12 April 2023 Economics & Growth | Rates | Real Estate | US

How to Predict US House Prices

By Sam van de Schootbrugge
5 min read

Summary

A Swiss Finance Institute research paper investigates the best variables for predicting US house prices.
They asses the backcasting, nowcasting, and forecasting abilities of over 70 macroeconomic and financial variables at low and high frequencies.
They find that the best predictors of house prices are high-frequency financial variables, such as REITs, the S&P 500, and construction firm returns.

Introduction

Housing demand has dropped sharply because of rising interest rates. This is already hurting house prices, with the UK and EU countries such as Germany and the Netherlands posting negative YoY changes.

How far could house prices fall? The housing market is: (a) a large source of wealth for households and a key determinant of consumer spending, (b) linked to the health of the banking sector, making it an important predictor of financial crises, and (c) forms a large part of many portfolios.

Predicting house prices is hard. The relationship between real interest rates and house prices suggests they could fall between 20–60% globally. However, a Swiss Finance Institute research paper recommends using high-frequency financial variables, rather than macroeconomic ones, to predict house prices.

Relative to financial predictors, macroeconomic variables offer limited improvement, with housing permits a notable exception. And during periods of economic stress and market volatility, financial predictors offer a far better indication of current house prices than, for example, the Case-Shiller index.

Predicting House Prices

Most real estate indices, including the industry standard Case-Shiller (CS) index, are released with several months’ delay. Therefore, home price valuations made today using the CS repeated-sales index from two months ago might miss relevant information.

The authors of a Swiss Finance Institute research paper examine what information can be elicited from other macroeconomic and financial variables to improve the accuracy of today’s house price valuations.

The question they ask is straightforward. Assuming the current month is March and the latest CS data is available for January, can house prices in February (backcasting), March (nowcasting), April and June (forecasting) be better predicted with additional macroeconomic and financial variables rather than just using past CS values?

The benchmark is a simple autoregressive model using past CS values, which they pitch against a forecasting model that includes an additional 71 macroeconomic and financial indicators.

Data, Variables, and Methodology

The authors collect data from 1987 to 2019 on seasonally adjusted S&P/Case-Shiller home price indices (HPI), 38 financial variables, and 33 widely used macroeconomic indicators (Appendix).

Financial variables include the value and equally weighted S&P 500 returns; the spread between the 30-year fixed-rate mortgage and the 30-year Treasury rate; the spread between the 10-year and three-month Treasury yield; returns to aggregate, mortgage, residential, equity, and hybrid REITs; and the credit spread. Data is available daily and monthly.

The macroeconomic variables are available monthly and include the Chicago Fed National Activity Index, the unemployment rate, inflation rate, non-farm payrolls, new privately owned housing units started and permits, total industrial production along with twelve disaggregated indices, the ISM Report on Manufacturing Business Employment, and various average weekly hours measures. For all these predictors, they use monthly growth rates.

Their empirical strategy has four dimensions. First, they assess which of the 71 variables can improve real estate price predictability. Second, they determine whether grouping predictors can yield even better results. Third, they use MIDAS (a mixed data sampling method) to examine whether higher-frequency, daily financial data can improve forecasting performance. Finally, they see whether their results hold across 19 US metropolitan statistical areas (MSA).

The authors estimate all models (180 different specifications) using ordinary least squares (OLS), both in-sample and out-of-sample. Variables contain at most three lags to avoid the full ‘kitchen-sink’ approach. They measure the performance of each specification by the mean square forecasting error (MSFE) relative to the benchmark autoregressive model. They break the performance down into backcasts, nowcasts, and forecasts.

Results

Starting with backcasting (that is, predicting real estate valuations from last month), S&P 500 returns, equity, aggregate and hybrid REITs returns, and most industry returns lead to significant improvements in performance relative to the benchmark.

Furthermore, all financial variables provide statistically valuable information for predicting today’s valuations (nowcasting). And for forecasting one- and three-month(s) ahead, financial variables improve prediction accuracy by 3.0–6.4%.

Overall, of the 114 distinct specifications the authors consider, 35 (31%) are significant at backcasting, 70 (61%) at nowcasting, 72 (63%) at forecasting one month ahead, and 56 (49%) at forecasting three months ahead. Across the predictors, including one- and two-month lags produces the largest increase in forecasting precision (Chart 1).

For the macroeconomic variables, only housing permits, industrial production of residential utilities, and manufacturing improve house price predictability. The results also hold at a local level and provide the best forecast for returns in Seattle.

Most financial predictors improve forecasting accuracy, while macroeconomic predictors do not. So, it is unsurprising that grouping variables leads to superior predicting performance by financial variables. When grouping, the authors recommend using a weighting scheme based on the inverse MSFE average.

Next are the results from MIDAS. Using higher-frequency (daily) financial data significantly improves nowcasting and forecasting performance – in most cases by at least 5%. Among the most robust MIDAS predictors are the returns of the S&P 500, the Ziman real estate indices, and industries such as construction and finance. These sectors are directly or indirectly related to the residential real estate market.

The authors find that predictors do particularly well in turbulent times – when they are most needed. They also find that the forecasting accuracy of a model which includes the above financial predictors improves when economic uncertainty increases (as measured by NBER recessions, VIX, and UoM Consumer Sentiment Index). For example, a $1mn property has an annual forecast error of around $33,300, which improves by $1,064 in normal months but $3,784 in crisis months.

Bottom Line

The authors take a crude approach to determining the best predictors of US house prices. However, it yields some helpful results. We often overweight macroeconomic information or housing fundamentals when forecasting the direction of prices.

Instead, in the short run, financial variables seem the best forecasters of residential real estate returns. In particular, a housing market on a downward trajectory should be seen in lower returns on REITS and construction firms.