Real house prices will take 90 years to reach the peak absent another housing bubble
One of the more esoteric debates on the fundamental value of houses centers on whether or not houses appreciate faster than the overall rate of inflation. Nobody who isn’t kool aid intoxicated believes house prices rise much more than the level of inflation because intelligent people understand trees cannot grow to the sky. If house prices consistently went up in value faster than price or wage inflation, over time, people would lose their ability to afford to buy houses because debt-to-income ratios would have to rise to accommodate the higher prices. No, this debate is whether or not prices rise slightly above inflation or not at all. The reason this debate is important is because if you project forward off the last bottom in 1997 to today, the slope of that line determines where fundamental value falls today. Even a small change in the underlying rate makes a large difference when projected 15 years into the future.
Dr. Shiller’s graph of real house prices suggests that prices have not risen when adjusted for inflation over the last 120 years. There have certainly been periods when this was true. When examining the last 30 years of data, it appears house prices have risen faster than inflation. In my opinion, there are two plausible answers for why house prices have risen faster than inflation over this period. First, there has been much more new construction over the last fifty years than in the 70 years which preceded it. New construction always sells for a higher value, so aggregate values go up. In theory, Case-Shiller corrects for this phenomenon by looking at the prices of individual homes. This is one reason Case-Shiller is more accurate than tracking the median to evaluate the relative price change of specific houses. Second, and more importantly, interest rates have declined dramatically and steadily over the last 25 years. Some amount of the gain is due to lower borrowing costs. House prices are bound to inflate faster than incomes rise when borrowing costs decline over a long period of time.
What I find most fascinating about this debate is the implication for future house prices. Eventually, peak buyers will see prices get back to their entry points. This nominal change in price will take anywhere from 10 to 25 years depending on the market — unless we inflate another housing bubble. However, since the value of the currency continually declines, it will be a very, very long time before these buyers see an inflation adjusted peak. Assuming there is a small increase in fundamental values over time relative to inflation, then it is possible to project when peak buyers will be made whole by the market. Based on projections from Calculated Risk extrapolated into the future, it will take until 2100 — a full 90 years — for peak buyers to break even. I hope they live long enough to see it.
by CalculatedRisk on 4/28/2012 04:14:00 PM
A year ago, Dave Altig asked Just how out of line are house prices?. Dr. Altig’s post featured both a price-to-rent graph and a real house price graph originally from the NY Times based on Professor Robert Shiller’s work.
The price-to-rent ratio graph Dr Altig presented seemed to show that house prices were getting back to normal, but the graph based on Professor Shiller’s work seemed to suggest that house prices could fall much further. Below is an updated graph from Shiller through Q4 2011.
The Shiller graph has suggested to many observers that house prices track inflation (i.e. that house prices adjusted for inflation are stable – except for bubbles).
Last year I pointed out the slope depends on the data series used, and that if Professor Shiller had used either Corelogic or the Freddie Mac house prices series, before Case-Shiller was available, there would a greater upward slope to his graph.
An upward slope to real prices makes sense to me as I’ve argued before: “In many areas – if the population is increasing – house prices increase slightly faster than inflation over time, so there is an upward slope for real prices.”
Increasing population has nothing to do with house prices increasing faster than inflation over time. If it did, Mexico City would have the most valuable real estate in the world. Obviously, it does not. What really drives up house prices is increasing wages. If businesses which employ high wage earners are drawn to an area, then house prices there will rise faster than the base rate of inflation. This is the main reason Orange County real estate has gone up so high so fast, and it is also why Orange County real estate will sputter along with the California economy.
This is the updated graph from Professor Shiller.
For the underlying data for the NY Times graphic, please see Professor Shiller’s Irrational Exuberance website.
It is important to realize that Professor Shiller used the quarterly Case-Shiller National index starting in 1987. From 1975 through 1986 he used what is now called the FHFA index. He used other price indexes in earlier periods.
The second graph shows the National Case-Shiller real prices and the CoreLogic HPI real prices (adjusted for CPI just like Shiller). For Q1, I used the February Corelogic index value.
The FHFA index used by Shiller was based on a small percentage of transactions back in the ’70s. If we look at the CoreLogic index instead, there is a clear upward slope to real house prices.
If Professor Shiller had used the Freddie Mac quarterly index back to 1970 (instead of the PHCPI), there would be more of an upward slope to his graph too. So it is important to understand that for earlier periods the data is probably less accurate.
The third graph shows the upward slope for both real price indexes. Even the Shiller “Irrational Exuberance” real price index has an upward slope (about 0.5% per year) – and the CoreLogic upward slope is steeper (about 1.5% per year).
Right now the real CoreLogic HPI is only slightly above the trend line (it could overshoot), and the Case-Shiller national index will probably be just above the trend line when the Q1 data is released.
This would suggest nominal prices are at the bottom (and real prices are close too). This is one reason I think the Case-Shiller and Corelogic house prices indexes probably stopped falling, NSA, in March 2012 (the March data will be released next month).
Look carefully at the last chart above. Let’s take the best case scenario from CoreLogic and assume real house prices rise 1.5% per year. The chart shows a reading of 78 in 1983 and a reading of 95 in 2012. That’s a 20% increase over that 30 year period. I believe most of that gain is due to declining interest rates, but assume it’s not. Assume Calculated Risk is right and it’s due to increasing population. Based on that rate of increase, it will take 90 years for house prices to reach the 2006 peak in inflation-adjusted dollars.
Peak buyers may solace themselves with nominal gains over the next decade or more, but in real terms, they are losers. Timing the housing market is important.
In case you missed the post back in January of 2008:
Today, we will look at two families, the Peakers and the Troughers (gotta love those names, right?) Both families have a combined family income of $150,000 per year, and they have both saved $100,000 they can put toward a downpayment on a house.
It is the Summer of 2006, and each family is looking at a $1,000,000 home. The Peakers think the property is a good deal, so they put their $100,000 down and borrow $900,000 with an adjustable-rate mortgage starting at 6% with a 10-year fixed period followed by a 20 year fully amortized payment at a new interest rate. Their monthly payments are $4,500 a month, but after all of the adjustments for taxes and fees and the other costs of ownership, their total monthly cost of ownership is $6,000 per month.
The Troughers, on the other hand, made the same calculation and decided that the cost was simply too high. They decided to rent and wait for prices to drop to rental parity. As it turns out, this $1,000,000 home can be rented for $3,000 per month (the cost of ownership was double the cost of rental in the summer of 2006.) In order to make this comparison apples-to-apples, the Troughers have decided that will live the same lifestyle as the Peakers, so they will put $3,000 per month into their downpayment fund while prices are dropping.
Fast forward to 2011: five years later, rents have been increasing at about 4% per year, so the Troughers are now paying $3,500 per month in rent, houses similar to the Peakers are now selling at rental parity which is about $560,000. During this five-year period, the Peakers and the Troughers have enjoyed exactly the same lifestyle: both have had use of a house with similar characteristics, and both have been living on the same amount of disposable income. The Peakers are now $340,000 underwater 5 years into their 10-year fixed term, and they are stressed about what will happen. The Troughers have a mountain of cash, and they are about to buy a home.
The Troughers have accumulated $325,000 in their downpayment fund by adding the rent savings each month and having this compound at 5% interest (the calculations are too cumbersome to post.) The Troughers now buy the comparable house for $560,000 using all of their $325,000 downpayment. This only leaves a $235,000 first mortgage. These Troughers are thrifty people, and in keeping with our all-things-being-equal example, the Troughers are going to continue to put away the same $6,000 a month the house is costing the Peakers. They will put $4,543 toward their mortgage, and the remainder toward other ownership costs. By making this $4,543 monthly payment — something they were used to doing from their 5 years of renting and saving — they will pay off the mortgage completely in 5 years.
Fast forward to 2016: It is now 10 years since the Peakers have purchased, and they still owe $900,000 on the house. Let’s assume they got very lucky, and we quickly inflated another housing bubble that brought the resale value of their home up to $1,000,000 — breakeven. The Peakers are facing a dramatically escalating payment as the 900,000 is about to convert to a fully-amortizing loan over the remaining 20 year term. Their payment will now rise to $6,447. Let’s hope they are making more money to pay for it.
Here we are in 2016, both families have enjoyed the same amount of spending money each month and the same lifestyle (remember the tax benefits are already figured in to the cost of ownership.) The Peakers have a $1,000,000 house on which they owe $900,000. They will either need to make a $6,447 payment or refinance again. The Troughers also own a $1,000,000 house, but their mortgage is completely paid off. Their only cost of ownership is reduced to taxes, insurance and maintenance. Whereas the Peakers are trying to figure out how they are going to make payments, the Troughers suddenly have $4,500 a month extra in their monthly budget, and their net worth is $900,000 higher than the Peakers.
What happens if we do not inflate another bubble, and comparable houses are only worth $800,000? What if interest rates go up to 8% or higher? The Troughers couldn’t care less, they are saving money versus renting, and they have plenty of equity; however, the Peakers are in trouble, and they may lose their home. People who bought at the peak are betting on appreciation, and they are betting against higher interest rates. Not a good bet to make when interest rates are near historic lows and prices relative to fundamental valuations are at unprecedented highs.
You can spin this example any number of ways, no matter how the Troughers save or spend their money, they will come out far, far ahead of the Peakers. They could either enjoy a better lifestyle (no mortgage equity withdrawal for the Peakers,) or save for retirement, or save for their larger downpayment. In the real world, those who did not buy at the peak can balance those options to best suit their needs and wants, the Peakers do not have these options. They are imprisoned in their house. Let’s hope it is a gilded cage.