These charts make me sigh a bit in relief. I'm a bit less bearish today, at least on the magnitude of the housing bubble. (Update: And when I say a bit less bearish, you must first consider that I was extremely bearish and still am.)
Okay, here's the premise. Let's say you are somewhat rational.
1. You decide what you want your initial housing payment to be. I have chosen 50% of the average weekly wages reported by the BLS x 52 / 12. I figure with two people working these days in most households (generally), that seems rather doable.
2. You decide to take a 30 year conventional mortgage.
3. You decide what you can afford based on what your payments are going to be on the loan plus other direct expenses such as property taxes and insurance (in this case I have used 1.5%).
That's the chart over time. As your wages went up, you could afford to pay more for the house. As interest rates fell, you could also afford to pay more for the house (this is a LOT of it). The bottom line is that you weren't paying more as a percent of your income, at least in this hypothetical (expected) chart.
This is what it looks like when you adjust for inflation. I'm not arguing that this is the actual inflation adjusted price of homes. I'm simply saying what I would expect to see as I look back in time knowing only the interest rate on the 30 year conventional mortgage over time.
Is it rational to expect home prices to double in real terms over the last 20+ years? Yes and no I suppose, depending on how you look at it. If taken from the perspective we all seem to embrace that it is all about "what do you want your payments to be" then I would argue yes, it is rational. If you take it from the perspective of "how much do you want to pay overall" or "how much do you want to pay relative to how much it costs to make" then I would argue that no, it isn't rational.
My parents bought their home in Spokane, WA in 1982. If anything, my parents home hasn't kept up with the first chart, although it is relatively close.
My Renton, WA house was built in 1991 (I didn't own it then) and has appreciated pretty much right in line with the first chart.
So that's the good news! Rising home prices aren't just an inflation story, in my opinion. It is much more an interest rate story. You know there will be bad news too though.
Heaven help us all if interest rates rise. We will find out just how irrational planning our life around "what do you want your payments to be" was. The tailwinds from 20+ years of falling interest rates would turn into some rather nasty headwinds. What could cause that? Oh, I don't know, a reluctance to loan people money with a 30 year contract when oil is pushing $90 a barrel and many are finding it hard to pay? Just a thought.
In any event, this does not imply that there hasn't been a great deal of speculation involved in select pockets (or perhaps even entire states) of this country. It also doesn't imply houses are fairly valued. To know that we'd need to know more about future wage growth and future interest rates. I'm not that hopeful on the former and I'm somewhat fearful of the latter (in the long-term).
20+ years of tailwinds are not something we can continue to count on more than likely. Maybe we better stop acting like we can.
Even lower interest rates aren't going to help all that much. The reason? As interest rates fall, property taxes (about 1%), principal repayment, and the premium/profits to the ones making the loans (about 2%) are taking up a larger and larger part of the overall housing payment. There just isn't all that much left to be milked by lowering interest rates further, at least in the grand scheme of things anyway. At 2% conventional mortgage rates (probably the bare minimum if Japan is any indicator), the inflation adjusted chart would peak at $260k and would put enormous risk on home prices should interest rates rise in the future.
And lastly, it wasn't a trivial exercise to reverse engineer the home price from the initial payment once property taxes and insurance appeared. Perhaps I'm missing an easier way, but I setup a separate spreadsheet with the rows being 0.1% increment interest rate changes and the home prices incrementing by $2k in the columns. The cells in the table represented the combined payment on the loan using Excel's PMT function and then added 1.5% of the column header (the price of the home) to it. Then, to find the expected price of a home I used the Match and Index functions to search the table based on the current month's mortgage interest rate and payment allowed (based on the weekly wages of that month). In any event, that's how I did it case you are curious. I should also point out that because of the complexity of the data there is a nontrivial risk that I might have made an error. As is always the case, please don't just blindly trust my charts. I'm most certainly not a machine. ;)
Update:
This chart shows where the wages allocated to the house purchase would initially go (as seen in the hypothetical charts at the top of this post). I'm also more comfortable that I did the math right in the other two charts because the numbers added up (+/- $20 in rounding error each month due to the $2k rounding I did in expected home prices) when I reengineered the reverse engineering.
Source Data:
FRB: Selected Interest Rates (Conventional mortgages)St. Louis Fed: Average Hourly Earnings: Total Private IndustriesSt. Louis Fed: Average Weekly Hours: Total Private IndustriesSt. Louis Fed: Consumer Price Index For All Urban Consumers: All Items
Posts
