Asset price bubbles and economic welfare

https://doi.org/10.1016/j.irfa.2016.01.011Get rights and content

Highlights

  • We examine whether or not asset price bubbles predict economic welfare.

  • Asset price bubbles both positively and negatively predict economic welfare.

  • Evidence that asset price bubbles are welfare-enhancing is much stronger.

  • Results are robust to out-of-sample forecasting.

  • Results are robust to structural breaks.

Abstract

In this paper, we provide the first empirical evidence on whether or not asset price bubbles predict economic welfare. Using a time-series model, we show that asset price bubbles both positively and negatively predict economic welfare, although the evidence that asset price bubbles are welfare-enhancing is much stronger. These results are also robust to out-of-sample forecasting as well as to a predictive regression model augmented by structural break dates.

Introduction

The global economy's connection with asset price bubbles is nothing new. The most recent relationship was witnessed during the 2008 global financial crisis, which resulted from the bursting of the US housing bubble. Asset price bubbles instigate economic repercussions because people invest during bubbles. There is no shortage of reasons why investment in bubbles is attractive. Perhaps the earliest work to document the reasons why people invest during bubbles is Bewley (1980). See also other influential studies on this subject, such as Kiyotaki and Moore (2008), Kocherlakota, 1992, Kocherlakota, 2009, and Wang and Wen (2009).3

In this paper, we examine whether asset price bubbles predict economic welfare. We define economic welfare as one that reflects the level of prosperity and living standards of an individual and/or a group of persons. It is also referred to as utility that is gained through the achievement of material goods and services (Samuelson & Nordhaus, 2004). We consider eight proxies for welfare, which include change in stocks, exports of goods and services, current price of GDP, government final consumption expenditure, gross fixed capital formation, imports of goods and services, private final consumption expenditure, and real GDP.4 Our focus is on six most developed countries, namely, Japan, Germany, Canada, France, the UK, and the USA. Our choice of these countries is motivated by the availability of consistent historical time-series data.

We contribute to the literature in two ways. First, we find that asset price bubbles exist. In this regard, our findings are consistent with the literature, which has documented asset price bubbles using other econometric methods, such as unit root and cointegration tests. These studies include Anderson, Brooks, and Katsaris (2010); Cerqueti and Costantini (2011); Cunando, Gil-Alana, and Perez de Garcia (2005); Evrensel and Kutan (2006); Gonzalez and James (2007); Koustas and Serletis (2005); McMillan (2007); and Sarno and Taylor (1999).5 However, our contribution to this literature is different. While this literature has confirmed bubbles, we not only confirm this but also extract bubbles over time. This is how we achieve this. Using a procedure developed by Phillips, Wu, and Yu (2011), we extract bubbles (or explosive behaviour). We proxy this with the time-varying t-test statistic that is used to test the null hypothesis of no explosive behaviour.6 In other words, because we use daily stock price (and dividend yield) data to search for bubbles, the day on which the t-test statistic that examines the null hypothesis of no bubble is greater the than critical value is characterised as the day of bubble.

To examine whether asset price bubbles predict economic welfare, we propose a bivariate time-series predictive regression model. Using this model, we test the null hypothesis of no predictability. We show that asset price bubbles can predict economic welfare in at least 50% of the regression models, suggesting that while bubbles do not predict all eight economic variables, they predict at least half of them. This is our second contribution. Although none of the studies specifically consider whether asset prices empirically contribute to economic welfare, there are theoretical evidence suggesting that bubbles could either positively (see, for instance, Ventura, 2012) or negatively (see, for instance, Chauvin, Laibson, & Mollerstrom, 2011) predict economic welfare. These studies are reviewed in Section 2. Our empirical findings confirm these theoretical predictions although we find greater evidence that bubbles positively predict welfare—therefore, in our story, bubbles are welfare-enhancing. Our empirical findings also contribute to another group of studies (see, for instance, Caballero et al., 2006, Jermann and Quadrini, 2007, Jerzmanowski and Nabar, 2008) that show that stock markets positively affect economic welfare through productivity growth.

As a final point, we confirm the robustness of our results in different ways. First, we undertake an out-of-sample forecasting exercise where we compare the bubble-based forecasting model with a constant-only based welfare model. From this exercise, we find that results are consistent with those obtained from in-sample tests. In in-sample tests, predictability was found in 52% of the cases, whereas in out-of-sample tests, the bubbles-based forecasting model beats the constant-only model in 46% of the cases.

Second, we are mindful of the fact that since we are using historical time-series data structural breaks (if present, which is likely, and we confirm this to be true) can distort the outcomes regarding the test of the null hypothesis of no predictability. The effect of structural breaks on predictability has been shown by Narayan and Gupta (2015); Narayan, Narayan, Popp, and Ahmed (2015). Motivated by this literature, we augment the predictive regression model with structural break dummy variables. These structural breaks relate to the welfare variables. This means that each time we run a predictive regression model, the dummy variables relate to the two structural breaks corresponding to the dependent (welfare) variable. Our results from structural break predictive regression models are consistent with those obtained from a model without structural breaks in that the null hypothesis of no predictability is rejected in 50% of the cases.

The balance of the paper is organised as follows. In Section 2, we discuss the theoretical motivation for the asset price bubble and welfare relationship. Based on this motivation, we propose an empirical framework that allows us to examine the null hypothesis of no predictability, where the asset price bubble is treated as the predictor variable. In Section 3, we discuss the main findings. In the final section, we provide concluding remarks.

Section snippets

Theory and empirical framework

There are two objectives of this section. First, we establish the link between asset price bubbles and economic welfare. There is a growing body of literature, albeit theoretical, that contends that asset price bubbles are either positively or negatively related to economic welfare. We attempt to capture this conceptual relationship. Second, using the theoretical framework, we propose a bubbles-welfare predictive regression model. In this model, we use asset price bubbles as a predictor

Preliminaries

Our final data set is quarterly. However, we do use daily time-series data on stock price indices and dividend yield to test for explosive behaviour (bubbles). Different countries have different sample sizes, dictated by data availability. The first column of each country in Table 1 notes the different start dates for each of the nine variables. The end date is 2014Q4 for all variables. The daily stock price index and the dividend yield are obtained from BLOOMBERG, while quarterly time-series

Concluding remarks

In this paper, we show how asset price bubbles predict economic welfare. We consider six developed countries' stock markets, devise an empirical framework to compute a proxy for asset price bubbles, and use as many as eight macroeconomic variables as proxies for economic welfare. We end up with a unique time-series quarterly data set on asset price bubbles and economic welfare. We then apply a time-series predictive regression model to this data set. We unravel a number of new insights. First,

Acknowledgments

We thank two anonymous referees of this journal and Professor Russell Smyth for helpful comments and suggestions on earlier versions of this paper. This research was supported by Deakin University's Strategic Research Centre scheme and we acknowledge financial support under this Scheme.

References (46)

  • Z. Koustas et al.

    Rational bubbles or persistent deviations from market fundamentals?

    Journal of Banking & Finance

    (2005)
  • D. Kumar

    Sudden changes in extreme value volatility estimator: Modelling and forecasting with economic significance analysis

    Economic Modelling

    (2015)
  • H. Lehkonen

    Bubbles in China

    International Review of Financial Analysis

    (2010)
  • J. Lewellen

    Predicting returns with financial ratios

    Journal of Financial Economics

    (2004)
  • M.-A. Luik et al.

    Bubbles over the U.S. business cycle: A macroeconometric approach

    Journal of Macroeconomics

    (2014)
  • D.G. McMillan

    Bubbles in the dividend-price ratio? Evidence from an asymmetric exponential smooth-transition model

    Journal of Banking & Finance

    (2007)
  • P.K. Narayan et al.

    Are Indian stock returns predictable?

    Journal of Banking & Finance

    (2015)
  • P.K. Narayan et al.

    Has oil price predicted stock returns for over a century?

    Energy Economics

    (2015)
  • L. Sarno et al.

    Moral hazard, asset price bubbles, capital flows and the East Asian crisis: The first tests

    Journal of International Money and Finance

    (1999)
  • N. Todorova

    The course of realized volatility in the LME non-ferrous metal market

    Economic Modelling

    (2015)
  • H.L. Vogel et al.

    An analytical review of volatility metrics for bubbles and crashes

    International Review of Financial Analysis

    (2015)
  • J. Ventura

    Bubbles and capital flows

    Journal of Economic Theory

    (2012)
  • K. Anderson et al.

    Speculative bubbles in the S&P500: Was the tech bubble confined to the tech sector?

    Journal of Empirical Finance

    (2010)
  • Cited by (0)

    This project was supported by funding provided by the Deakin University's Strategic Research Centre Scheme.

    1

    Tel.: + 61 3 92 446 180.

    2

    Tel.: + 61 3 92 446 871.

    View full text