Tim Duy takes a look at housing and finds himself agreeing with David Altig:
I have tended to shy away on the housing story, leaving it to other bloggers such as Calculated Risk. I am sympathetic to those who view it as a bubble, but have limited my comments to likely Fed reaction to such a bubble.
Today, however, I was unsettled by the combination of weak durable goods numbers, strong housing numbers, and this morning’s Wall Street Journal piece regarding global capital flows into the US housing market. Like many, I see the need for an eventual rebalancing – a shifting away from consumption and housing and toward investment – of growth in the years ahead. Consequently, I would be displeased to see that the Fed’s contractionary campaign failed to yield such a result. But if the housing sector stays strong while other sectors stagnate, the day of reckoning is pushed further into the future.
With that in mind, I wanted to get a better sense of the Fed’s impact on the housing market. To be sure, this is an exercise in “optical econometrics,” but I think the data tells an interesting story just the same. [Note: I used the Fed St. Louis Fed site as my data source.] First, a look at 30 year mortgage rates and housing starts:
Quite honestly, this picture doesn’t tell me much. Sure, the rate spike in the early 1980s corresponds to a housing downturn, and low rates today correspond to strong housing starts, but why exactly is housing in a downtrend during the late 1980’s even as rates are falling? I felt I was missing some appropriate measure of monetary policy. So I took another pass at the data, using the spread between mortgage rates and the fed funds rate:
I lagged the spread 12 months. The most visible feature in this chart is the strong correlation between the lagged spread and housing starts during the 1970s and early 80’s. That relationship, however, begins to break down in the late 1980’s. Of course, I am not surprised that the data would exhibit a structural break. After all, deregulation and financial innovations have radically altered the mortgage markets. Lower down payments, expanded uses of ARMs, declining lending standards, etc. have all played a part in altering the link between interest rates and housing starts. Has the international sector played a role? For that, I went back to a relationship I stumbled upon a year ago or so: that between the current account (as a percentage of GDP) and housing starts.
Interesting, no? The relationship between the current account and housing starts during the 1970’s and early 1980’s is likely spurious. Recessions are consistent with both weak housing and improving CA deficits. But the relationship tightens up dramatically in the second half of the 1980’s, and note that this past recession saw neither significant improvement in the CA deficit nor deterioration in housing starts. So, I don’t think this is entirely a spurious relationship, especially for the last 20 years. Something structural is likely happening.
All in all, it suggests to me that international factors – specifically, the willingness of foreign investors to place their capital into the US – have a significant place in explaining the consumption binge/CA deficit/low interest rate issue. This places me in the camp of David Altig, who had a similar take on today’s WSJ article.
Of course, the international angle only increases the difficulty of the Fed’s job – the willingness of foreign capital to flow into the US could mean the Fed will end up strangling the non-housing sectors of the economy to keep overall inflation expectations in line.