- Title
- Idiosyncratic volatility and returns in international equity markets
- Creator
- Goy, Oliver Charles
- Relation
- University of Newcastle Research Higher Degree Thesis
- Resource Type
- thesis
- Date
- 2021
- Description
- Research Doctorate - Doctor of Philosophy (PhD)
- Description
- Idiosyncratic volatility (IV) is the volatility that is individual to a stock, identified by the residuals from an asset pricing model. IV has received substantial attention in the literature, with studies examining the IV anomaly and the trend in IV over time. One area of contention amongst these studies is the method used to estimate (and price) IV. Implementing a range of commonly used asset pricing models and alternative specifications of risk factors, this thesis investigates the effect of these different estimation methods on the aggregated distribution of IV. For a sample of 23 developed nations over a 28-year period, the results show that distributions of IV are statistically different when estimated by different pricing models in all countries except Germany. Similarly, local versions of risk factors produce statistically different distributions of IV to non-local risk versions of risk factors. Additionally, eight countries have IV distributions that are substantially affected when controlling for illiquidity. As a result, multiple methods of IV estimation and liquidity controls are recommended when carrying out an international study. The trend in IV over time has implications on the benefits of portfolio diversification and the relative performance of under-diversified portfolios over time. To date, there is no consensus regarding IV time-trend findings, with earlier studies identifying a positive trend, but more recent research showing that the trend is time-varying and sample-period dependent. By mimicking sample period breakpoints from previous studies, the results of this thesis show that IV is positively trending in most countries during the 1990s, negatively trending during the 2000s, and has no statistically significant trend during the 2010s. These results are robust to the choice of IV estimation model, specification of local or global risk factors, macro-economic shocks and liquidity controls. Therefore, non-linear changes to risk variables are likely candidate explanations for the trend in IV over time. Using a novel control for market volatility, the IV-to-total-return-volatility ratio (IV/MKT) is assessed for its time-trend properties. Most countries have a positive trend in IV/MKT during the 1990s, a negative trend during the 2000s, and a positive trend during the 2010s. Changes in IV and market volatility were not found to be uniform over time, indicating a change to portfolio diversification benefits. As the proportionality of unsystematic risk to systematic risk is decreasing during the 2000s, the unsystematic risk reduction benefits from diversification are lessening. However, idiosyncratic risk is increasing relative to systematic risk during the 1990s and 2010s, improving diversification benefits. Whilst global investors can diversify across markets, thereby reducing some systematic risks, IV/MKT trends lower during economic crises, evidencing higher systematic risk portions of price volatilities internationally. However, IV/MKT displays mean reversion characteristics post-crises, affecting the time-trend findings. Using a zero-investment IV anomaly trading strategy, where low-IV portfolios are bought and high-IV portfolios are sold, this thesis provides economically large positive risk-adjusted returns in 20 developed countries. Low-IV portfolio outperformance is a consistent feature in 19 out of the 20 countries, whereas high-IV portfolio underperformance is only found in six countries. The underperformance of high-IV portfolios as a driver for the IV anomaly is found in the US but is not synchronous internationally. Instead, positive risk-adjusted returns can be made with a long-only IV anomaly trading strategy in low-IV portfolios. Additionally, the IV anomaly is present in portfolios of large stocks, indicating a viable international investing strategy with fewer market frictions. Whilst the method of IV estimation is found to affect the subsequent distribution of IV, this thesis shows that the statistical significance of the IV anomaly is robust to the choice of pricing model and specification of risk factors in most countries. Finally, temporary high variance regimes generally do not affect an IV anomaly trading strategy; returns to low-IV and high-IV portfolios are negatively impacted during crises, but the changes are proportional, leaving the profitability of a zero-investment strategy unchanged.
- Subject
- idiosyncratic volatility; idiosyncratic volatility anomaly; stock market; international equity markets
- Identifier
- http://hdl.handle.net/1959.13/1426844
- Identifier
- uon:38479
- Rights
- Copyright 2021 Oliver Charles Goy
- Language
- eng
- Full Text
- Hits: 1143
- Visitors: 1443
- Downloads: 319
Thumbnail | File | Description | Size | Format | |||
---|---|---|---|---|---|---|---|
View Details Download | ATTACHMENT01 | Thesis | 2 MB | Adobe Acrobat PDF | View Details Download | ||
View Details Download | ATTACHMENT02 | Abstract | 196 KB | Adobe Acrobat PDF | View Details Download |