Digital library of construction informatics and information technology in civil engineering and construction
Digital library
Paper: eres2016_267
| Paper title: | Housing Prices at the time of QEs in California Effect of Mortgage Rates vs Bond Yields |
| Authors: | Kim, Taewon; Daniel Lee, Yalan Feng, Donald Keenan |
| Summary: | Mortgage interest rates in general move in tandem with Treasury bond yields. The corporate bond index is known to have a strong correlation with treasury yields, especially with those of five-year treasuries (see WSJ market data center, for example). It is not hard to infer then that there exists a close correlation between mortgage rates and the bond yield index, which suggests that one could use the corporate bond index (CBI) as a barometer of mortgage rates, as corporate rates are more easily available than mortgage rates, which are more localized. During the last decade-long Great Recession, however, the usually close relationship between bond yields and mortgage rates broke down at the various stages of the many Quantitative Easing (QE) programs. The degree of the divergence between these two rates during this period has been established in recent research. In this paper, we investigate how this divergence between the two rates affected housing prices differently during 2007-2014. To that end, we compare the impact on housing prices of mortgage rates on one hand and of the impact of CBI on the other, in California’s largest cities during this period. Our research shows that, once the close relationship between the two rates broke down as the result of QEs, significant divergence develops in their respective correlation coefficients on housing prices. We show the degree to which using the bond yield index as a barometer of mortgages rates can distort housing price estimates during the period of divergence between these two rates.Our two models are as follows:DLHPit = β0 + β1×DLUNRit + β2×DLUNRit(-1) + β3×DLMRit + β4×DLMRit(-1) + β5×DLMRit(-2) + β6×Recess1 + β7×Recess2 + β8×HPD1 + β9×HPD2+ β10×DLHPit(-1)+ β11×DLHPit(-2)+ β12×DLHPit(-3)+ β13×DLHPit(-4) + uit + εit (1)DLHPit= β0 + β1×DLUNRit + β2×DLUNRit(-1) + β3×DLCBIit + β4×DLCBIit(-1) + β5×DLCBIit(-2) + β6×Recess1 + β7×Recess2 + β8×HPD1 + β9×HPD2+ β10×DLHPit(-1)+ β11×DLHPit(-2)+ β12×DLHPit(-3)+ β13×DLHPit(-4) + uit + εit (2) whereDLHPit= the natural logarithm of housing price at time t DLUNRi= the natural logarithm of the unemployment rate at time tDLMRit = the first difference of natural logarithm of mortgage rate at time t. This data is not available for individual cities over timeRecess1 = recession period as defined by the National Bureau of Economic Research, i.e. Recess1 = 1 for period from March 2001 to November 2001, and 0 otherwise (NBER)Recess2 = recession period as defined by NBER, i.e. Recess2 = 1 for period from December 2007 to June 2009, and 0 otherwise (NBER).HPD1 = 1 for Feb. 1996 to Feb. 2006 and HPD1 = 0 otherwise (Period 1 in which housing prices were rising from February 1996 to February 2006)HPD2 = 1 for March 2006 to Jan 2012, HPD2= 0 otherwise. (Period 2 in which housing prices were falling from March 2006 to January 2012)DLCBI = the logarithm of the corporate bond yield indexIn addition, ui represents between-city errors created by all other unobserved time-invariant variables that influence the dependent variable. The term εit is the random disturbance for the ith city at tth time period with E (εit) = 0. It is assumed that ԑit is uncorrelated with the independent variables and with ui and that COVεit,εis = 0 for t ≠ s.In both models, the dependent variable is the housing value. Our sample cities are sixty-one cities in California with the population of over 100,000 as of 2005.They are Anaheim, Antioch, Bakersfield, Berkeley, Burbank, Chula Vista, Concord, Corona, Costa Mesa, Daly City, Downey, El Monte, Elk Grove, Escondido, Fairfield, Fontana, Fremont, Fresno, Fullerton, Garden Grove, Glendale, Hayward, Huntington Beach, Inglewood, Irvine, Lancaster, Long Beach, Los Angeles, Modesto, Moreno Valley, Norwalk, Oakland, Oceanside, Ontario, Orange, Oxnard, Palmdale, Pasadena, Pomona, Rancho Cucamonga, Richmond, Riverside, Roseville, Sacramento, Salinas, San Bernardino, San Diego, San Francisco, San Jose, Ventura, Santa Ana, Santa Clara, Santa Clarita, Santa Rosa, Simi Valley, Stockton, Sunnyvale, Thousand Oaks, Torrance, Vallejo, Visalia and West Covina. But Elk Grove doesn’t have data and the list is reduced to 61 in our data set.The results show that, while both mortgage rates and the corporate bond yield index both have a strong explanatory power on housing values as can be expected, the regression coefficients become consistently weaker as the regression coverage period approaches the start of the QE (Table 1 & 2). We note however, as the sample period nears the housing collapse of the 2000s, the corporate bond yield index coefficients decrease more precipitously than the mortgage rate coefficients i.e. the effect of the corporate bond yield index on housing prices becomes noticeably and consistently weaker over time, more so than the mortgage rate coefficients do (see Table 1). One could explain this as the natural byproduct of the Treasury buy-back QE policies which suppressed the bond rates more than they affected mortgage rates, hence the divergence in values of these two variables as noted in Kim, Lee and Tran (2015). Keywords: QEs, Correlation of the mortgage rate and corporate bond yield index, economic recession, housing recession, housing value |
| Type: | paper session |
| Year of publication: | 2016 |
| Series: | ERES:conference |
| Download paper: | |
| Citation: | Kim, Taewon; Daniel Lee, Yalan Feng, Donald Keenan (2016). Housing Prices at the time of QEs in California Effect of Mortgage Rates vs Bond Yields . 23rd Annual European Real Estate Society Conference in Regensburg, Germany, http://itc.scix.net/paper/eres2016_267 |
hosted by University of Ljubljana
includes:
W78
includes:
W78
inspired by SciX, ported by Robert Klinc [2019]