I'm basically going off of the simplest equation I can think of to judge how overpriced housing is: median house price vs median wages (or median household income).
What I get is this:
https://fred.stlouisfed.org/graph/fredgraph.png?g=qMrO
And this:
https://fred.stlouisfed.org/graph/fredgraph.png?g=n2Gt
The two are fairly similar, though HI is showing more long-term price inflation.
For wages: You'll note the mid-00s bubble is quite clear here, as is the glut predating S&L in the 80s. Those match up nicely, along with the relative non-bubble in the 90s.
For household income: this looks extremely frothy and above the 05/06 peak in frothness. Arguably household income is a far better measure here.
I'm having a hard time reading this as not-a-bubble, even when I've not really heard it called a bubble much in the press. Looking at trendlines, there's a definite continuity of "normal" prices that extends from post-S&L through the 90s, then into post-2008, and (in wages) a continuity of "bubble" prices that extends from S&L into ~06 into more recent times, or just (in household income) a continuation of the 05 bubble that looks to have not fully popped with extreme monetary measures taken in the aftermath. It does look to have leveled off a bit lately but still be above the longer run "normal" average by far.
Of course there are other factors, particularly interest rates being so low allowing more demand to enter the housing market through marginal consumers, along with more purchases for things like AirBnBs, both of which should drive up prices above prior norms. Also demographics, with millennials finally entering the housing market. Also construction/code standards, which have been steadily rising over this period, which should increase both building costs and perceived long-term value. I suppose I'm ignoring the impact of high-value home building in recent years in driving up prices across the board, alongside foreign RE investment inflows.
Ultimately, this all begs the question as to what the long-run "normal" ratio should be. The farthest back the data series go shows a level around 260:1 for wages and 3.6:1 for incomes, but at a time with a funds rate around 10-14%; we've basically been around 0% for most of the past 12 years, only seeing it rise alongside the dip in ratio near the end.
My opinion is that while there may be fundamentals to support a gradual long-term rise in this ratio (what the rational ceiling will be is anyone's guess), much of the most recent cycle activity has been based on historically abnormal interest rates and marginal/speculative buying/building. While rates are unlikely to change for the foreseeable future, I would expect to see a sharp contraction in the market alongside price drops as soon as we had entered any "normal but deep" recession, to say nothing of a pandemic event. Arguably that slide had already begun with rising rates after Q4 2017, and I expect it would have eventually steepened until it burned off much of the "excess" house price over that long-run "normal" trend.
Of course, now we have a lot of other factors to consider due to the lockdowns, which should only further steepen this drop, likely under that long-run "normal" trend. Of course, with a collapse in the median wage/HI, even if the ratio simply "returns to the norm" the actual house price point for that is going to be considerably lower than otherwise due to the drop in wages.
Basically, RE is probably not a great bet right now. Prove me wrong.
What I get is this:
https://fred.stlouisfed.org/graph/fredgraph.png?g=qMrO
And this:
https://fred.stlouisfed.org/graph/fredgraph.png?g=n2Gt
The two are fairly similar, though HI is showing more long-term price inflation.
For wages: You'll note the mid-00s bubble is quite clear here, as is the glut predating S&L in the 80s. Those match up nicely, along with the relative non-bubble in the 90s.
For household income: this looks extremely frothy and above the 05/06 peak in frothness. Arguably household income is a far better measure here.
I'm having a hard time reading this as not-a-bubble, even when I've not really heard it called a bubble much in the press. Looking at trendlines, there's a definite continuity of "normal" prices that extends from post-S&L through the 90s, then into post-2008, and (in wages) a continuity of "bubble" prices that extends from S&L into ~06 into more recent times, or just (in household income) a continuation of the 05 bubble that looks to have not fully popped with extreme monetary measures taken in the aftermath. It does look to have leveled off a bit lately but still be above the longer run "normal" average by far.
Of course there are other factors, particularly interest rates being so low allowing more demand to enter the housing market through marginal consumers, along with more purchases for things like AirBnBs, both of which should drive up prices above prior norms. Also demographics, with millennials finally entering the housing market. Also construction/code standards, which have been steadily rising over this period, which should increase both building costs and perceived long-term value. I suppose I'm ignoring the impact of high-value home building in recent years in driving up prices across the board, alongside foreign RE investment inflows.
Ultimately, this all begs the question as to what the long-run "normal" ratio should be. The farthest back the data series go shows a level around 260:1 for wages and 3.6:1 for incomes, but at a time with a funds rate around 10-14%; we've basically been around 0% for most of the past 12 years, only seeing it rise alongside the dip in ratio near the end.
My opinion is that while there may be fundamentals to support a gradual long-term rise in this ratio (what the rational ceiling will be is anyone's guess), much of the most recent cycle activity has been based on historically abnormal interest rates and marginal/speculative buying/building. While rates are unlikely to change for the foreseeable future, I would expect to see a sharp contraction in the market alongside price drops as soon as we had entered any "normal but deep" recession, to say nothing of a pandemic event. Arguably that slide had already begun with rising rates after Q4 2017, and I expect it would have eventually steepened until it burned off much of the "excess" house price over that long-run "normal" trend.
Of course, now we have a lot of other factors to consider due to the lockdowns, which should only further steepen this drop, likely under that long-run "normal" trend. Of course, with a collapse in the median wage/HI, even if the ratio simply "returns to the norm" the actual house price point for that is going to be considerably lower than otherwise due to the drop in wages.
Basically, RE is probably not a great bet right now. Prove me wrong.