Quantpedia Premium Update – 18th March

New strategies:

Period of rebalancing: Daily
Markets traded: currencies, equities
Instruments used for trading: futures, ETFs
Complexity: Complex strategy
Backtest period: 1996-2018
Indicative performance: 7.5%
Estimated volatility: 10%

Source paper:

Park, Yang-Ho: Informed Trading in Foreign Exchange Futures: Payroll News Timing
https://ssrn.com/abstract=3983210
Abstract:
This paper studies informed trading about U.S. payrolls in the foreign exchange (FX) futures market. I find that speculators such as hedge funds are more likely to be sellers than buyers of FX futures ahead of good U.S. payroll news and thus appear to have earned significant gains around payroll announcements. In contrast, hedgers—in particular, dealers—appear to have provided liquidity to speculators. I show that mimicking speculators’ FX exposures around payroll announcements can add a large economic gain to various reference portfolios. My analysis also uncovers that information in FX trading is long-lived and differs along the U.S. business cycle.

#729 –Reversal Effect in India

Period of rebalancing: Monthly
Markets traded: equities
Instruments used for trading: stocks
Complexity: Simple strategy
Backtest period: 1997-2018
Indicative performance: 47.74%
Estimated volatility: 36.21%

Source paper:

Ali Fayyaz Munir, Mohd Edil Abd. Sukor, and Shahrin Saaid Shaharuddin: Adaptive Market Hypothesis and TimevaryingContrarian Effect: Evidence From Emerging Stock Markets of South Asia
https://journals.sagepub.com/doi/pdf/10.1177/21582440211068490
Abstract:
This study contributes to the growing debate on the relation between varying stock market conditions and the profitability of stock market anomalies. We investigate the effect of changed market conditions on time-varying contrarian profitability in order to examine the presence of the Adaptive Market Hypothesis (AMH) in South Asian emerging stock markets. The empirical findings reveal that a strong contrarian effect holds in all the emerging markets. We also find the stock return opportunities vary over time based on contrarian portfolios. We show that contrarian returns strengthen during the down state of market, higher volatility and crises periods, particularly during the Asian financial crisis. Interestingly, the market state instead of market volatility is the primary predictor of contrarian payoffs, which contradicts the findings of developed markets. We argue that the linkage arises from structural and psychological differences in emerging markets that produce unique intuitions regarding stock market anomalies returns. The overall findings on the time-varying contrarian returns in this study provide partial support to AMH. Another significant outcome of this study implies that investors in South Asian emerging markets, like investors in the developed markets, could not adapt to evolving market conditions. Therefore, contrarian profits often exist, and persistent weak-form market inefficiencies prevail in these markets.

#730 – Volatility Decomposition and Mutual Fund Returns

Period of rebalancing: Monthly
Markets traded: equities
Instruments used for trading: funds
Complexity: Moderately complex strategy
Backtest period: 2006-2019
Indicative performance: 14.52%
Estimated volatility: 32.33%

Source paper:

Vafai, Nima and Rakowski, David A., Mutual Fund Performance and the Sources of Portfolio Volatility
https://ssrn.com/abstract=4005351
Abstract:
We conduct a volatility decomposition to identify the source of performance differences between low volatility and high volatility mutual funds. A higher level of return covariance of fund holdings is associated with more fund-level exposure to the idiosyncratic volatility effect. Average security-level variance of fund holdings is only weakly associated with idiosyncratic volatility. However, the average security-level variance of a fund’s holdings is closely tied to a fund’s exposure to the beta anomaly. We demonstrate that our measure of the within-portfolio covariance of fund holdings is useful in evaluating fund-level performance measures and exposure to volatility anomalies.

#731 –Stock Trading Rule that Produces Higher Returns with Lower Risk

Period of rebalancing: Monthly
Markets traded: equities, bonds
Instruments used for trading: ETFs, bonds, funds
Complexity: Moderately complex strategy
Backtest period: 1928-2008
Indicative performance: 6.78%
Estimated volatility: not stated

Source paper:

Prentis, Eric L.: Evidence on a New Stock Trading Rule that Produces Higher Returns with Lower Risk
https://ssrn.com/abstract=3949604
Abstract:
This new stock market trading rule uses three steps to remove random unsystemic risk from stock price data to smooth volatility. Proving empirically that a technical analysis relative maxima and minima trading rule for an S&P 500 Index portfolio substantially beats a naïve buy-and-hold policy, at significantly lower risk. Calling key theories in economics and finance into question. The new trading rule succeeds because of market participants’ emotions. Investor fear and panic selling plunges stock prices downward below equity intrinsic values at market bottoms.

#732 –Predicting Performance Using Consumer Big Data

Period of rebalancing: Daily
Markets traded: equities
Instruments used for trading: stocks
Complexity: Simple strategy
Backtest period: 2009-2020
Indicative performance: 9.29%
Estimated volatility: 14.29%

Source paper:

Froot, Kenneth and Kang, Nambo and Ozik, Gideon and Sadka, Ronnie: Predicting Performance Using Consumer Big Data
https://scholar.harvard.edu/files/kenfroot/files/Predicting_Performance_Using_Consumer_Big_Data-Aug18.2021.pdf?fbclid=IwAR3PCsYpCD4Uc5oyfpBPONFR30jOXVzCzWT5M5c9nX3UN1RJlzBcIq0T7rA
Abstract:
To predict firms’ fundamentals, the authors construct three proxies for real-time corporate sales from fully distinct information sources: In-store foot traffic (IN-STORE), web traffic to companies’ websites (WEB), and consumers’ interest level in corporate brands and products (BRAND). The authors demonstrate that trading using these proxies, estimated for a sample of 330 firms over 2009–2020, result in significant net-of-transaction-costs profitability. During the pandemic, WEB activity increases significantly while there is remarkable decrease in IN-STORE, reflecting the migration of consumers from physical stores toward online. The results suggest that the information contained in IN-STORE and BRAND is not immediately available to investors, while the WEB information is diffused more quickly, and that overall information diffusion worsened during the pandemic.

New research papers related to existing strategies:

#71 – Short Term Reversal with Futures
#21 – Momentum Effect in Commodities

Zhi Da, Ke Tang, Yubo Tao and Liyan Yang: Financialization and Commodity Markets Serial Dependence
https://www3.nd.edu/~zda/commodity.pdf
Abstract:
Recent financialization in commodity markets makes it easier for institutional investors to trade a portfolio of commodities via various commodity-indexed products. We present several pieces of novel causal evidence that daily exposure to such index trading results in price overshoots and reversals, as reflected in a negative daily return autocorrelations, only among commodities in that index. This is because index trading propagates non-fundamental noises to all indexed commodities. We present direct evidence for such noise propagation using commodity news sentiment data.

#12 – Pairs Trading with Stocks

Wei-Lun Kuo, Tian-Shyr Dai and Wei-Che Chang: Solving Unconverged Learning of Pairs Trading Strategies with Representation Labeling Mechanism
http://ceur-ws.org/Vol-3052/paper4.pdf
Abstract:
A pairs trading strategy (PTS) constructs a market-neutral portfolio whose value typically moves back and forth around a mean price level; investors short (long) the portfolio when its value reaches the upside (downside) opening threshold and close the position when the value reverts to the mean to earn the price difference. Recent machine learning models select the open and stop-loss thresholds either heuristically or chosen from a limited set, which significantly limits the investment performance. We address this by creating a wider set of open/stop-loss threshold recommendations that generally cover all possible scenarios; but regression- or classification-based deep learning methods for recommending thresholds fail to converge. Thus, we design a representative labeling mechanism that selects representative open and stop-loss thresholds from all possible optimal thresholds according to the selection frequencies of the thresholds and the 𝑘-means algorithm. Experiments suggest that training the multi-scale residual network with stock pairs relabeled by representative thresholds yields better investment performance than other methods in the literature.

#376 – Combining Fundamental FSCORE and Equity Short-Term Reversals

Salas Najera, Carlos: The Evolution of Fundamental Scoring Models and Machine Learning Implications
https://ssrn.com/abstract=3945477
Abstract:
“Man+Machine” is a series of articles where the reader can find guidance on how to bridge the gap between fundamental analysis knowledge and new data science/ML/AI methods. This first article will provide an overview of the historical evolution of systematic fundamental scoring models, and an introductory analysis of how Machine Learning (ML) is transforming and enhancing these traditional indicators. This article is designed to briefly introduce readers without prior knowledge of fundamental scoring models.

#582 – Carbon Risk in the Cross Section of Corporate Bond Returns

Liu, Lu: A Study on the Effectiveness of China Bond Market ESG Investment Practice
https://ssrn.com/abstract=4008732
Abstract:
At present, the concept of ESG investment has attracted much attention among market entities. Based on the bond valuation and ChinaBond ESG scores of China’s credit bond issuers, this paper calculates and analyzes the premium risk, excess return and risks of ESG investment in the bond market, so as to test the effectiveness of ESG investment in China’s bond market. The results show that ESG risk premium exists in the bond market; bonds issued by issuers with higher ESG scores show a higher return on and lower risk to investment, namely producing higher ESG excess return and lower excess risk. The conclusion is still sound based on the control of credit rating. This paper not only provides empirical evidence for the bond market to approve ChinaBond ESG evaluation, but also provides useful reference for in-depth dissemination of the ESG investment philosophy.

#685 – Boosted Trees and Cryptocurrency Return Prediction

Attanasio, Giuseppe, Cagliero, Luca, Garza, Paolo and Baralis, Elena: Quantitative cryptocurrency trading: exploring the use of machine learning techniques
https://iris.polito.it/retrieve/handle/11583/2749758/516302
Abstract:
Machine learning techniques have found application in the study and development of quantitative trading systems. These systems usually exploit supervised models trained on historical data in order to automatically generate buy/sell signals on the financial markets. Although in this context a deep exploration of the Stock, Forex, and Future exchange markets has already been made, a more limited effort has been devoted to the application of machine learning techniques to the emerging cryptocurrency exchange market. This paper explores the potential of the most established classification and time series forecasting models in cryptocurrency trading by backtesting model performance over a eight year period. The results show that, due to the heterogeneity and volatility of the underlying financial instruments, prediction models based on series forecasting perform better than classification techniques. Furthermore, trading multiple cryptocurrencies at the same time significantly increases the overall returns compared to baseline strategies exclusively based on Bitcoin trading.

#536 – Machine Learning Stock Picking

Hambly, Ben M. and Xu, Renyuan and Yang, Huining: Recent Advances in Reinforcement Learning in Finance
https://ssrn.com/abstract=3971071
Abstract:
The rapid changes in the finance industry due to the increasing amount of data has revolutionized the techniques on data processing and data analysis and brought new theoretical and computational challenges. In contrast to classical stochastic control theory and other analytical approaches for solving financial decision-making problems that heavily reply on model assumptions, new developments from reinforcement learning (RL) are able to make full use of the large amount of financial data with fewer model assumptions and to improve decisions in complex financial en- environments. This survey paper aims to review the recent developments and use of RL approaches in finance. We give an introduction to Markov decision processes, which is the setting for many of the commonly used RL approaches. Various algorithms are then introduced with a focus on value and policy based methods that do not require any model assumptions. Connections are made with neural networks to extend the framework to encompass deep RL algorithms. Our survey concludes by discussing the application of these RL algorithms in a variety of decision-making problems in finance, including optimal execution, portfolio optimization, option pricing and hedging, market making, smart order routing, and robo-advising.

#536 – Machine Learning Stock Picking

Kelly, Bryan T. and Malamud, Semyon and Zhou, Kangying: The Virtue of Complexity in Machine Learning Portfolios
https://ssrn.com/abstract=3984925
Abstract:
We theoretically characterize the behavior of machine learning portfolios in the high complexity regime, i.e. when the number of parameters exceeds the number of observations. We demonstrate a surprising “virtue of complexity:” Sharpe ratios of machine learning portfolios generally increase with model parameterization, even with minimal regularization. Empirically, we document the virtue of complexity in US equity market timing strategies. High complexity models deliver economically large and statistically significant out-of-sample portfolio gains relative to simpler models, due in large part to their remarkable ability to predict recessions.

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

Factor Performance in Cold War Crises – A Lesson for Russia-Ukraine Conflict

The Russia-Ukraine war is a conflict that has not been in Europe since WW2. And it has great implications not only on human lives but also on security prices. It bears numerous characteristics of the cold war crises, where two nuclear powers (Soviet Union and USA/NATO) were often very close to hot war or were waging a proxy war in 3rd countries. We thought it might be wise to look at similar periods from the past to understand what happens in such situations. We selected five events and analyzed the performance of main equity factors (market, HML, SMB, momentum & 2x reversal) and energy and fixed income proxy portfolios.

Full vs. Synthetic Replication and Tracking Errors in ETFs

The growth of passive investing and ETFs is indisputable. Consequently, this boom also affects financial markets (e.g., market elasticity or by creating predictable buys and sells) and assets that ETFs track. Even though all passive ETFs aim to replicate some benchmark index, there are two distinct approaches to doing so. The first approach is directly replicating the benchmark (by buying underlying assets) either by full direct replication or sampling. The second approach consists of synthetic replication using derivatives – most commonly by total return swaps (or futures). How do replication methods influence tracking error?