Academic year 2022-23: I encourage students with an interest in the areas of empirical finance, asset pricing, and other topics related to time series econometrics to select me as the dissertation supervisor. In particular, I am willing to supervise the following topics, though you are more than welcome if you consider any other related topic within the above areas or want to amend or change these topics.
Topic A. Testing the random walk hypothesis, the efficient market hypothesis, and/or the predictability of stock returns and/or cryptocurrencies: The Efficient Market Hypothesis (EMH) states that market prices reflect all available information (Samuelson, 1965; Fama, 1970), and no one can beat the market because market prices are not predictable. There are different forms of EMH, one of which is the Random Walk Hypothesis (RWH). Based on the RWH, the stock market moves randomly. On the other hand, behavioural economists suggest that due to overreaction, panic, human errors, etc., markets are not always efficient. The recent empirical results concerning market inefficiency are mixed. For example, Durusu-Ciftci et al. (2019) review the literature and the methodologies and conclude in favour of the RWH. However, Hill and Motegi (2019) report against RWH for the U.S. and U.K. markets during financial crises. In this context, to reconcile EMH and behavioural economics, the Adaptive Market Hypothesis (AMH) supposes that market inefficiency is time-varying. This may motivate you to investigate the AMH versus the EMH by characterizing the evolution of market inefficiency over time.
- Dehghani, M., Cho, S., & Hyde, S. (2022). Friedman plucking model and Okun’s law. Third chapter of the PhD thesis, The University of Manchester, Manchester.
- Durusu-Ciftci, D., Ispir, M. S., & Kok, D. (2019). Do stock markets follow a random walk? New evidence for an old question. International Review of Economics & Finance, 64, 165-175.
- Fama, E.F. (1970). Efficient capital markets: a review of theory and empirical work. The Journal of Finance 25, 383–417.
- Hill, J. B., & Motegi, K. (2019). Testing the white noise hypothesis of stock returns. Economic Modelling, 76, 231-242.
- Samuelson, P. A. (1965). Rational theory of warrant pricing. Industrial Management Review, 6, 13-31.
Topic B. Modeling volatility in financial markets (stocks and cryptocurrencies) and their differences: Engle (1982) captured volatility clustering by developing an ARCH model. Since then, several versions of this model have been applied to capture the volatility of different financial assets. Bollerslev (1986), Engle’s student, elaborated the ARCH into the GARCH. Glosten, Jaganathan, and Runkle (1993) test asymmetric volatility by developing TGARCH. Given the growing interest in cryptocurrency, it is appealing to apply and/or elaborate those models to a set of data for different cryptocurrencies.
- Engle, R. (2004). Risk and volatility: Econometric models and financial practice. American economic review, 94(3), 405-420.
- Engle, R. F. (1982). Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation. Econometrica: Journal of the econometric society, 987-1007.
- Bollerslev T. (1986). Generalized autoregressive conditional heteroscedasticity Journal of econometrics., 31(3), 307-327.
- Glosten, L. R., Jagannathan, R., & Runkle, D. E. (1993). On the relation between the expected value and the volatility of the nominal excess return on stocks. The journal of finance, 48(5), 1779-1801.
- Tsay, R. S. (2005). Analysis of financial time series. John wiley & sons.
Topic C. Modeling bubbles in financial markets (stocks and cryptocurrencies): The history of financial markets hints at many episodes of speculative bubbles. Speculative bubbles are often defined as a positive deviation from the fundamental price (intrinsic value) that is followed by a burst. The dot.com bubble in the late-1990s, the real estate bubble in 2005, and the cryptocurrency bubbles in 2017 and 2021 are examples of notorious bubbles. Blanchard and Watson (1983), Tirole (1985), Diba and Grossman (1988), and Johansen et al. (2000) present “rational bubbles,” several models that attempt to rationalize the formation of speculative bubbles in the stock markets. Based on this model, speculative bubbles occur even if all investors have rational expectations and know that the bubble will eventually burst. Recently, Cheah and Fry (2015) repurposed the Johansen et al. (2000) model to inspect bubbles in bitcoin markets. Can we repurpose other models, e.g., Diba and Grossman (1988), and apply them to a set of cryptocurrencies to characterize the crypto-bubbles?
- Blanchard, O. J., & Watson, M. W. (1983). Bubbles, rational expectations and speculative markets, NBER Working Paper 0945.
- Cheah, E. T., & Fry, J. (2015). Speculative bubbles in Bitcoin markets? An empirical investigation into the fundamental value of Bitcoin. Economics letters, 130, 32-36.
- Diba, B. T., & Grossman, H. I. (1988). Explosive rational bubbles in stock prices?. The American Economic Review, 78(3), 520-530.
- Johansen, A., Ledoit, O., & Sornette, D. (2000). Crashes as critical points. International Journal of Theoretical and Applied Finance, 3(02), 219-255.
- Tirole, J. (1985). Asset bubbles and overlapping generations. Econometrica: Journal of the Econometric Society, 1499-1528.
Topic D. CAPM with a time-varying beta: The main aim of the Capital Asset Pricing Model (CAPM) is to estimate the cost of capital (required return) for firms according to the market risk premium (market excess return). According to Fama and French (2004), beta is a measure of the sensitivity of an asset with respect to the variation in the market return. Later, Fama and French (1996) proposed a three-factor model by adding a size factor and a book-market ratio factor to the market excess return factor. Nevertheless, there is ample evidence that beta is not stable in the long run (Groenwold and Fraser, 1999; Chen and Huang, 2007). If this is the case, CAPM suffers from this misspecification. How to test for the stability of the betas and how to model time-varying betas is what this research will be built on. An easy way is to define a rolling window and estimate beta for each window. You will also use structural break tests to check the potential breakpoint in beta before and after an event (Covid-19 is a good example). Another model is to estimate the CAPM with a time-varying coefficient by using the Kalman filter (in MATLAB and R, there is code to run this model). Page 510 of the book written by Tsay explains this model. Additionally, a Markov-switching model for beta is an idea that comes to mind.
- Fama, E. F., & French, K. R. (2004). The capital asset pricing model: Theory and evidence. Journal of economic perspectives, 18(3), 25-46.
- Fama, E. F., & French, K. R. (1996). Multifactor explanations of asset pricing anomalies. The journal of finance, 51(1), 55-84.
- Groenewold, N., & Fraser, P. (1999). Time-varying estimates of CAPM betas. Mathematics and Computers in Simulation, 48(4-6), 531-539.
- Tsay, R. S. (2005). Analysis of financial time series. John wiley & sons.
- Chen, S.-W. and Huang, N.-C. (2007). “Estimates of the ICAPM with regime-switching betas: evidence form four pacific rim economies”, Applied Financial Economics, 17: 313- 327.
Data and methodology: Regarding the data, individual stock returns, market returns, and cryptocurrency returns or their indices are accessible in the Wharton Research Data Service (WRDS), Bloomberg Terminal, and other online databases. Depending on the topic, to address the research question(s), you will use one or some of the econometric models, including linear models, ARCH and GARCH models, correlation models, vector auto regressive, principal component analysis, structural break tests, random walk tests, asymmetric random walk models, Markov switching, state-space models, value at risk, etc.
Pre-requisite: Basic econometrics background gained through passing BMAN-71122 (time-series econometrics) and/or BMAN-70211 (cross-sectional econometrics) is expected, though motivation to learn and apply new methods is more important. It is essential to know a programming language (e.g., MATLAB, R, or Python), either by taking the corresponding university course or by self-studying. The logic and syntax of the above languages are similar enough that knowing one is enough to learn another. This is the student’s choice to select one of the above languages depending on the topic, method, and availability of the codes and packages. Although I advise using one of the above to improve programming skills, it is acceptable to use STATA.
Academic year 2021-22:
I encourage students with an interest around the areas of empirical finance and assert pricing, empirical macroeconomics and business cycles, monetary policy and other topics that is related to the time series econometrics. Regarding the research topics, I am willing to supervise the following topics. If you consider any other research topics within the above areas or want to amend/change these topics, you are more than welcomed.
- Financial markets asymmetric return distribution and asymmetric volatility: Leverage effect and volatility effect.
- Financial markets persistent and asymmetric volatility clustering.
- Speculative bubbles in asset prices, particularly stock markets and cryptocurrencies.
- Flash crashes (steep fall in a stock price at a higher frequency).
- Stock market crash (steep fall in the stock market at a lower frequency).
- Testing random walk hypothesis and efficient market hypothesis (Does market incorporates all of the available information?).
- CAPM with constant coefficient or time-varying coefficients.
- Forecasting trend and cycle output for the U.S., the U.K., or other economies.
- Quantity theory of money, zero lower bound, conventional and unconventional monetary policy before and after COVID-19.
- Asset prices, inflation, quantity and velocity of money.
Regarding the methodology, you may apply one or some methods to address the research question(s). I am familiar with these methods: Linear models (e.g., Regression), ARCH and GARCH models, correlation models, Vector Auto Regressive (VAR), Principal Component Analysis (PCA) and factor models, structural break tests, random walk tests, asymmetric random walk, Markov Switching, state-space models and trend cycle decomposition, Value at Risk (VaR), event study, difference in difference regression.