- Title
- Return predictability of emerging stock markets using combination forecast and regime switching models
- Creator
- Bahrami, Afsaneh
- Relation
- University of Newcastle Research Higher Degree Thesis
- Resource Type
- thesis
- Date
- 2017
- Description
- Research Doctorate - Doctor of Philosophy (PhD)
- Description
- This study provides a comprehensive examination of stock return predictability in advanced emerging markets. These markets offer unique investment opportunities to international investors as they are weakly integrated with developed markets and generally yield robust returns as a result of financial rapid economic growth. However, emerging markets typically underperform developed markets in terms of transparency in financial reporting, investor protection provisions and extent of financial liberalisation. Since all of these factors are inversely related to stock return predictability, emerging markets may exhibit a higher degree of return predictability than their developed counterparts. The extant literature is over-represented by studies of return predictability in the context of developed markets, and more importantly, most studies provide return forecasts from an individual predictive model with time-invariant parameters. This study provides comprehensive evidence of return predictability for ten advanced emerging markets (Brazil, Czech Republic, Hungary, Malaysia, Mexico, Poland, South Africa, Taiwan, Thailand and Turkey) and overcomes methodological shortcomings of previous studies in this area. More specifically, return predictability is examined in this study using three sets of predictor variables: financial, macroeconomic and technical. These variables are theoretically motivated and applied across each of the ten advanced emerging markets to ensure consistency of the results. The predictor variables are used in models with a single predictor variable (univariate predictive model), and models with all relevant predictor variables (kitchen-sink regression model). Models with time-invariant parameters suggest that financial variables provide the best in-sample return predictability, while macroeconomic variables provide the best out-of-sample return predictability. Overall, the results are consistent with the previous findings in developed markets that none of the univariate predictive models are able to consistently outperform a historical average benchmark. This study then applies more recent methodologies that reduce forecast error variance (combination forecast method), allow model parameters to vary over time (Markov regime-switching models) and integrate the combination forecast method with a Markov regime- switching model. To the best of our knowledge these methodologies have not been used to test the predictability of returns in advanced emerging markets. The results provide consistent evidence of in-sample stock return predictability particularly when using Markov regime- switching models. Evidence of out-of-sample stock return predictability is also found when applying a combination forecast or a Markov regime-switching model. However, the strongest evidence of out-of-sample return predictability is found by combining forecasts from the individual regime-switching forecast returns. These findings are important for fund managers and investors attempting to improve investment performance through higher expected returns and risk diversification opportunities offered by emerging markets. This study shows that a risk-averse investor can attain utility gains by using forecast returns from the combination forecast and regime-switching models. Further, evidence of stock return predictability is important for researchers to develop more realistic asset pricing models for emerging markets.
- Subject
- return predictability; in-sample; regime switching model; combination forecast; emerging markets; stock market; out-of-sample
- Identifier
- http://hdl.handle.net/1959.13/1336114
- Identifier
- uon:27551
- Rights
- Copyright 2017 Afsaneh Bahrami
- Language
- eng
- Hits: 1111
- Visitors: 1437
- Downloads: 438
| Thumbnail | File | Description | Size | Format | |||
|---|---|---|---|---|---|---|---|
| View Details Download | ATTACHMENT01 | Thesis | 1 MB | Adobe Acrobat PDF | View Details Download | ||
| View Details Download | ATTACHMENT02 | Abstract | 445 KB | Adobe Acrobat PDF | View Details Download |