New strategies:
#963 – Why Do US Stocks Outperform EM and EAFE Regions?
Period of rebalancing: Monthly
Markets traded: equities
Instruments used for trading: ETFs, funds
Complexity: Simple strategy
Backtest period: 1994-2023
Indicative performance: 7.14%
Estimated volatility: 19.5%
Source paper:
Dujava, Cyril and Vojtko, Radovan: Why Do US Stocks Outperform EM and EAFE Regions?
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4694624
Abstract:
The article “Why Do US Stocks Outperform EM and EAFE Regions?” from Quantpedia explores the reasons behind the historical outperformance of US stocks compared to emerging markets (EM) and developed markets (DM) outside of the United States. The article highlights several factors that contribute to this phenomenon, including: Productivity: The US is home to many highly productive companies with solid returns on capital. These companies generate higher profits and, consequently, higher returns for investors.
Market Size and Liquidity: The US stock market is the most significant and liquid in the world, with a wide variety of stocks. This liquidity makes it easier for investors to buy and sell stocks, which can lead to more efficient pricing and better returns.
Political and Regulatory Stability: The US has a stable political and regulatory environment, providing investors with confidence and reducing risk. This stability can attract more investment to the US market, further boosting returns.
Currency Strength: The US dollar is the world’s reserve currency, giving it stability and attractiveness to investors. This can make US stocks more attractive to international investors, contributing to their outperformance.
Market Cycle: The performance of US stocks tends to follow a cyclical pattern, with periods of outperformance followed by periods of underperformance. Various factors influence these cycles, including economic conditions, interest rates, and investor sentiment.
#964 – Harvesting Volatility Risk Premia and Crisis Alpha via ETFs
Period of rebalancing: Daily
Markets traded: bonds, commodities, equities
Instruments used for trading: ETFs
Complexity: Complex strategy
Backtest period: 2008-2023
Indicative performance: 22.2%
Estimated volatility: 24.13%
Source paper:
Sadik, Sheikh: A Tactical Strategy using ETFs: Harvesting Volatility Risk Premia & Crisis Alpha
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4666899
Abstract:
I discuss a systematic approach to investing in volatility risk premia through ETFs. The reason behind ETFs is mostly due to my personal interest in volatility investing and being able to manage the risk of the strategy on my own without any capital constraint. However, I will be using futures contracts mostly to backfill the data for my backtest but the main idea is focused around futures contracts whose exposure can be obtained through ETFs. I construct a primary model for harvesting volatility risk premia and next overlay a meta model for risk management using ridge regression. I also incorporate a CTA program using selected commodity ETFs to harvest crisis alpha.
#965 – Exploration of Long Short ETF Momentum Strategies
Period of rebalancing: Monthly
Markets traded: bonds, commodities, currencies, equities, REITs
Instruments used for trading: ETFs
Complexity: Simple strategy
Backtest period: 2006-2023
Indicative performance: 8.08%
Estimated volatility: 10.54%
Source paper:
Vojtko & Pauchlyová: Exploration of Long Short ETF Momentum Strategies
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4710334
Abstract:
CTA funds, recognized for their effectiveness as hedges during crises, are traditionally diversified and involve going long and short on various asset classes. This paper aims to explore trend-following strategies by creating a CTA proxy using ETFs to represent CTA strategies. This allows for a comprehensive assessment of individual substrategies across diverse assets. We identify that the optimal approach involves following trends in all asset classes, except for stocks. Here, adopting a long-only position yields the most favorable results.
#966 – Investment Carry
Period of rebalancing: Weekly
Markets traded: currencies
Instruments used for trading: CFDs, forwards, futures
Complexity: Complex strategy
Backtest period: 1993-2023
Indicative performance: 6.42%
Estimated volatility: 9.4%
Source paper:
Hertrich, Daniel: Beyond the Forward Premium: the Reward-risk Trade-off of Taking the Long Position Abroad or Domestically
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4577434
Abstract:
Based on the ratio between expected currency excess returns and the conditional value-at-risk of spot returns, we construct the carry classifier, an indicator that distinguishes between two foreign exchange market regimes. In funding carry regimes, investment currencies are unable to adequately compensate for their negative exposure to global FX volatility, preventing the carry risk factor from explaining the cross-section of G10 currency excess returns. However, in funding carry regimes, we find a significant reversal in the future excess returns of ‘high-risk loser’ and ‘low-risk winner’ currencies, consistent with a future correction of investor’s past overreaction to idiosyncratic excess returns.
#967 – Shortable Stocks Factor in Emerging Market Equities
Period of rebalancing: Monthly
Markets traded: equities
Instruments used for trading: stocks
Complexity: Very complex strategy
Backtest period: 1994-2023
Indicative performance: 15.73%
Estimated volatility: 17.46%
Source paper:
Li, Xiaoming: Manacled Short Sellers and Return Premium: New Evidence
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4201779
Abstract:
Investigating the short-selling regulation of the Honk Kong market, we document that shortable stocks, on average, earn significantly higher returns than non-shortable stocks. However, loadings of stocks/portfolios on the shortable minus non-shortable misvaluation factor SMN predict a significant negative return premium in the cross section of returns. We measure SMN by applying both value- and return-weighted methods with various time lags. We propose a behavioural model to rationalize our results. The model shows that, if investors are overconfident regarding short-selling regulatory factor signals, it is possible to detect a positive average/abnormal return but a negative future return premium on SMN.
#968 – Night Minus Day Return of Factor Portfolios
Period of rebalancing: Intraday
Markets traded: equities
Instruments used for trading: stocks
Complexity: Very complex strategy
Backtest period: 1996-2020
Indicative performance: 94.73%
Estimated volatility: 21.72%
Source paper:
Lu, Zhongjin and Qin, Zhongling: A One-factor Model for Expected Night-minus-day Stock Returns
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3854046
Abstract:
We propose a one-factor model for expected night-minus-day (NMD) returns and use it to examine the economic forces underlying NMD return predictabilities. Our model successfully prices a large set of NMD portfolios with a cross-sectional R2 around 80%. Consistent with the absence of near-arbitrage opportunities, the pricing factor has substantial exposure to the dominant common risks in the NMD return space. Finally, we link the pricing factor to retail order imbalances at the market open and the required returns from liquidity provision. Our findings point to a pricing equilibrium in which liquidity providers require compensation for accommodating sentiment-driven demand.
#969 – Pragmatic Asset Allocation Model for Semi-Active Investors
Period of rebalancing: Quarterly
Markets traded: bonds, commodities, equities
Instruments used for trading: ETFs, funds, futures
Complexity: Simple strategy
Backtest period: 1927-2023
Indicative performance: 10.73%
Estimated volatility: 11.38%
Source paper:
Vojtko, Radovan and Javorská, Juliána: Pragmatic Asset Allocation Model for Semi-Active Investors
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4715415
Abstract:
This paper introduces the Pragmatic Asset Allocation strategy, a novel approach to Global Tactical Asset Allocation (GTAA), offering investors a practical balance between market participation and simplified portfolio management through quarterly rebalancing. The strategy, implemented in six steps, incorporates Momentum, Trend, Cash, Yield Curve, Hedging Portfolio, and Stop Loss mechanisms. Backtested from December 1927 to August 2023, the strategy yields an annual return of 10.73% and a Sharpe ratio 0.93. Key features of the Pragmatic Asset Allocation strategy, such as tranching and tax optimization rules, make it a straightforward and accessible option for investors seeking global diversification and effective risk management with minimal hands-on involvement.
New research papers related to the existing strategies:
#871 – Traditional Carry in Cryptocurrencies
Fan, Zhenzhen and Jiao, Feng and Lu, Lei and Tong, Xin: The Risk and Return of Cryptocurrency Carry Trade
https://ssrn.com/abstract=4666425
Abstract:
This paper provides a comprehensive analysis of the risk and return dynamics associated with cryptocurrency carry trade. The cross-sectional carry trade strategy using cryptocurrencies yields an annualized return of 49.3% with a Sharpe ratio of 0.81. Our investigation shows that the cryptocurrency returns cannot be explained by prevailing crytocurrency factors, including market, size, momentum, volatility, downside-risk, and platform collapse risks. Moreover, our analysis fails to identify any substantial connection between these returns and geopolitical risks or fiat-currency carry trade. Our findings suggest that a significant portion of cryptocurrency carry trade returns can be regarded as a premium for equity market volatility risk. Our research highlights the inter-asset class linkages between equity risk factors and cryptocurrency returns.
#381 – Blended Factors in Cryptocurrencies
Borri, Nicola and Massacci, Daniele and Rubin, Mirco and Ruzzi, Dario: Crypto Risk Premia
https://ssrn.com/abstract=4154627
Abstract:
This paper studies risk premia in a large cross-section of cryptocurrency. We characterize the stochastic discount factor in terms of latent factors and obtain risk premia estimates for a large set of observable factors that are robust to omitted variable and measurement error. These are particularly relevant issues in the study of this novel asset class. We show that several price-based crypto and non-crypto factors are priced in the cross-section of crypto returns. Most importantly, we find that macroeconomic risk is priced in cryptocurrency: coins with lower payoffs at times of higher geopolitical risk or negative macro shocks are riskier and offer higher returns. These findings uncover clear linkages to other asset classes and to macroeconomic conditions, thus providing a broad perspective on crypto risk premia.
#421 – Earnings Response Elasticity
Kottimukkalur, Badrinath and Nallareddy, Suresh and Venkatachalam, Mohan: The Changing Information Content of Aggregate Earnings
https://ssrn.com/abstract=4068820
Abstract:
Motivated by the structural shift in the relation between inflation and the output gap, we investigate whether the information content of aggregate earnings for inflation and output gap changes over time. We find that, during 1970–2001, aggregate earnings are positively associated with future inflation but not with the output gap. In recent periods, however, higher aggregate earnings are associated with a higher output gap but are unrelated to inflation. Given the changing information content of aggregate earnings for macroeconomic outcomes, we posit and find that the association of aggregate earnings with risk-free rates, bond yields, risk premiums, and aggregate stock returns also differs systematically across time. In sum, our findings suggest that the informational content of aggregate earnings is conditional on the inflation–output gap relationship.
#670 – Machine Learning Pairs Trading Strategy
Figueira, Miguel and Horta, Nuno: Machine Learning-Based Pairs Trading Strategy with Multivariate
https://ssrn.com/abstract=4295303
Abstract:
Pairs trading is one of the most popular arbitrage investment strategies. By monitoring a pair of two assets that closely follow each other, the trader acts when the pair presents an imbalance, profiting when the stocks converge to their equilibrium. Despite the rising popularity of Machine Learning in financial applications, most pairs trading strategies are still based in exhaustive search methods and rigid trading heuristics. This work creates pairs by addressing conflicting objectives, maximizing profit and minimizing risk, and explores multivariate pairs, composed of more than 2 stocks. Two Elitist genetic algorithms, NSGA II and III, are used to create pairs, achieving returns of up to 9,7% p.a., and proving to be robust to market crashes. Furthermore, multivariate pairs are found to be inferior to the traditional two-stock pairs. Additionally, a forecasting-based trading strategy is developed, attempting to improve the standard trading technique. An ARIMA and XGBoost models are used to forecast the spread in a trend-based trading strategy. The forecasting strategy yields returns of 23% p.a. and beats the market during 2018 and 2019. The system is tested in the Real Estate sector of the S&P500, from 2018 to December 2021.
#685 – Boosted Trees and Cryptocurrency Return Prediction
Oyedele, Azeez A. and Ajayi, Anuoluwapo and Oyedele, Lukumon and Bello, Sururah A. and Jimoh, Kudirat O.: Performance Comparison of Deep Learning and Boosted Trees for Cryptocurrency Closing Price Prediction
https://ssrn.com/abstract=4094652
Abstract:
The emergence of cryptocurrencies has drawn significant investment capital in recent years with an exponential increase in market capitalization and trade volume. However, the cryptocurrency market is highly volatile and burdened with substantial heterogeneous datasets characterized by complex interactions between predictors, which may be difficult for conventional techniques to achieve optimal results. In addition, volatility significantly impacts investment decisions; thus, investors are confronted with how to determine the price and assess their financial investment risks reasonably. This study investigates the performance evaluation of a genetic algorithm tuned Deep Learning (DL) and boosted tree-based techniques to predict several cryptocurrencies’ closing prices. The DL models include Convolutional Neural Networks (CNN), Deep Forward Neural Networks, and Gated Recurrent Units. We assess and benchmark the performance of the DL models with boosted tree-based models on six cryptocurrency datasets from multiple data sources using relevant performance metrics. The results reveal that the CNN model has the least mean average percentage error of 0.06 and produces a consistent and highest explained variance score of 0.97 (on average) compared to other models. Hence, CNN is more reliable with limited training data and easily generalizable for predicting several cryptocurrencies’ daily closing prices. Also, the results will help practitioners obtain a better understanding of crypto market challenges and offer practical strategies to lower risks.
And several interesting free blog posts that have been published during the last 2 weeks:
Are Cryptocurrencies Exposed to Traditional Factor Risks?
Cryptocurrencies are attracting much attention, even becoming a priority for many high-net-worth investors. The introduction of the new spot Bitcoin ETFs simplifies access to this asset class, and as cryptos are included in more and more portfolios, industry practitioners look for models that can help assess how big a portion of clients’ portfolios allocate to this new asset class. Factor risk models are an industry standard for understanding other main asset classes, and authors of today’s presented research (Akbari, Ekponon, and Guo, revised 2024) provide useful insights into which factor risks can explain the variation in cryptos returns.
The main take-away? We can definitely shred the idea that crypto stands on its own, acting independently and in isolation from other financial world vehicles. Overall, these findings provide the evidence that well-known factor risks can explain crypto market returns and that a strong link exists between the crypto market and traditional asset classes.
Join the Race: Quantpedia Awards 2024 Await You
Two weeks ago, we promised you a surprise, and now it’s finally time to unveil what we have prepared for you :).
Our Quantpedia Awards 2024 aims to be the premier competition for all quantitative trading researchers. If you have an idea in your head about systematic/quantitative trading or investment strategy, and you would like to gain visibility on the professional scene, then submit your research paper, and you can compete for an attractive list of prizes. All info about the prizes, submission process, expert committee, and our partners are described in detail on our dedicated subpage: Quantpedia Awards 2024. But we will also give you a quick overview in this blog post.
Improving FX Carry Strategy with Exotic Currencies and the Frontier Markets
Forex markets lure retail traders into a game of “hunting pips” with high leverage and high turnover scalping strategies, in which small traders often lose more than they can afford. But there are other ways of trading currencies. The smart money knows how to exploit interest rate spreads that this asset class offers by employing the FX Carry Trade strategy. In the past decade, the low interest rates of the most developed countries made the FX Carry strategy less profitable, but as inflation returned, higher interest rates returned in some countries, too, and with them, the interest rate spreads widened. And FX Carry is back, and the question stands: Can we improve this well-known trading style?
How to Build a Systematic Innovation Factor in Stocks
The aim of this article is multifold. It aims to answer the research question: does a portfolio consisting of top innovators outperform the S&P 500 index? To address this question, a strategy of investing long in top innovators according to their ranking is developed, and its performance is compared to that of the broad-based index. Based on the common belief that higher innovativeness carries higher risk, it aims to evaluate the volatility associated with innovative stocks. Additionally, it aims to analyze the impact of sector factors on the portfolio’s performance. Finally, it conducts a comparative analysis between the portfolio’s performance and that of the ARK Innovation ETF (ARKK), which specifically focuses on investing in companies relevant to the theme of disruptive innovation.
Plus, the following trading strategies have been backtested in QuantConnect in the previous two weeks:
485 – Toxical Releases and Stock´s Performance
959 – Volume Weighted Average Price (VWAP) as Precise Trend-Following Indicator for Day-Traders
963 – Why Do US Stocks Outperform EM and EAFE Regions?
969 – Pragmatic Asset Allocation Model for Semi-Active Investors
