Three essays on asset pricing
<p>This thesis encompasses three original research studies in asset pricing, accompanied by an introduction, a comprehensive literature review, and concluding remarks. The first two studies delve into the term structure of equity returns, while the third study centres on tax-loss harvesting wi...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
2024
|
| Subjects: |
| _version_ | 1811140437660925952 |
|---|---|
| author | Li, W |
| author2 | Schmalz, M |
| author_facet | Schmalz, M Li, W |
| author_sort | Li, W |
| collection | OXFORD |
| description | <p>This thesis encompasses three original research studies in asset pricing, accompanied by an introduction, a comprehensive literature review, and concluding remarks. The first two studies delve into the term structure of equity returns, while the third study centres on tax-loss harvesting with ETFs.</p>
<p>The term structure of equity return volatility exhibits temporal variability, affecting the term structure of equity returns through the volatility feedback effect and explaining the cyclicality of the equity return term structure. By analysing the dividend strip futures, the first study finds that the volatility feedback effects of dividend strips decrease with the horizon. Using realised and implied volatilities as business cycle indicators, the study substantiates the pro-cyclical nature of the term structure of equity returns. The decomposition of cyclicality shows that the pro-cyclicality comes from the high relative sensitivity of short-duration volatility. Notably, the predictable cyclicality presents a novel criterion for testing macro-finance models, ultimately leading to the rejection of the rare disaster model proposed by Gabaix (2012).</p>
<p>The subsequent original research uncovers a puzzling phenomenon termed the "short-duration equity return puzzle"—short-duration dividend strips exhibit high conditional Sharpe ratios during crises, surpassing theoretical upper bounds. Using dividend prices and forecasts, this study calculates the required rate of return and conditional Sharpe ratio for dividend strips from 2002 to 2021. Notably, during crisis periods, the required rate of return for 1-year dividends peaks at 55\%, accompanied by a conditional Sharpe ratio exceeding 14, far surpassing theoretical predictions by mainstream macrofinance models. This anomaly persists across various horizons and remains robust to measurement errors and transaction costs.</p>
<p>The third study uncovers a new source of tax efficiency for ETFs—using highly correlated ETFs to harvest capital losses without violating the wash-sale rule. By exploiting the tax loophole, investors can potentially earn a better return than the index. This research reveals that the tax-loss trading volume of highly correlated ETFs accounts for 20.7% of their total trading volume. Tax-loss harvesting is negatively related to past returns, especially for recent and negative ones. ETFs with high past volatility have higher tax-loss trading volumes, while smaller and less liquid ones have lower tax-loss trading volumes. A parsimonious model is developed to elucidate the relationship between tax-loss harvesting and past price movements. Simulations with the model predict an annual tax revenue loss of 0.52% of assets under management for highly correlated ETFs, equivalent to approximately 25 billion USD in 2021.</p> |
| first_indexed | 2024-09-25T04:21:58Z |
| format | Thesis |
| id | oxford-uuid:56749e1b-072c-44c5-886a-90480137e8c2 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-09-25T04:21:58Z |
| publishDate | 2024 |
| record_format | dspace |
| spelling | oxford-uuid:56749e1b-072c-44c5-886a-90480137e8c22024-08-19T11:40:00ZThree essays on asset pricingThesishttp://purl.org/coar/resource_type/c_db06uuid:56749e1b-072c-44c5-886a-90480137e8c2FinanceEnglishHyrax Deposit2024Li, WSchmalz, MWilson, M<p>This thesis encompasses three original research studies in asset pricing, accompanied by an introduction, a comprehensive literature review, and concluding remarks. The first two studies delve into the term structure of equity returns, while the third study centres on tax-loss harvesting with ETFs.</p> <p>The term structure of equity return volatility exhibits temporal variability, affecting the term structure of equity returns through the volatility feedback effect and explaining the cyclicality of the equity return term structure. By analysing the dividend strip futures, the first study finds that the volatility feedback effects of dividend strips decrease with the horizon. Using realised and implied volatilities as business cycle indicators, the study substantiates the pro-cyclical nature of the term structure of equity returns. The decomposition of cyclicality shows that the pro-cyclicality comes from the high relative sensitivity of short-duration volatility. Notably, the predictable cyclicality presents a novel criterion for testing macro-finance models, ultimately leading to the rejection of the rare disaster model proposed by Gabaix (2012).</p> <p>The subsequent original research uncovers a puzzling phenomenon termed the "short-duration equity return puzzle"—short-duration dividend strips exhibit high conditional Sharpe ratios during crises, surpassing theoretical upper bounds. Using dividend prices and forecasts, this study calculates the required rate of return and conditional Sharpe ratio for dividend strips from 2002 to 2021. Notably, during crisis periods, the required rate of return for 1-year dividends peaks at 55\%, accompanied by a conditional Sharpe ratio exceeding 14, far surpassing theoretical predictions by mainstream macrofinance models. This anomaly persists across various horizons and remains robust to measurement errors and transaction costs.</p> <p>The third study uncovers a new source of tax efficiency for ETFs—using highly correlated ETFs to harvest capital losses without violating the wash-sale rule. By exploiting the tax loophole, investors can potentially earn a better return than the index. This research reveals that the tax-loss trading volume of highly correlated ETFs accounts for 20.7% of their total trading volume. Tax-loss harvesting is negatively related to past returns, especially for recent and negative ones. ETFs with high past volatility have higher tax-loss trading volumes, while smaller and less liquid ones have lower tax-loss trading volumes. A parsimonious model is developed to elucidate the relationship between tax-loss harvesting and past price movements. Simulations with the model predict an annual tax revenue loss of 0.52% of assets under management for highly correlated ETFs, equivalent to approximately 25 billion USD in 2021.</p> |
| spellingShingle | Finance Li, W Three essays on asset pricing |
| title | Three essays on asset pricing |
| title_full | Three essays on asset pricing |
| title_fullStr | Three essays on asset pricing |
| title_full_unstemmed | Three essays on asset pricing |
| title_short | Three essays on asset pricing |
| title_sort | three essays on asset pricing |
| topic | Finance |
| work_keys_str_mv | AT liw threeessaysonassetpricing |