Search

The three main ways to estimate future volatilities include the implied volatility of option prices, time-series volatility models, and neural network models. This project investigates whether there are economically meaningful differences between those approaches. Seminal time-series models like the GARCH, as well as recurrent neural network models like the LSTM are investigated to forecast volatilities. An eventual informational advantage over the market’s expectation of future volatility in the form of implied volatility is sought after. Through trading strategies involving options, as well as investment vehicles that emulate the VIX, it is attempted to trade volatility in a profitable way.

Previous research on residual momentum indicated that it performs well compared to total momentum strategies as well as during market turmoil. This paper analyzed the performance of a residual momentum strategy based on two different factor models and multiple weighting schemes. Evidence is found of a volatility weighted residual momentum strategy outperforming the S&P 500 Index over a time span of 20 years and generating a statistically significant alpha. It failed, however, to outperform the S&P 500 Index in the out-of-sample period that ranges from 2011 to .2021 and thus opens the door to further enhance the signal

Both momentum and value strategies earn consistent and significant premia and are negatively correlated, with their equal weight combination improving the risk-return trade-off. This paper shows that allocation based on market volatility further improves the risk-return trade-off, particularly by limiting the large drawdowns momentum experiences in market crashes, where value tends to perform better. Both long-short strategy legs achieve comparably low Sharpe ratios in the past 20 years. There is no clear picture of high momentum stocks performing better than their low momentum counterparts, similar for value, which seems to off-set the long-short returns, while the long legs perform comparably well. The group report tests the combination of five different sub strategies, resembling the performance of a multi-strategy hedge fund benchmarked against the popular buy-and-hold S&P 500 investing approach. The sub-strategies are: residual momentum, value including intangibles, value and momentum, volatility forecasting, and a long short-term memory strategy, the latter two being machine-learning-based, and all investing in the U.S. universe. The combined strategy’s performance is analyzed by three weighting schemes: equal-weight, momentum, and mean variance, resulting in a gamut of robustness and performance. The combined strategies reap diversification benefits, thereby giving investors a superior risk-reward trade-off compared to the buy-and hold S&P 500 approach

This study re-investigates the relationship between the equity premium and variables, that have been either examined for predictive qualities by the literature or are used by active fund managers at the time of the study, with state-of-the-art machine learning models to uncover potential non-linear relationships. Results show that within the U.S. market, no variable, whether suggested by the literature or by active fund managers, is able to forecast the equity premium. However, it appears as though there is a predictive relationship between European long-term yields and the European equity premium.

This report aims to analyze the performance of along-short investment strategy over a 12-year sample period. The strategy is constructed upon the approach of Fama and French (1992) to value investing and the marginality of firms, as advanced by Chan and Chen (1991). A short-term reversal component is subsequently implemented to take advantage of the interplay between value and marginality. Notwithstanding considerable barriers to its implementation, and while past returns do not necessarily predict future returns, the strategy delivers the best risk-adjusted return relative to considered benchmarks, averaging an annual return of 12.47% and 9.54%of annual volatility.

This report examines the performance of a quantitative investment strategy in commodity futures that incorporates term-structure information. A weekly rebalanced long-short portfolio of distant futures contracts based on the roll-yield provides significantly higher risk-adjusted returns than the second month contracts. Controlling for reasonable transaction costs and excluding the most illiquid contracts, the performance is superior to the second month contracts and especially to the benchmarks like the GSCI index and long-only GSCI constituents portfolio. Hence, it is a valuable addition to our group strategy which combines it with four equity strategies and significantly outperforms the S&P500 on a risk-adjusted basis.

The group report analyses the joint impact of Momentum, Value, Size, and Low Volatility in factor-based investing in sovereign bond emerging and developed markets, develops an ESG factor and assesses its effect across all strategies. This paper focuses on a Value based strategy implemented with OAS and finds that controlling for credit risk is essential to capture positive returns exclusive of OAS. Results show that ESG captures bonds risk premium and allows to better control for risk, presenting a new approach to sovereign bonds market and complementing the already existing literature that shows the impact of ESG on corporate credit spreads.

The group report analyses the joint impact of Momentum, Value, Size, and Low Volatility in factor-based investing in sovereign bond emerging and developed markets, develops an ESG factor, and assesses its effects across several strategies. This individual report focuses on a Size approach and finds that the total value of government debt can be used to achieve higher risk-adjusted returns. Across the 13 years of data, the results show that ESG scores have little impact on top of the Size factor. This new integrated approach to sovereign bond markets adds to the recent and increasing literature on ESG investing.

The group report analyses the joint impact of Momentum, Value, Size, and Low Volatility in factor-based investing in sovereign bond emerging and developed markets, develops an ESG factor and assesses its affect across all strategies. This paper explores different Momentum based strategies in sovereign emerging bond markets. Results show that cross-sectional momentum strategies can exhibit meaningful risk-adjusted returns above the market. The alternative market time-series strategy explored in this paper revealed to capture dynamics of asset class preferences. The ESG integration improved the risk-return performances, reducing the presence of outliers from the universe, thus suggesting ESG materiality for fixed income investing.

The group report analyses the joint impact of Momentum, Value, Size, and Low-Volatility in factor-based investing in sovereign bond emerging and developed markets, develops an ESG factor and assesses its effect across all strategies. This paper focuses on a low-volatility strategy, with volatility being measured as the standard deviation of the yield-to-maturity, and finds that controlling for credit risk and interest rate risk is fundamental to apply it. Results show that the integration of the ESG factor complements the low-volatility strategy and produces higher risk-adjusted returns.

This study examines whether it is possible to classify the sentiment of earnings conference calls of U.S. publicly traded companies not by using standard metrics such as standardized unexpected earnings, but by but using the general sentiments, opinions and affective states present in the earnings calls. This classification task is attempted using the naïve Bayes classifier. Results show that due to the high signal to noise ratio present in the earnings calls, the classifier of choice was unable to adequately distinguish positive earnings calls from negative ones and vice-versa. Nonetheless, the classifier did shed light on the extent to which company CEOs tend to be overly optimistic when partaking in the conference calls.