Quantpedia Premium Update – August 2nd

New strategies:

#898 – Convenience Yield Risk Factor Predicts Commodity Futures Returns

Period of rebalancing: Monthly
Markets traded: commodities
Instruments used for trading: CFDs, futures
Complexity: Moderately complex strategy
Backtest period: 1959 – 2018
Indicative performance: 6.93%
Estimated volatility: 15.07%

Source paper:

Prokopczuk, Marcel and Symeonidis, Lazaros and Wese Simen, Chardin and Wichmann, Robert: Convenience Yield Risk
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4341471
Abstract:
We develop a framework to quantify the convenience yield risk (CYR) inherent to each commodity futures market. Implementing our approach, we document that our novel CYR measure is informative about future commodity returns. In panel regressions, the CYR predicts future returns with a positive sign. Economically, a strategy that opens long positions in commodity markets with a higher than median CYR signal and sells the remaining commodities yields an average return of 6.93% per year. The performance of the CYR strategy cannot be explained by exposure to existing commodity strategies or other variables that capture changes in the investment opportunity set.

#899 – Conditional Currency Momentum Portfolios

Period of rebalancing: Monthly
Markets traded: currencies
Instruments used for trading: CFDs, forwards, futures
Complexity: Complex strategy
Backtest period: 1983-2022
Indicative performance: 5.03%
Estimated volatility: 9.59%

Source paper:

Iwanaga, Yasuhiro and Sakemoto, Ryuta: Conditional Currency Momentum Portfolios
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4411616
Abstract:
Currency momentum portfolios have not generated positive returns since the global financial crisis due to the strong demand of U.S. dollars. We propose conditional currency momentum strategies that incorporate information about the average forward discount, the currency market volatility, and the return dispersion of currency portfolios. Our strategy goes long in the momentum portfolio only when the average forward discount is positive, the volatility is low, and the return dispersion is low. We reveal that the conditional one-month currency momentum portfolio raises the Sharpe ratio by 0.69 and the certainty equivalent return by 6.6% per annum.

#900 – Gamma Factor Premium

Period of rebalancing: Monthly
Markets traded: equities
Instruments used for trading: stocks
Complexity: Complex strategy
Backtest period: 2019-2021
Indicative performance: 34.09%
Estimated volatility: 17.89%

Source paper:

Stagnol, Lauren and Ben Abdallah, Marc-ali and Herfroy, Patrick: Equity Convexity and Unconventional Monetary Policy
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4034299
Abstract:
In this paper, we intend to gain an understanding of the drivers of stock convexity, also known as gamma. First, using a bottom-up – firm-level – approach, we showcase that stock fundamentals, in particular metrics related to value (captured by the price-to-book ratio) and historical volatility, allow us to efficiently discriminate between convex and concave stocks. Building on this result, we investigate the ties between the gamma premium and traditional risk factors. Second, we adopt a top-down – macroeconomic driven – framework, to understand which economic environment is the most favorable to convexity: we highlight the importance of the short-term interest rate, the VIX, but also oil price dynamics in a univariate cointegrating vector. These variables share long-term relationships. We then evaluate the ability of different models to forecast future convexity premium dynamics. Finally, we seek to employ these signals in the design of a systematic long convexity strategy and show that it leads to significantly improved risk-adjusted returns compared to a capitalization-weighted benchmark, especially in turbulent markets. Convexity exposure appears particularly relevant in a context of monetary policy normalization.

#901 – The High Resolution Term Structure of Stock Return Predictability

Period of rebalancing: Monthly
Markets traded: equities
Instruments used for trading: stocks
Complexity: Very complex strategy
Backtest period: 1984-2019
Indicative performance: 18.3%
Estimated volatility: 22.06%

Source paper:

Skouras, Spyros: The High Resolution Term Structure of Stock Return Predictability
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4411451
Abstract:
The dependence of cross-sectional stock returns on their past displays striking patterns when past returns are defined over granular intervals. A regression of monthly returns on 1600 daily lags reveals quarterly and annual waves and an inverted-U pattern in coefficient estimates that are useful in a monthly investment strategy. Daily return regressions reveal periodic predictability spikes and that predictability is jointly affected at all lags by market conditions. These patterns drive standard reversal and momentum effects and contribute to seasonality effects in monthly returns. They therefore have significant implications for theoretical models of such effects

#902 – Arbitrage Opportunities from MSCI Index Reconstitutions In Asian Stock Markets

Period of rebalancing: Daily
Markets traded: equities
Instruments used for trading: stocks
Complexity: Complex strategy
Backtest period: 2006-2021
Indicative performance: 10.1%
Estimated volatility: –

Source paper:

Qian, Shuoge and Chang, Xin and Luo, Jiang and Peng, Jiaxin and Tan, C W: Arbitrage Opportunities from MSCI Index Reconstitutions In Asian Stock Markets
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4476422
Abstract:
Using the reconstitution of MSCI indices in seven Asian markets from 2006 to 2021, we discover arbitrage opportunities arising from index-tracking funds’ efforts to minimize tracking errors around the dates when index reconstitution changes become effective (i.e., effective dates). We document pronounced abnormal returns and trading volume on the last trading day before the effective date. Arbitrageurs can exploit this predictable pattern of stock price changes and earn sizable abnormal returns if they long the added and short the deleted stocks on the announcement date and close their positions at the end of the day before the effective date. Additional analysis reveals how index-tracking investors and arbitragers trade against each other to shape stock prices and equity-lending activities around MSCI index reconstitutions.

New research papers related to existing strategies:

#822 – Negative ESG Premium in Chinese Stock Market

Ni, Yinan and Sun, Yanfei: Environmental, Social, and Governance Premium in Chinese Stock Markets
https://ssrn.com/abstract=4222515
Abstract:
Financial markets have increasingly adopted the concept of ESG (environmental, social, and governance); this paper studies the evolving effect of corporate ESG performance on the stock returns in China’s stock markets. Utilizing the Paris Agreement and China’s President Xi’s pledge to achieve carbon neutrality by 2060 as ESG shocks, we find that firms with lower ESG scores provide higher stock returns after the announcement of the Paris Agreement. Furthermore, the effect of ESG performance heightens after Xi’s pledge. Using sorted portfolios and Fama–French factor models, we find that investors are rewarded for bearing ESG-related risks. Our estimated monthly ESG risk premium is between 0.52% and 0.61%, while state-owned firms with larger market capitalizations and better financial and operational performance tend to have better ESG performance.

Qin, Kaiyue: ESG Rating, Investor Attention, and Stock Returns in China
https://ssrn.com/abstract=4312543
Abstract:
We study whether ESG ratings can predict stock returns in China. We find marginal evidence that stocks with higher ESG ratings have lower future returns. In addition, we explore the cross-sectional and time-series heterogeneities of the relationship between ESG and stock returns. We find the predictability of ESG rating is stronger for stocks in the industries that are important to climate change, and the relationship is only significant after 2018 when climate change is evident for investors in China. Overall, our findings show that investors’ attention is crucial for the stock return predictability of ESG ratings in China.

Zhang, Ning and Zhang, Yue and Zong, Zhe: Fund ESG Performance and Downside Risk: Evidence from China
https://ssrn.com/abstract=4323648
Abstract:
Whether responsible investing reduces portfolio risk remains open to discussion. We study the relationship between ESG performance and downside risk at fund level in the Chinese equity mutual fund market. We find that fund ESG performance is positively associated with fund downside risk during the period between July 2018 and March 2021, and that the positive relationship weakens during the COVID-19 pandemic. We propose three channels through which fund ESG performance could affect fund downside risk: (i) the firm channel in which the risk-mitigation effect of portfolio firms’ good ESG practices could be manifested at fund level, (ii) the diversification channel in which the portfolio concentration of high ESG-rated funds could amplify fund downside risk, and (iii) the flow channel in which fund ESG performance may attract greater investor flows that could reduce fund downside risk. We show evidence that the observed time-varying relationship between fund ESG performance and downside risk is driven by the relative force of the three channels.

#536 – Machine Learning Stock Picking

Noguer i Alonso, Miquel and Zoonekynd, Vincent: Equity Machine Factor Models
https://ssrn.com/abstract=4310924
Abstract:
We examine in this paper the training and test set performance of several equity factor models with a dataset of 20 years of data, 1,200 stocks and 100 factors. First, we examine several models to forecast expected returns, which can be used as baselines for more complex models: linear regression, linear regression with an L1 penalty (lasso), constrained linear regression, xgboost and artificial neural networks. Second, we present a unified framework for portfolio construction, leveraging machine learning for the whole pipeline, from the factor data to the portfolio weights, which scales to a large number of assets and predictors. The results we obtain are interesting and non trivial to interpret; non linear models models offer a more balanced outcome considering test set Sharpe ratio and turnover but linear unconstrained models show a good performance in the test set. We introduce a model-free reinforcement learning model, which uses factors to find the portfolio weights maximizing the information ratio.

#628 – Social Media Sentiment Factor

Dim, Chukwuma and Sangiorgi, Francesco and Vilkov, Grigory: Media Narratives and Price Informativeness
https://ssrn.com/abstract=4323093
Abstract:
We show that an increase in stock return exposure to media attention to narratives, measured with standard methods for extracting topic attention from news text, leads to a lower stock price informativeness about future fundamentals. Empirically, narrative exposure explains over 86% of idiosyncratic variance in the cross-section, and both narrative exposure and non-systematic information channels—idiosyncratic variance and variance related to public information—decrease stock price informativeness. Moreover, stocks with high narrative exposure demonstrate elevated trading volume. To rationalize the empirical results, we suggest a mechanism based on disagreement among investors arising due to the differential processing of information in media narratives.

And several interesting free blog posts have been published during last 2 weeks:

Optimal Market Making Models with Stochastic Volatility

The emergence of high-frequency trading has led to improvements in numerous algorithmic trading strategies. Consequently, there is a growing demand for quantitative analysis and optimization techniques to develop these strategies. We present a paper by Aydoğan et al. (2022), which discusses the derivation of the optimal prices for HFT to execute the limit buy and sell orders where a stochastic volatility model generates the mid prices of the assets in the market.

Top Models for Natural Language Understanding (NLU) Usage

In recent years, the Transformer architecture has experienced extensive adoption in the fields of Natural Language Processing (NLP) and Natural Language Understanding (NLU). Google AI Research’s introduction of Bidirectional Encoder Representations from Transformers (BERT) in 2018 set remarkable new standards in NLP. Since then, BERT has paved the way for even more advanced and improved models.

We discussed the BERT model in our previous article. Here we would like to list alternatives for all of the readers that are considering running a project using some large language model (as we do 😀 ), would like to avoid ChatGPT, and would like to see all of the alternatives in one place. So, presented here is a compilation of the most notable alternatives to the widely recognized language model BERT, specifically designed for Natural Language Understanding (NLU) projects.

Exploring the Factor Zoo with a Machine-Learning Portfolio

The latest paper by Sak, H. and Chang, M. T., and Huang, T. delves into the world of financial anomalies, exploring the rise and fall of characteristics in what researchers refer to as the “factor zoo.” While significant research effort is devoted to discovering new anomalies, the study highlights the lack of attention given to the evolution of these characteristics over time. By leveraging machine learning (ML) techniques, the paper conducts a comprehensive out-of-sample factor zoo analysis, seeking to uncover the underlying factors driving stock returns. The researchers train ML models on a vast database of firm and trading characteristics, generating a diverse range of linear and non-linear factor structures. The ML portfolio formed based on these findings outperforms entrenched factor models, presenting a novel approach to understanding financial anomalies. Notably, the paper identifies two subsets of dominant characteristics – one related to investor-level arbitrage constraint and the other to firm-level financial constraint – which alternately play a significant role in generating the ML portfolio return.

How Well Do Factor Investing Funds Replicate Academic Factors?

Cremers, Liu, B. Riley (Apr 2023) share their view on and try to answer the question: how well do factor investing funds perform? They conclude that, on average, factor-investing funds do not outperform. But using active characteristic share (ACS)—an adaption of Cremers and Petajisto’s (2009) original active share measure—, the authors demonstrate that the factor investing funds that match indexes the most have significantly better performance. An equal-weighted portfolio of factor investing funds in the lowest tercile of ACS outperforms an equal-weighted portfolio of funds in the highest tercile by 3.82% per year (t-stat = 3.89) using the CAPM and by 1.08% per year (t-stat = 2.01) using the CPZ6 model.