Do cay and cayMS predict stock and housing returns? Evidence from a nonparametric causality test

Abstract

We use a nonparametric causality-in-quantiles test to compare the predictive ability of the consumption-wealth ratio (cay) and the Markov Switching version (cay MS) for excess and real stock and housing returns and their volatility. Our results reveal strong evidence of nonlinearity and regime changes in the relationship between asset returns and cay or cay MS , which corroborates the relevance of this econometric framework. Moreover, both cay or cay MS are found to predict only excess stock returns over its entire conditional distribution, with the latter being a strong predictor only at certain quantiles. As for the housing market, these two consumption-wealth ratios only predict the volatility of real housing returns, with cay MS outperforming cay over the majority of the conditional distribution.

Consumption, asset wealth, equity premium, term spread, and flight to quality

“…The improvement in “out‐of‐sample” forecasting accuracy is noticeable vis‐à‐vis the “constant expected returns” benchmark model. As in the case of equity returns (Balcilar et al, ; Caporale & Sousa, ; Lettau & Ludvigson, ; Sousa, ), government bond returns (Afonso & Sousa, ; Sousa, ), and housing risk premium (Balcilar et al, ; Caporale & Sousa, ; Caporale et al, ), our evidence corroborates that equity and term premia and the premium vis‐à‐vis the “safe‐haven” asset are time‐varying.…”

“…For example, Afonso and Sousa () show that forecasts both stock returns and government bond yields, and Caporale and Sousa () and Caporale, Sousa, and Wohar () investigate the predictive power of for housing risk premia. Balcilar, Gupta, Sousa, and Wohar () account for regime changes in the relationship between the consumption–wealth ratio (and its Markov switching version) and housing returns (as well as their volatility) using a nonparametric causality‐in‐quantiles test.…”

The role of economic policy uncertainties in predicting stock returns and their volatility for Hong Kong, Malaysia and South Korea

“…4 This causality-in-quantiles test that we employ in this paper combines the frameworks of kth order nonparametric causality of Nishiyama et al, (2011) and nonparametric quantile causality of Jeong et al, (2012), and hence, can be considered to be a more general version of Nishiyama et al's (2011) test. As pointed out by Balcilar, Gupta, Sousa and Wohar (2017), the causality-inquantiles approach employed in our study has following novelties: Firstly, it is robust to misspecification errors as it detects the underlying dependence structure between the examined time series; this could prove to be particularly important, as it is well known (and as we also show below) that stock returns display nonlinear dynamics. Secondly, via this methodology, we test for causality that may exist in the tails of the joint distribution of the variables, thus not only for causality-in-mean (1st moment).…”

Predicting stock market movements with a time-varying consumption-aggregate wealth ratio

“…At this stage it must be pointed out that, as is standard practice in the literature on stock returns predictability, the above-mentioned studies rely on linear predictive regression frameworks when predicting stock returns and volatility based on the consumption-aggregate wealth ratios. But, as has recently been shown by Bekiros and Gupta (2015) and Balcilar et al (2017a), the relationship between stock-market movements and cay or cay MS is in fact nonlinear, and hence, the linear models used in the literature are misspecified and results derived from them cannot be relied upon. Interestingly, Bekiros and Gupta (2015) investigate the predictability of US stock returns and its volatility emanating from cay and cay MS using the k-th order conditional mean-based nonparametric causality test of Nishiyama et al (2011), and find no evidence of predictability.…”

“…Then, as part of our second objective, as is standard practice in the literature, we use cay TVP to predict the US equity premium over the quarterly period from 1952:Q2 to 2015:Q3, and compare its performance with the classical cay and its Markovswitching version, cay MS . Given the evidence of nonlinearity in the relationship between consumption-aggregate wealth ratios and excess stock returns as shown by Bekiros and Gupta (2015), and following Balcilar et al (2017a), we conduct the predictability analysis based on the k-th order nonparametric causality-in-quantiles test that has been recently developed by Balcilar et al (2016a).…”