I am a macroeconomist with both theoretical and empirical interests. My primary research concentrates on real estate economics and macroeconometrics. I also have research that focuses on explosive dynamics and asset predictability.
In my job market paper, Real Estate and Construction Sector Dynamics in the Business Cycle, I investigate the business cycle, property-price, and investment dynamics when there is competition between households and firms for real estate. I include a construction sector into a DSGE model, which uses land, capital and labour and undertakes the production of both commercial and residential real estate. Specifically, I introduce sectoral heterogeneity by differentiating between two groups of entrepreneurs - consumption good and construction sector. This market ¬¬structure activates a `real estate substitution channel’, where economic disturbances which alter the demand for one type of real estate, by affecting the overall costs of real estate production, endogenously create a substitution with its counterpart. For example, an increase in demand for residential real estate also increases the cost of producing commercial structures which reduces the amount demanded by firms. In turn, this crowds out commercial real estate which affects the goods market in a similar way to an adverse aggregate supply shock. The estimated model reveals that housing preference shocks explain the largest part of the variation in property prices and residential investment, while technology shocks primarily drive commercial real estate prices.
Another chapter of my PhD examines the sentimental part of housing markets, using narrative evidence. In the paper Sentimental Housing Markets, I investigate the causal effect of consumer confidence on the housing market dynamics. I study the role of expectations and the influence of sentiment in the housing markets through survey evidence on consumer behaviour. I adopt an external instrument approach that is using mass fatalities to identify exogenous variations in consumer confidence. The occurrence of these tragic events and media coverage to the wider public can potentially create a wave of fear and pessimism that can affect the behaviour of the consumers, which can impact on the economy. I find that adverse sentiment shocks can negatively affect housing demand with a strong and prolonged reduction of house prices and new houses sold. The deterioration of sentiments is worsening homeownership conditions, causes a response of monetary policy, and exacerbates real consumption spending. I isolate the effect of the housing market by conducting a counterfactual experiment that restricts the effect of the house prices and new houses sold. I evaluate the quantitative effect of the housing market by measuring the difference between the restricted and the unrestricted model. The housing market can propagate the effect of the sentiment shock to the rest of the economy. The effect becomes particularly evident on longer horizons, specifically after one year, where the deviation from the unrestricted model becomes substantial.
In parallel, I also work on the development of bubble detection tests in asset markets and asset price modelling. In my paper, Speculative bubbles in segmented markets: Evidence from Chinese cross-listed stocks, which is coauthored with Eftymios Pavlidis and published in the International Journal of Money and Finance I propose a novel approach for testing for rational speculative bubbles in segmented capital markets. The basic idea is that, under capital controls, heterogeneity of speculative expectations across international equity markets causes financial assets with identical cash flow promises to trade at different prices. Because these deviations from the law of one price inherit the properties of the speculative bubble process, they display periods of explosive dynamics and have predictive power for future movements in equity prices in sample. These two hypotheses can be examined empirically using sequential unit root tests and predictive regressions. An attractive feature of this approach for bubble detection is that it does not require the specification of a model for market fundamentals, thus mitigating the well-known joint hypothesis problem. The focus of the paper is on mainland Chinese companies that cross-list shares in Hong Kong. China is an ideal setting for our analysis because of the significant restrictions on capital movements imposed by the authorities and the turbulent behaviour of its stock market over the last decades.
I have several other projects and papers that are still in the early stages of development. For example, as a natural extension to my job market paper, I investigate the role of monetary policy in affecting the behaviour of real estate investment and property prices, as opposed to other, possibly non-fundamental factors that drive house prices up and down, such as bubbles. In this paper, Real Estate and Monetary Policy, I carry out a structural analysis using a Structural Vector Autoregression approach (SVAR). The SVAR focus on the modelling of the construction sector, where both types of real estate are estimated jointly. A unified approach help understand the potential spillovers and comovements between the two real estate sectors. I focus on monetary policy, real estate demand and credit supply (borrowing) in the two markets to understand similarities and differences in a systematic manner.
My job market paper outlines the importance of the housing preference shock, especially in periods preceding economic and financial turbulence. Thus, in another paper, I apply a non-linear (Interacted) Panel VAR model to examine the asymmetries in the transmission of housing demand shocks on the real economy and financial stress. The empirical findings suggest that the impact of a house price shock is significantly stronger in periods that the housing market is under distress. To define the periods of distress in the housing market, we employ various methodological approaches such as the House Prices-at-Risk, the exuberance indicators, and also recession and systematic stress periods. The results are in line with the existing literature that supports that the wealth effect (and the marginal propensity to consume) is more substantial in periods of a distressed housing market. Regarding the financial system, we find evidence in favour of the deviation hypothesis theory. This channel supports that in the presence of significant misalignment in the market, an increase in the house prices will further increase financial stress.
Furthermore, in a recent research project named Predictability of International House Price Returns, I examine the predictability of the quarterly growth rates of the house price index (HPI), in a panel of 24 countries. In this research, I consider a broad range of macroeconomic and financial variables and set of variables that may capture the impact of international financial flows and common global business cycle factors. The methodology used utilizes a new instrumental variable-based Wald statistic (IVX-AR) which accounts for serial correlation and heteroscedasticity in the error terms of the linear predictive regression model. Preliminary results show a high degree of heterogeneity on the predictability among countries. However, the spread between the long- and short-term interest rates, residential investment, and disposable income constitute robust predictors of the growth rate of HPI.
Finally, I also work on software development which is employed to create code in an open-source language like R. I develop statistical packages that incorporate state-of-art econometric techniques and facilitate scientific research. Among other packages, I developed and maintain the R package exuber for testing and date-stamping periods of mildly explosive dynamics (exuberance) in time series, using univariate and panel recursive unit root tests. There is a Dallas Fed working paper exuber: Recursive Right-Tailed Unit Root Testing with R that provides illustrations of the package using artificial series and a panel on international house prices.