Big Ideas

What are the best ways to apply backtesting in the development of investment strategies? In their July 2024 paper entitled “The Three Types of Backtests”, Jacques Joubert, Dragan Sestovic, Illya Barziy, Walter Distaso and Marcos Lopez de Prado offer guidance on backtesting best practices by reviewing:

• Strengths and weaknesses of the three main types of backtests (walk-forward, resampling and Monte Carlo simulation).
• Ways to enhance the quality of backtests.
• Mitigation of the confounding effects of running multiple tests on the same data.

Based on theoretical and practical considerations, they conclude that:

Keep Reading

Do widely used associational (rather than causal) methods used by researchers to specify factor models of asset returns work? In their March 2024 paper entitled “The Case for Causal Factor Investing”, Marcos Lopez de Prado, Alex Lipton and Vincent Zoonekynd describe the shortcomings of associational methods of factor model development. They address p-hacking (data snooping), with focus on interferences from variables called colliders (causally influenced by two or more variables) and confounders (influencing both dependent and independent variables). They further describe what can be done to correct these shortcomings. Based on logical/mathematical analysis and the body of financial markets research, they conclude that:

Keep Reading

Technological disruption (as experienced with widespread electrification and the rise of the world-wide web, and imagined for artificial intelligence) is a recurring feature of human history. Such disruptions presents risks and opportunities for investors. How can investors manage such risk? In their February 2024 paper entitled “Technological Disruption and Long-Term Investors: Managing Risk and Opportunities”, Alistair Barker, Ashby Monk and Dane Rook describe approaches to managing investment risks from technological disruptions of varying scales and velocities. Using outputs of interviews with 20 elite long-term investors worldwide, they find that:

Keep Reading

Some exchange-traded funds (ETF) focus on capturing potentially attractive factor premiums or thematic niches. Their histories offer a way to test these concepts live. We have conducted many such tests, listed here to offer a global view.

1. “U.S. Equity Premium?” – evidence from simple tests on about 21 years of data suggests that stock market leadership shifts between the U.S. and other developed markets over time, but the U.S. may be better overall.
2. “Tech Equity Premium?” – evidence from simple tests on 24 years of data suggests long boom, short bust for a tech/innovation-concentrated portfolio. It does not support belief in risk-adjusted outperformance.
3. “Measuring the Size Effect with Capitalization-based ETFs” – evidence from simple tests of capitalization-based ETFs with nearly 22 years of data offers little support for belief in a long-term, reliably exploitable size effect among U.S. stocks.
4. “Do Equal Weight ETFs Beat Cap Weight Counterparts?” – evidence from simple tests on some equal-weight U.S. equity ETFs offers little support for belief that equal weighting substantially and reliably beats capitalization weighting on a net basis.
5. “Measuring the Value Premium with Value and Growth ETFs” – evidence from simple tests with 21.6 years of available data does not support belief that investors reliably capture a value premium via popular value-growth ETFs.
6. “Are Equity Momentum ETFs Working?” – available evidence on attractiveness of momentum-oriented U.S. stock and sector ETFs is less than compelling.
7. “Are Stock Quality ETFs Working?” – available evidence offers little support for belief that quality ETFs reliably beat respective benchmarks.
8. “Are Low Volatility Stock ETFs Working?” – available evidence on attractiveness of low volatility stock ETFs is mixed, with recent data undermining belief in reliability of low volatility outperformance.
9. “Are Equity Multifactor ETFs Working?” – available evidence offers very little support for belief that equity multifactor ETFs beat their benchmarks, or that they offer material diversification with comparable performance.
10. “Are Hedge Fund ETFs Working?” – evidence on attractiveness of hedge fund-oriented ETFs is mostly negative.
11. “Are Managed Futures ETFs Working?” – available evidence on attractiveness of managed futures ETFs in aggregate (but with recent short-sample exceptions) suggests that any benefits from diversification of equities and fixed income are unlikely to compensate for poor absolute returns.
12. “Best Safe Haven ETF?” – evidence from simple tests over available and common sample periods suggests that silver, gold, longer-term U.S. Treasuries and investment grade corporate bonds are safe havens, while crude oil is clearly not.
13. “Do High-dividend Stock ETFs Beat the Market?” – evidence from data for high-dividend U.S. stock ETFs does not support belief that high-dividend stocks reliably outperform the broad U.S. stock market.
14. “Are ESG ETFs Attractive?” – available evidence suggests that ESG ETFs do not perform much differently from selected benchmarks.
15. “How Are Renewable Energy ETFs Doing?” – available evidence on attractiveness of renewable energy ETFs is adverse overall, but with bursts of market outperformance perhaps due to novelty.
16. “How Are Robotics-AI ETFs Doing?” – available evidence is that robotics-AI ETFs are less attractive than the broader technology exposure offered by QQQ.
17. “How Are AI-powered ETFs Doing?” – available evidence does not support belief that ETFs using AI to select and weight assets are particularly attractive.
18. “Are iShares Core Allocation ETFs Attractive?” – available evidence regarding attractiveness of iShares Core Asset Allocation ETFs is mixed to negative.
19. “Are Target Retirement Date Funds Attractive?” – evidence offers little support for belief that target retirement date mutual funds are preferable to simple stocks-bonds diversification.
20. “How Are TIPS ETFs Doing?” – available evidence on attractiveness of TIPS ETFs is mostly favorable after the recent inflation burst, with shorter duration funds offering more reliable inflation protection.
21. “Are Equity Index Covered Call ETFs Working?” – available evidence on attractiveness of equity index covered call ETFs as either substitutes for or diversifiers of underlying stock indexes is generally adverse.
22. “Are Equity Put-Write ETFs Working?” – available evidence on attractiveness of equity put-write ETFs is adverse.
23. “Are IPO ETFs Working?” – available evidence on attractiveness of IPO ETFs is mixed, requiring very high risk tolerance of interested investors.
24. “Are Preferred Stock ETFs Working?” – available evidence on attractiveness of preferred stock ETFs relative to a 60-40 stocks-bonds portfolio is largely negative.
25. “Do Convertible Bond ETFs Attractively Meld Stocks and Bonds?” – available evidence suggests that convertible bond ETFs sometimes outperform and sometimes underperform a conventional 60-40 stocks-bonds portfolio.
26. “Do ETFs Following Gurus/Insiders Work?” – available evidence on attractiveness of guru/insider-following stock ETFs is mostly adverse.
27. “Congressional Trade Tracking ETFs” – limited available evidence suggests that investors should choose a fund mimicking holdings of Democrat rather than Republican members of Congress.
28. “The Long and Short of Jim” – available evidence does not support belief that funds based on Jim Cramer’s stock/market recommendations reliably produce attractive short-term returns.
29. “Live Test of the Stock Market Overnight Move Effect” – early evidence does not support belief in exploitability of the overnight move effect.

The upshot of the above items is that academic factor research and thematic speculations rarely translate to outperformance when implemented with ETFs.

A global caution is that the period since 2009 is strong for broad equity indexes, driven by a few large-capitalization firms. This trend may not persist.

Do linear factor model specification choices inherently produce out-of-sample underperformance of investment strategies seeking to exploit factor premiums? In their January 2024 paper entitled “Why Has Factor Investing Failed?: The Role of Specification Errors”, Marcos Lopez de Prado and Vincent Zoonekynd examine whether standard practices induce factor specification errors and how such errors might explain actual underperformance of popular factor investing strategies. They consider potential effects of confounding variables and colliding variables on factor model out-of-sample performance. Based on logical derivations, they conclude that: Keep Reading

What is the state of machine learning in finance? In their July 2023 paper entitled “Financial Machine Learning”, Bryan Kelly and Dacheng Xiu survey studies on the use of machine learning in finance to further its reputation as an indispensable tool for understanding financial markets. They focus on the use of machine learning for statistical forecasting, covering regularization methods that mitigate overfitting and efficient algorithms for screening a vast number of potential model specifications. They emphasize areas that have received the most attention to date, including return prediction, factor models of risk and return, stochastic discount factors and portfolio choice. Based on the body of machine learning research in finance, they conclude that: Keep Reading

What happens as more and more web-scraped training data for Large Language Models (LLM), such as ChatGPT, derives from outputs of predecessor LLMs? In their May 2023 paper entitled “The Curse of Recursion: Training on Generated Data Makes Models Forget”, Ilia Shumailov, Zakhar Shumaylov, Yiren Zhao, Yarin Gal, Nicolas Papernot and Ross Anderson investigate changes in LLM outputs as training data becomes increasingly LLM-generated. Based on simulations of this potential trend, they find that: Keep Reading

In his 2023 book, The Uncertainty Solution: How to Invest with Confidence in the Face of the Unknown, author John Jennings seeks “to provide individual investors with mental models that will help them make better investment decisions, practice better investment behavior, and be better consumers of investment advice… This book is not about how to invest but rather how to think about investing. It is the culmination of my thirteen-year quest for investment wisdom… The mental models in this book describe the investment world as full of uncertainty, wild randomness, unpredictability, and pitfalls. There’s no easy path. But mental models that embrace reality—that take the world as it is, not how we think it is or want it to be—will make you a better investor and a better consumer of investment advice.” Based on his many years of wealth management experience, especially during the 2007-2008 Financial Crisis, he concludes that:

Keep Reading

For the last half-century, financial researchers have fallen short of scientific rigor, focusing on associations not supported by theory and not amenable to falsification. Is there is hope for finance to become a scientific field? In his April 2023 paper entitled “The Hierarchy of Empirical Evidence in Finance”, Marcos Lopez de Prado proposes a hierarchy of empirical evidence that gives greatest scientific weight to methods allowing falsification of causal claims. He addresses:

• Why associations alone do not constitute scientific knowledge.
• The importance of causality and how statistical methods enable the falsification of causal claims.
• How to extract causal effects from observational studies in fields like finance with inherent barriers to controlled experimentation.
• An example of examining causality in finance.

He ultimately translates the distinction between associational and causal claims into a hierarchy of empirical evidence in finance. Based on the philosophy of science and the constraints of studying complex financial systems, he concludes that: Keep Reading

Which industries are most exposed to disruption by artificial intelligence (AI) language models such as ChatGPT? In the April 2023 version of their paper entitled “How will Language Modelers like ChatGPT Affect Occupations and Industries?”, Edward Felten, Manav Raj and Robert Seamans focus the previously developed AI Occupational Exposure (AIOE) measure on models such as by ChatGPT, relating model capabilities to 52 human abilities and thereby to human occupations and industries. Applying this adaption, they find that:

Keep Reading