Approach Example: Housing Prices - Los Angeles
Previously we discussed the primary factors we believe influence the stock market, located here. The simplest illustration of how these factors work can actually be done with housing prices.
The only two considerations will be a Long Term Reversal Factor and a simple Intermediate Trend Factor. We will be creating both for Los Angeles housing prices because the data series are longer for this area than most areas and the variation in housing prices over the past 20 years in this area makes it worthy of analysis for possible mispricing.
Many people buy a house to live in it. However, a house has value because an owner could theoretically rent it to someone. In fact, most homeowners are really renting to themselves, but typically without the paperwork that comes from dealing with a separate renter.
The way we created a Long Term Reversal Factor is by taking housing prices and dividing by rental incomes. The S&P/Case Shiller housing price indices have the best methodology of all the housing price indices because they actually look at how a particular home has done over time without the distortions caused by things such as the mix (large versus small ) of houses sold for a particular period. These indices also have the added benefit of having a futures market where investors have made estimates on where housing prices will be as close as 3 months out and as far as 4 years out.
For the rental income data, Owner’s Equivalent Rent from the Bureau of Labor Statistics shows how rents have changed over time. For our purposes here, both of these data sets represent good approximations of home prices and rental incomes.
The graph below shows Los Angeles housing prices since 1987 with the series continued using current futures market forecasts. In addition, we have included where housing prices would have been if they moved with rental prices at the same price/rent ratio as the peak in the market in the late 1980s. Similar we did the same thing for the trough of the market in the late 1990s. The middle line takes an average of the peak and trough price/rent ratio. We then extrapolated these three lines out to 2023 based on future rental price increases being the same as past rental price increases.
The Long Term Reversal Factor is created by dividing by the housing price line by the middle line from above, which we will assume is the correct housing price (The data series is quite short to be too sure of this assumption). In the following analysis we will assume when housing prices are higher than the middle line the Long Term Reversal Factor is negative. Conversely, when housing prices are below the Long Term Reversal Factor is positive.
The Intermediate Trend Factor is much simpler to create. We will assume that if the Long Term Reversal Factor is higher (lower) than it was a year ago the Intermediate Term Trend Factor is positive (negative). Housing prices are very persistent making this simple Factor very useful. The primary issue with it is that housing price data are typically reported with a lag measured in months.
The graph below shows each combination on a graph of the Long Term Reversal Factor.
The usefulness of the Factors can also be shown in the performance of housing prices, measured relative to rents. The annualized performance numbers, with returns lagged two month to account for data lags, are as follows:
Housing price changes relative to rent during each Factor Combination (annualized, lagged two months) Trend positive, Reversal positive: 8.98% Trend positive, Reversal negative: 6.55% Trend negative, Reversal negative: -9.61% Trend negative, Reversal positive: -3.99%
While not shown above, taking into account the higher rental income when the Long Term Reversal Factor is positive versus negative would make the performance spread between those conditions even greater.
The above chart shows that Los Angeles housing prices currently have both Factors negative. Based on the numbers above, the decline this has been associated with in the past to housing prices relative to rents has been greater than the increase in rents, causing nominal housing price declines.
The dynamics of housing prices are that they will probably continue to fall relative to rents until they stop falling relative to rents. Once prices do start rising faster than rents, the strength of the rebound should be related to the depth of the prior decline.
Rather than guess what future housing price changes will be, the framework presented here is an example of what someone could create to guide investment decisions. A similar analysis applied to the stock market guides the creation of our ideal stock market exposure.
Ideal Stock Market Exposure
Our Long Term Reversal Factors and Intermediate Trend Factors remain negative. Our Short Term Trend Factors remain positive. Our ideal stock market exposure remains 20%.
