The mortgage crackdown caused the Great Recession and the 2008 financial crisis. When Ben Bernanke assured us that the subprime crash was contained in 2007, he was correct. The Fed was responsible for the decline in residential investment up to that point. The decline wasn’t necessary. But, we were primed for recovery by the end of 2007.
Figure 1 shows residential (red) and nonresidential (orange) investment as a percentage of GDP. One important adjustment to make here, which I have left as an exercise for the reader, is that one effect of the housing shortage is that brokers commissions on real estate sales are bloated. That used to be a negligible amount, but by 2005, it was up to about 1.4% of GDP. The BEA includes those commissions in its aggregate measure of residential investment.
When you pull that out of the aggregate figure, residential investment as a percentage of GDP ran within a very thin range from 1960 to 2005. It wasn’t unusually high in 2005. And, it was already at the historical low end of the range and falling sharply by the start of the Great Recession.
From that point on, the Fed’s errors were downstream of the mortgage crackdown. The crackdown created a financial crisis and a recession, and the Fed treated the wealth shock, the drop in credit-funded spending, construction unemployment, etc. as inevitable reversals of 2005 excesses instead of novel and unnecessary consequences of 2008 policy choices.
They didn’t realize that extreme policy errors in 2008 required extreme countermeasures from the Fed in 2008. At the end of the day, the Fed is the last mover, and is responsible for deviations from stable growth.
Their policy choices in 2008 would have probably been fine under normal circumstances. Suddenly cutting 15 million families out of the mortgage markets isn’t normal circumstances.
In Figure 1, I have marked the point in every business cycle when the yield curve un-inverted with dashed vertical lines. Inverted yield curves (long-term interest rates lower than short-term interest rates) usually point to coming recessions, and they un-invert when the Fed lowers the Fed Funds overnight target rate to a level lower than long-term rates.
As you can see in Figure 1, like clockwork, whenever the yield curve un-inverts (the green line moves below the black line, and the gray line moves above zero), residential investment bottoms and recovers. As Ed Leamer famously pointed out, housing investment leads us into and out of recessions.
(The only time that the recovery didn’t match up exactly with the yield curve was in 1990, but if you look at the fed funds rate (green) and the 10-year yield (black), the un-inversion was very tepid. They moved very closely together, even though the Fed Funds rate was slightly lower than the 10-year yield. And, when the Fed Funds rate then finally sharply diverged from the 10-year yield, residential investment bottomed in the first quarter of 1991.)
You can see that the yield curve was un-inverted by the first quarter of 2008. And residential investment was… (crickets chirping).
In Figure 2, I show residential investment as a percentage of GDP, the yield curve slope, and the median credit score on new mortgages.
Residential investment should have sharply recovered in early 2008. The Fed did what normally would need to be done. They should have been beside themselves when the construction market didn’t react. Of course, in 2007, when Ed Leamer had famously pointed out that housing investment leads us into and out of recessions, he also, unfortunately, told them to expect this time to be different.
So, they weren’t beside themselves even though they should have been. They actually were forecasting that the usual policy might avoid a recession but that housing wouldn’t lead the way this time. A terrible error, to be sure.
But, it was only an error because of the mortgage crackdown. (Here is a post from last year that was for subscribers only, which I have now made available to all readers, which includes some detail on the Fed’s housing derangement even well after the crisis. They interpreted the negative demand shock from the mortgage crackdown as a positive supply shock, since it left the country littered with millions of unsold foreclosures.)
So, in every single business cycle for decades, residential investment turns around and starts to rise when the yield curve un-inverts (where the dashed vertical black line is) – like clockwork.
In 2025, residential investment is still no higher than it was in the first quarter of 2008.
Borrower quality was pretty stable before 1999. Basically, the credit score (or hypothetical credit score in earlier decades) on newly originated mortgages was 720 points (+/-10) for decades and then since 2008 it has been 770 points (+/-10), and as you can see, that happened very quickly from mid-2007 to mid-2009. It happened right when the un-inverted yield curve would always be expected to lead to a housing recovery.
The mortgage crackdown caused the Great Recession and financial crisis. All else equal, without the mortgage crackdown, 2008 probably looks something like the 2000 cycle. A short downturn with minimal employment disruptions.
How about today? The yield curve has technically been inverted by this measure, but it’s a weird inversion. Normally, the yield curve flattens and then slopes up as short term rates decline faster than long-term rates do. This cycle, the inversion has been the result of the weird bump at the short end of the curve that is left over from the transitory inflation period.
To be honest, I think we still are in the transitory inflation period in terms of residential investment inputs, which is probably why the high Fed Funds rate hasn’t been recessionary. There isn’t capacity for housing starts to lead us to new growth.
When that capacity starts to return, the yield curve will be firmly positively sloped. Long-term yields will be higher than short-term yields. The forward yield is already very elevated relative to the forward yield. I expect yields to decline because of Trump’s aggressive economic meddling. But, the curve isn’t likely to invert.
There seem to be two schools of thought currently. Either a recession could cause housing markets and new home sales to decline, or high interest rates will. This would be a real recession, not a nominal one. The yield curve isn’t going to flatten as a pre-requisite. Yields may go down as a result of declining economic conditions, but the Fed won’t be flattening the curve by raising short-term rates. If conditions lead long-term rates to decline below 4%, employment and inflation indicators will be signalling the .
As capacity expands, residential investment will rise. Again, looking at Figure 1, there is no historical precedent for a deep cyclical decline in residential investment when the yield curve is sloping up. To the extent that demand softens, the main effect will be to lower input costs. There is a lot of room for those costs to decline. In most reasonable scenarios, homes will be constructed, near the capacity to build them, and they will be sold at reasonable margins for the foreseeable future.