired. When that happened, I did think that it would be the peak for real house prices;

– Either one of either nominal or inflation-adjusted prices should qualify. The global economy was entering a debt-induced Depression, and the previous one had led to deflation as well. If the bet had been in real terms and deflation had also occurred, then nominal house prices could fall 40 per cent and yet real house prices could actually rise (if the fall in the CPI was greater than 40 per cent). Since this was just a bet about house prices — rather than a combined bet about house prices and deflation — a fall of that magnitude in either should real or nominal prices should qualify; and

– The end couldn’t be any temporary new peak in prices — even in a crash, house prices do sometimes wobble, especially in real terms if deflation occurs (as is evident in both the Japanese and US data shown in Figure 1). The best way to define it was by the duration of the bet — which was 10-15 years — but that’s too long, so maybe a compromise was necessary. Rob Burgess’s suggestion of a 20 per cent fall over five years made sense (though even that was in Rory’s favour, since it was in nominal terms only, and used the minimum time horizon I specified rather than the average).

Figure 1: House prices in Japan, US and Australia since their peaks

But I thought I could negotiate terms? Ha! There’s no contract when you’ve been shanghaied. As the house price frenzy gathered pace in 2009 thanks to Rudd’s Boost, I realised that I’d done the proverbial innocent-stroll-along-the-US-West-Coast by agreeing to this bet, without having lawyers negotiate the terms beforehand. Any attempt to clarify what I meant by the bet would be pilloried by the property lobby in the media — I’d be portrayed as trying to weasel out of my commitment. So when it became obvious that Rudd’s Boost was going to drive house prices to a new high, I decided to do The Walk to Kosciuszko — even though I didn’t believe I had lost the bet — simply to shut the property lobby up.

Real house prices peaked in June 2010 (see Figure 2), and I did think then that prices would fall from that peak by enough for me to win the bet as I had defined it.

Figure 2: The bet, its nuances, and The Walk to Kosciuszko

They did duly fall by 10 per cent over the next two years — which was well on the way to meeting my expectations of a 20 per cent drop over five to seven years. But now they are rising again in a renewed bubble, and I expect they’ll soon exceed the previous peak.

I also expect that this new bubble will ensure that, when the hot air finally leaves the Australian Housing Balloon, its deflation will be too slow to result in a 40 per cent fall below the June 2010 level by June 2025—which is the Statute of Limitations on my original call. So I expect to lose the bet, on my own terms.

I’ll discuss why I think I will be proven wrong in the next instalment.

PS I noted on my Debtwatch blog last week that my mortgage accelerator indicator was still rising, indicating sustained price rises were afoot. Here’s the mortgage acceleration and house price change graph from that post, updated for the most recent ABS data, which was released last Monday.

Figure 3: Mortgage acceleration and house price change

Steve Keen is author of Debunking Economics and the blog Debtwatch and developer of the Minsky software program. 

Graham