ABSTRACT
This study investigates the directional predictability of financial indicators for home sales across tranquil (1984–2005) and volatile (1972–1983 and 2006–2013) periods. We find that the mortgage rate has directional predictability for both existing and newly built home sales for up to 2005. The federal funds rate generally has directional predictability for existing (newly built) home sales in 1984–2005 (1972–1983). The term spread has directional predictability for home sales in 1972–1983 but generally not in the tranquil period of 1984–2005. Further, unlike mortgage and federal funds rates, the term spread has directional predictability for home sales in 2006–2013 and thus can help the Fed with useful information (assuming that this trend continues).
Acknowledgement
The authors would like to thank an anonymous referee for helpful comments and suggestions.
Notes
1 See, for instance, Sims (Citation1980) and Bernanke and Blinder (Citation1992) for short-term interest rates, and Estrella and Hardouvelis (Citation1991), Estrella, Rodrigues, and Schich (Citation2003), Estrella (Citation2005) and Christiansen (Citation2013) for term spreads (the difference between long- and short-term interest rates).
2 We reach similar conclusions when we adjust the term spread for the mean value calculated over the sample periods. The mean values are 1.156, 1.963, 1.936 and 1.734 for, respectively, 1972.3–1983.4, 1984.1–2005.4, 2006.1–2013.4 and 1972.3–2013.4. Since these mean values are not known at the time of the forecast, our approach in using the moving average for adjusting the term spread is more appropriate.
3 Consistent with the orthogonality test results in , our results (not reported here) indicate that the change in house prices does not have directional predictability for home sales for 1972–2013. The same is true for both the tranquil and volatile periods.