Big Ideas

How does a large sample of stock return anomalies fare in recent replication testing? In their October 2018 paper entitled “Replicating Anomalies”, Kewei Hou, Chen Xue and Lu Zhang attempt to replicate 452 published U.S. stock return anomalies, including 57, 69, 38, 79, 103, and 106 anomalies 57 momentum, 69 value-growth, 38 investment, 79 profitability, 103 intangibles and 106 trading frictions (trading volume, liquidity, market microstructure) anomalies. Compared to the original papers, they use the same sample populations, original (as early as January 1967) and extended (through 2016) sample periods and similar methods/variable definitions. They test limiting influence of microcaps (stocks in the lowest 20% of market capitalizations) by using NYSE (not NYSE-Amex-NASDAQ) size breakpoints and value-weighted returns. They consider an anomaly replication successful if average high-minus-low tenth (decile) return is significant at the 5% level, translating to t-statistic at least 1.96 for pure standalone tests and at least 2.78 assuming multiple testing (accounting for aggregate data snooping bias). Using required anomaly data and monthly returns for U.S. non-financial stocks during January 1967 through December 2016, they find that:

Keep Reading

What do basic U.S. stock market return statistics say about consistency of equity risks and predictability of returns? We define basic statistics as first through fourth moments of the return distribution: mean (average), standard deviation, skewness and kurtosis. For tractability, we calculate these four statistics month-by-month based on daily returns. Using daily closes of the Dow Jones Industrial Average (DJIA) since January 1930 and the S&P 500 Index since January 1950, both through September 2018, we find that: Keep Reading

Patrick Donohoe introduces his 2018 book, Heads I Win, Tails You Lose: A Financial Strategy to Reignite the American Dream, by stating that the book: “…will teach you many of the principles and strategies to help discover your own path to financial freedom. Most importantly, it will show you the mindset required to carry out a successful plan. …almost everything you will gain from this book conflicts with what the typical financial planner, financial celebrity, and most financial publications tell you to do. …You will…discover how to pivot the foundation of your wealth to…the private mutual insurance company.” Based on his experience, market research and many examples, he concludes that: Keep Reading

Brian Portnoy introduces his 2018 book, The Geometry of Wealth: How To Shape A Life Of Money And Meaning, by stating that the book is: “…a story told in three parts,…from purpose to priorities to tactics. Each step has a primary action associated with it. The first is adaptation. The second is prioritization. The third is simplification. …The principle that motors us along the entire way is what I call ‘adaptive simplicity,’ a means of both rolling with the punches and and cutting through the noise.” Based on his two decades of experience in the mutual fund and hedge fund industries, including interactions with many investors, along with considerable cited research (much of it behavioral), he concludes that: Keep Reading

In what areas does machine learning have advantages over conventional financial/investment analysis? In his June 2018 presentation entitled “Nine Financial Applications of Machine Learning”, Marcos Lopez de Prado summarizes investing-related areas in which well-supervised machine learning outperforms conventional methods. Based on relevant research and his experience, he asserts that: Keep Reading

Data snooping bias is pervasive in published research and quantitative investment strategies. Should investors resign themselves to the consequence that investment managers/funds offer products picked mostly on past luck? In his May 2018 presentation package entitled “How the Sharpe Ratio Died, and Came Back to Life”, Marcos Lopez de Prado introduces an approach to Sharpe ratio estimation via backtesting that would enable academia, regulators and investors to distinguish between strategies that probably work and those that probably do not. Based on the evolution of Sharpe ratio estimation approaches, he concludes that: Keep Reading

Is there a tractable way of estimating the level of data snooping bias in investment strategy studies and thereby correcting for it? In their April 2018 paper entitled “Detection of False Investment Strategies Using Unsupervised Learning Methods”, Marcos Lopez de Prado and Michael Lewis summarize and validate an approach for estimating snooping bias derived from backtesting multiple strategies on the same data and using that estimate to correct for the bias. The approach involves estimating the overall scope and dispersion of multiple backtests based on correlation clusters within known backtests. Focusing on Sharpe ratio as the key performance metric, they validate their approach via Monte Carlo simulations. Based on derivations and simulations, they conclude that: Keep Reading

Is use of a sampling interval much shorter than input variable measurement interval a useful statistical practice in financial markets research? In the April 2018 update of their paper entitled “Long Horizon Predictability: A Cautionary Tale”, flagged by a subscriber, Jacob Boudoukh, Ronen Israel and Matthew Richardson examine statistical reliability gains from overlapping measurements of long-horizon variables (such as daily or monthly sampling of 5-year returns or 10-year moving average earnings). They employ the widely used cyclically adjusted price earnings ratio (CAPE, or P/E10) for some examples. Based on illustrations and mathematical derivations, they conclude that: Keep Reading

What is the best way to put strategic edges and tactical views into investment portfolios? In their March 2018 paper entitled “Model Portfolios”, Debarshi Basu, Michael Gates, Vishal Karir and Andrew Ang describe and illustrate a three-step optimized asset allocation process incorporating investor preferences and beliefs that is rigorous, repeatable, transparent and scalable. The three steps are: 

1. Select a benchmark portfolio matched to investor risk tolerance via simple combination of stocks and bonds. They represent stocks with a mix of 70% MSCI All World Country Index and 30% MSCI USA Index. They represent bonds with Barclays US Universal Bond Index. In their first illustration, they focus on 20-80, 60-40 and 80-20 stocks-bonds benchmarks, rebalanced quarterly.
2. Construct a strategic portfolio with the same expected volatility as the selected benchmark but generates a higher long-term Sharpe ratio by including optimized exposure to styles/factors expected to outperform the market over the long run. Key inputs are long-run asset returns and covariances plus a risk aversion parameter. In their first illustration, they constrain the strategic model portfolio to have the same overall equity exposure and regional equity exposures as the selected benchmark.
3. Add tactical modifications to the strategic portfolio by varying strategic positions based on short-term expected returns and risks. In their second illustration, they employ a 100-0 stocks-bonds benchmark consisting of 80% MSCI USA Net Total Return Index and 20% MSCI USA Minimum Volatility Net Total Return Index. The corresponding strategic portfolio reflecting long-term expectations is an equally weighted combination of value, momentum, quality, size and minimum volatility equity factor indexes. They specify short-term return and risk expectations based on four indicators involving: economic cycle variables; aggregate stock valuation metrics; factor momentum; and, dispersion of factor measures (such as difference in valuations between value stocks and growth stocks). They apply these indicators to underweight or overweight strategic positions using an optimizer. They rebalance these portfolios monthly. 

For their asset universe, they focus on indexes accessible via Exchanged Traded Funds (ETFs). Using monthly data for five broad capitalization-weighted equity indexes, six broad bond/credit indexes of varying durations and six style/factor (smart beta) equity indexes as available during January 2000 through June 2017, they find that: Keep Reading

What methods are available to suppress data snooping bias derived from testing multiple strategies/strategy variations on the same set of historical data? Which methods are best? In their March 2018 paper entitled “Systematic Testing of Systematic Trading Strategies”, Kovlin Perumal and Emlyn Flint survey statistical methods for suppressing data snooping bias and compare effectiveness of these methods on simulated asset return data and artificial trading rules. They choose a Jump Diffusion model to simulate asset return data, because it reasonably captures volatility and jumps observed in real markets. They define artificial trading rules simply in terms of probability of successfully predicting next-interval return sign. They test the power of each method by: (1) measuring its ability not to choose inaccurate trading rules; and, (2) relating confidence levels it assigns to strategies to profitabilities of those strategies. Using the specified asset return data and trading rule simulation approaches, they conclude that: Keep Reading