Momentum Investing

Which works best, strategies comparing past returns among assets (relative or cross-sectional) or strategies requiring positive past raw/excess returns (intrinsic or absolute or time series)? In their May 2015 paper entitled “Cross-Sectional and Time-Series Tests of Return Predictability: What is the Difference?”, Amit Goyal and Narasimhan Jegadeesh investigate differences between relative and intrinsic past return strategies, focusing on individual U.S. common stocks. For relative return strategies, they construct portfolios that are long (short) stocks with above-average (below-average) past returns, with the long and short sides weighted equally. For intrinsic return strategies, they construct portfolios that are long (short) stocks with past returns greater (less) than the risk-free rate, with each stock weighted equally. They consider past return measurement (ranking) intervals and holding intervals each ranging from one month to 60 months. They also compare relative and intrinsic past return strategies across and within global asset classes (17 equity indexes, 21 bond indexes, 24 commodity spot series and 8 currencies). Finally, they apply the relative and intrinsic concepts to individual U.S. stock financial ratios (book/market, gross profit/assets, asset growth and accruals/assets), with relative comparing once a year to the average for all stocks and intrinsic comparing once a year stock-by-stock to respective median values over the past five years. Using data for a broad sample of U.S. common stocks above the 20th percentile of NYSE market capitalization during 1946 through 2013, and for the asset class series during 1985 through 2013, they find that: Keep Reading

Is there strong evidence for price momentum within and across all major asset classes over the long run? In the May 2015 version of their paper entitled “215 Years of Global Multi-Asset Momentum: 1800-2014 (Equities, Sectors, Currencies, Bonds, Commodities and Stocks)”, Christopher Geczy and Mikhail Samonov examine the momentum effect for very long price histories within and across major asset class universes consisting of: 47 country equity markets; 301 global equity sectors; 43 government bond series; 76 commodities; 48 currencies; and, 34,795 U.S. stocks. Their baseline momentum metric is return from 12 months ago to two months ago, with two skipped months to avoid any short-term reversal, followed by a one-month holding interval. Their baseline momentum portfolio is long (short) the third of assets with highest (lowest) momentum, equally weighted and reformed monthly. Their benchmarks are the equally weighted returns of all assets in the universes. Using monthly price data for the specified assets as available during 1800 through 2014 (215 years), they find that: Keep Reading

Do factors that predict returns in U.S. stock data also work on global stock markets at the country level? In the May 2015 version of their paper entitled “Do Quantitative Country Selection Strategies Really Work?”, Adam Zaremba and Przemysław Konieczka test 16 country stock market selection strategies based on relative market value, size, momentum, quality and volatility. For each of 16 factors across these categories, they sort country stock markets into fifths (quintiles) and measure the factor premium as return on the highest minus lowest quintiles. They consider equal, capitalization and liquidity (average turnover) weighting schemes within quintiles. They look at complementary large and small market subsamples, and complementary open (easy to invest) and closed market subsamples. Using monthly total returns adjusted for local dividend tax rates in U.S. dollars for 78 existing and discontinued country stock indexes (primarily MSCI) during 1999 through 2014, they find that: Keep Reading

Is intermediate-term asset class momentum a useful way to generate inputs (return, volatility and correlation forecasts) for a multi-class mean-variance optimization strategy? In their May 2015 paper entitled “Momentum and Markowitz: a Golden Combination”, Wouter Keller, Adam Butler and Ilya Kipnis test the effectiveness of using intermediate-term lookback intervals (1 to 12 months) to generate monthly long-only mean-variance optimized portfolios. They argue that such lookback intervals are more likely than conventional long (multi-year) intervals to provide forecasts that persist during one-month portfolio holding intervals. They name their approach Classical Asset Allocation (CAA). To test CAA, in addition to adopting the practical long-only constraint, they further:

1. Select from the efficient frontier a target annualized portfolio volatility of either 10% (aggressive) or 5% (conservative).
2. Forecast asset returns by averaging results from lookback intervals of 1, 3, 6 and 12 months.
3. Forecast covariances (volatility-correlation relationships) from a 12-month lookback interval.
4. Cap portfolio weights for risky assets at 25%, but do not cap weights for 3-month U.S. Treasury bills (T-bills) and 10-year U.S. Treasury notes (T-notes).
5. Consider three universes of 8, 16 and 39 asset class proxies.
6. Use equal weighting (EW) of all assets in a universe as a benchmark.

They introduce an optimizer program to streamline calculation of optimal portfolio weights. Using monthly total returns for 39 indexes spanning multiple asset classes as available during January 1914 through December 2014, they find that: Keep Reading

Can investors amplify stock return momentum by screening past winners and losers based on return skewness? In their April 2015 paper entitled “Expected Skewness and Momentum”, Heiko Jacobs, Tobias Regele and Martin Webee explore the interaction of expected stock return skewness and momentum. They measure expected skewness as maximum daily return over the preceding month, which predicts future skewness more accurately than does past skewness. Their benchmark is a conventional momentum portfolio that is each month long (short) the fifth, or quintile, of stocks with the highest (lowest) returns from 12 months ago to one month ago. To test the interaction of expected skewness and momentum, they first sort stocks into quintiles based on expected skewness and then sort each expected skewness quintile into quintiles based on momentum (25 total portfolios). Their skewness-weakened momentum portfolio is long (short) winners (losers) with relatively high/positive (low/negative) expected skewness. Their skewness-enhanced momentum portfolio is long (short) winners (losers) with relatively low/negative (high/positive) expected skewness. Using daily and monthly returns for a broad sample of U.S. common stocks (excluding very small and illiquid stocks) during January 1926 through December 2011, they find that: Keep Reading

When does value investing work and how does it work best? In the April 2015 initial draft of their paper entitled “Fact, Fiction, and Value Investing”, Clifford Asness, Andrea Frazzini, Ronen Israel and Tobias Moskowitz address areas of confusion about value investing. They describe value as the tendency of cheap securities to outperform expensive ones based on some valuation method. They broadly specify the value premium as the return achieved by holding or overweighting cheap securities and shorting or underweighting expensive ones. They focus on systematic (mechanical), diversified value strategies based on quantified metrics such as book-to-market ratio or earnings-price ratio. Their context is firm belief that such strategies are great investments. Based on academic studies and simple tests with recent data, largely from Kenneth French’s data library, they conclude that: Keep Reading

How sensitive is the performance of the “Simple Asset Class ETF Momentum Strategy” to selecting ranks other than winners and to choosing a momentum ranking interval other than five months? This strategy each month ranks the following eight asset class exchange-traded funds (ETF), plus cash, on past return and rotates to the strongest class:

PowerShares DB Commodity Index Tracking (DBC)
iShares MSCI Emerging Markets Index (EEM)
iShares MSCI EAFE Index (EFA)
SPDR Gold Shares (GLD)
iShares Russell 1000 Index (IWB)
iShares Russell 2000 Index (IWM)
SPDR Dow Jones REIT (RWR)
iShares Barclays 20+ Year Treasury Bond (TLT)
3-month Treasury bills (Cash)

Available data are so limited that sensitivity test results may mislead. With that reservation, we perform two robustness/sensitivity tests: (1) comparison of returns for all nine ranks of winner through loser based on a ranking interval of five months and a holding interval of one month (5-1); and, (2) comparison of winner returns for ranking intervals ranging from one to 12 months (1-1 through 12-1) and for a six-month lagged six-month ranking interval (12:7-1) per “Isolating the Decisive Momentum (Echo?)”, all with one-month holding intervals. Using monthly adjusted closing prices for the asset class proxies and the yield for Cash over the period July 2002 (or inception if not available then) through April 2014 (154 months), we find that: Keep Reading

The order of the first and second place winners is now reversed from that shown at the close yesterday because of a price change on Yahoo!Finance after 4:00PM. Keep Reading

Which moving average rules and measurement (lookback) intervals work best? In the March 2015 version of his paper entitled “Market Timing with Moving Averages: Anatomy and Performance of Trading Rules” Valeriy Zakamulin compares market timing rules based on different kinds of moving averages, including simple momentum. He first compares the mathematics of these rules to identify similarities and differences. He then conducts very long run out-of-sample tests of a few trading rules with distinct weighting schemes to measure their market timing effectiveness. He tries both an expanding window (inception-to-date) and rolling windows to discover optimal lookback intervals. He uses Sharpe ratio as his principal performance metric. He estimates one-way trading friction as a constant 0.25%. Using monthly returns for the S&P Composite Index and for the risk-free asset during January 1860 through December 2009, he finds that: Keep Reading

Which stock momentum return predictor works best? In his March 2015 paper entitled “Momentum Crash Management”, Mahdi Heidari compares the crash protection effectiveness of seven stock momentum return predictors, categorized into two groups: 

1. Overall stock market statistics: prior-month market return; change in monthly market return; volatility of market returns (standard deviation of weekly returns for the past 52 weeks); cross-sectional dispersion of daily stock returns for the past month; and, market illiquidity (value-weighted average of the monthly averages of daily price impacts of trading for all stocks).
2. Momentum return series statistics: volatility of momentum returns (standard deviation of monthly returns over the past six months); and monthly change in volatility of momentum returns.

He measures momentum conventionally by first ranking all stocks by their returns from 12 months ago to one month ago and then after the skip-month forming a hedge portfolio that is long (short) the value-weighted tenth of stocks with the highest (lowest) past returns. He next tests the power of the above seven variables to predict the resulting monthly momentum return series. Finally, he tests dynamic momentum risk management strategies that execute the conventional momentum strategy (go to cash) when each of the seven predictors is below (above) the 90 percentile of its values over the last five years. Using daily and monthly returns, daily trading volumes and shares outstanding for a broad sample of U.S. common stocks during January 1926 through December 2013, he finds that: Keep Reading