Momentum Investing

How lucky would a asset class picker with no skill have to be to match the performance of the Simple Asset Class Momentum Strategy (SACEMS), which each month picks winners from a set of eight exchange-traded funds (ETF) plus cash based on total returns over a specified lookback interval. To investigate, we run 1,000 trials of a “strategy” that each month allocates funds to one, the equally weighted two or the equally weighted three of these nine assets picked at random. We focus on gross compound annual growth rate (CAGR) and gross maximum drawdown (MaxDD) as key performance statistics. Using monthly total (dividend-adjusted) returns and for the specified assets during February 2006 (limited by DBC) through June 2019, we find that:

Keep Reading

Some traders use a Relative Strength Index (RSI) range to identify trend and RSI extremes to signal turning points. How long should they require that RSI remain in range, and how often should they require that RSI recapture a momentum threshold? In his December 2018 paper entitled “Finding Consistent Trends with Strong Momentum – RSI for Trend-Following and Momentum Strategies”, Arthur Hill systematically tests the predictive power of 14-day RSI range and momentum signals on S&P 500 stocks. Specifically, he tests each of the following five signals over lookback intervals of 25, 50, 75, 100 and 125 trading days:

1. RSI Bull Range: RSI between 40 and 100.
2. RSI Bear Range: RSI between 0 and 60.
3. RSI Bull Momentum: highest high value of RSI greater than 70.
4. RSI Bear Momentum: lowest low value of RSI less than 30.
5. RSI Bull Range-Momentum: combination of 1 and 3.

For example, 25-day RSI Bull Range signals buy at the close when 14-day RSI has been between 40 and 100 over the last 25 trading days and sell at the open when it next crosses below 40. His performance metrics are gross Success Rate (frequency of positive/negative returns after buy/sell signals) and gross Profit/Loss Ratio (average gain of successful trades divided by average loss of failed trades). Using daily prices for historical S&P 500 stocks during July 1998 through June 2018, he finds that:

Keep Reading

The Simple Asset Class ETF Momentum Strategy (SACEMS) each month picks winners based on total return over a specified ranking (lookback) interval from the following eight asset class exchange-traded funds (ETF), plus cash:

1. PowerShares DB Commodity Index Tracking (DBC)
2. iShares MSCI Emerging Markets Index (EEM)
3. iShares MSCI EAFE Index (EFA)
4. SPDR Gold Shares (GLD)
5. iShares Russell 2000 Index (IWM)
6. SPDR S&P 500 (SPY)
7. iShares Barclays 20+ Year Treasury Bond (TLT)
8. Vanguard REIT ETF (VNQ)
9. 3-month Treasury bills (Cash)

Based on findings in “SACEMS Portfolio-Asset Addition Testing”, a subscriber proposed adding iShares JPMorgan Emerging Market Bond Fund (EMB) to this set. To investigate, we revisit relevant analyses and conduct robustness tests, with focus on the equal-weighted (EW) Top 3 SACEMS portfolio. Using monthly dividend-adjusted closing prices for asset class proxies and the yield for Cash during February 2006 (when all ETFs in the baseline universe are first available) through June 2019, we find that: Keep Reading

How reliable and variable are the most widely accepted long-short factor premiums across asset classes? Can investors time factor premium? In their June 2019 paper entitled “Factor Premia and Factor Timing: A Century of Evidence”, Antti Ilmanen, Ronen Israel, Tobias Moskowitz, Ashwin Thapar and Franklin Wang examine multi-class robustness of and variation in four prominent factor premiums:

1. Value – book-to-market ratio for individual stocks; value-weighted aggregate cyclically-adjusted price-to-earnings ratio (P/E10) for stock indexes; 10-year real yield for bonds; deviation from purchasing power parity for currencies; and, negative 5-year change in spot price for commodities.
2. Momentum – past excess (relative to cash) return from 13 months ago to one month ago.
3. Carry – front-month futures-to-spot ratio for equity indexes since 1990 and excess dividend yield before 1990; difference in short-term interest rates for currencies; 10-year minus 3-month yields for bonds; and, percentage difference in prices between the nearest and next-nearest contracts for commodities.
4. Defensive – for equity indexes and bonds, betas from 36-month rolling regressions of asset returns versus equal-weighted returns of all countries; and, no defensive strategies for currencies and commodities because market returns are difficult to define.

They each month rank each asset (with a 1-month lag for conservative execution) on each factor and form a portfolio that is long (short) assets with the highest (lowest) expected returns, weighted according to zero-sum rank. When combining factor portfolios across factors or asset classes, they weight them by inverse portfolio standard deviation of returns over the past 36 months. To assess both overfitting and market adaptation, they split each factor sample into pre-discovery subperiod, original discovery subperiod and post-publication subperiod. They consider factor premium interactions with economic variables (business cycles, growth and interest rates), political risk, volatility, downside risk, tail risk, crashes, market liquidity and investment sentiment. Finally, they test factor timing strategies based on 12 timing signals based on 19 methodologies across six asset classes and four factors. Using data as available from as far back as February 1877 for 43 country equity indexes, 26 government bonds, 44 exchange rates and 40 commodities, all through 2017, they find that: Keep Reading

Is there a common optimal set of ranking (lookback) interval, holding interval, weighting scheme and skip-month rule for long-short stock momentum strategies across European country markets? In her May 2019 paper entitled “Are Momentum Strategies Profitable? Recent Evidence from European Markets”, Anastasia Slabchenko identifies optimal parameter sets from 576 long-short stock momentum strategy variations in each of France, Germany, Greece, Italy, Netherlands, Portugal, Spain, Sweden and UK. Variations derive from:

• Past return ranking (lookback) intervals of 1 to 12 months.
• Holding intervals of 1 to 12 months.
• Value weighted (VW) or equal-weighted (EW) momentum portfolios.
• Skip-month or no skip-month between ranking and holding intervals.

She defines the optimal variation as that generating the highest average gross monthly return. Using end-of-month closing prices for stock samples from each country, excluding financial stocks and stocks priced less than one euro, during December 1989 through January 2018, she finds that: Keep Reading

Do simple factor models help explain future return variations across different cryptocurrencies, as they do for stocks? In their April 2019 paper entitled “Common Risk Factors in Cryptocurrency”, Yukun Liu, Aleh Tsyvinski and Xi Wu examine performances of cryptocurrency (coin) counterparts for 25 price-related and market-related stock market factors, broadly categorized as size, momentum, volume and volatility factors. They first construct a coin market index based on capitalization-weighted returns of all coins in their sample. They then each week sort coins into fifths based on each factor and calculate average excess return for a portfolio that is long (short) coins in the highest (lowest) quintile. Finally, they investigate whether any small group of factors accounts for returns of all significant factors. Using daily prices in U.S. dollars and non-return variables (excluding top and bottom 1% values as potential errors/outliers) for all coins with market capitalizations over $1 million dollars from Coinmarketcap.com during January 2014 through December 2018 (a total of 1,707 coins, growing from 109 in 2014 to 1,583 in 2018), they find that:

Keep Reading

What steps should investors consider to mitigate impact of inevitable large U.S. stock market corrections? In their May 2019 paper entitled “The Best of Strategies for the Worst of Times: Can Portfolios be Crisis Proofed?”, Campbell Harvey, Edward Hoyle, Sandy Rattray, Matthew Sargaison, Dan Taylor and Otto Van Hemert compare performances of an array of defensive strategies with focus on the eight worst drawdowns (deeper than -15%) and three NBER recessions during 1985 through 2018, including:

1. Rolling near S&P 500 Index put options, measured via the CBOE S&P 500 PutWrite Index.
2. Credit protection portfolio that is each day long (short) beta-adjusted returns of duration-matched U.S. Treasury futures (BofAML US Corp Master Total Return Index), scaled retrospectively to 10% full-sample volatility.
3. 10-year U.S. Treasury notes (T-notes).
4. Gold futures.
5. Multi-class time-series (intrinsic or absolute) momentum portfolios applied to 50 futures contract series and reformed monthly, with:
   • Momentum measured for 1-month, 3-month and 12-month lookback intervals.
   • Risk adjustment by dividing momentum score by the standard deviation of security returns.
   • Risk allocations of 25% to currencies, 25% to equity indexes, 25% to bonds and 8.3% to each of agricultural products, energies and metals. Within each group, markets have equal risk allocations.
   • Overall scaling retrospectively to 10% full-sample volatility.
   • With or without long equity positions.
6. Beta-neutral factor portfolios that are each day long (short) stocks of the highest (lowest) quality large-capitalization and mid-capitalization U.S. firms, based on profitability, growth, balance sheet safety and/or payout ratios.

They further test crash protection of varying allocations to the S&P 500 Index and a daily reformed hedge consisting of equal weights to: (1) a 3-month time series momentum component with no long equity positions and 0.7% annual trading frictions; and, (2) a quality factor component with 1.5% annual trading frictions. For this test, they scale retrospectively to 15% full-sample volatility. Throughout the paper, they assume cost of leverage is the risk-free rate. Using daily returns for the S&P 500 Index and inputs for the specified defensive strategies during 1985 through 2018, they find that:

Keep Reading

Do all kinds of assets and long-short equity factor premiums exhibit exploitable time series (intrinsic or absolute momentum)? In their September 2018 paper entitled “Trends Everywhere”, Abhilash Babu, Ari Levine, Yao Hua Ooi, Lasse Pedersen and Erik Stamelos test intrinsic momentum on 58 traditional (studied in prior research) assets, 82 alternative (futures, forwards, and swaps not previously studied) assets and 16 long-short equity factors. They include only reasonably liquid (investable) assets and strategies. For equity factors, they each month: (1) classify over 4,000 U.S. common stocks as big or small according to NYSE median market capitalization; (2) within each size group, reform for each factor a value-weighted hedge portfolio that is long (short) the 30% of stocks with the highest (lowest) expected returns; and, (3) for each factor, average big and small hedge portfolio returns. They focus on a 12-month lookback interval for calculating momentum, taking a long (short) position in an asset/factor with positive (negative) return over this interval. For comparability of assets, they scale each position to an estimated 40% annualized volatility based on exponentially-weighted squared past daily returns. They assess diversification potentials by looking at pairwise correlations between momentum series, and between portfolios of momentum series and benchmark indexes (S&P 500 Index, MSCI World Index, Barclays Aggregate Bond Index and S&P GSCI Index). Using daily excess returns for the selected assets, factors and benchmarks as available during January 1985 through December 2017, they find that:

Keep Reading

Does return persistence of individual stocks predict associated option returns? In their March 2019 paper entitled “Stock Return Autocorrelations and the Cross Section of Option Returns”, Yoontae Jeon, Raymond Kan and Gang Li investigate relationships between equity option returns and return autocorrelations of underlying stocks. They consider call options, put options and straddles (long both a call and a put with the same strike price). Each month on standard option expiration date, they:

• Measure one-step monthly stock return autocorrelations using a 36-month rolling window of monthly returns for U.S. stocks with over 20 monthly observations.
• Rank stocks (and respective options) by autocorrelation into fifths (quintiles).
• Construct a hedge portfolio that is long (short) the equal-weighted or market capitalization-weighted stocks in the top (bottom) quintile of autocorrelations, to calculate stock portfolio return as a control variable.
• Construct corresponding hedge portfolios of call options, put options or straddles, limiting choices to reasonably liquid options with moneyness closest to 1.0 and time to expiration closest to 30 days. 
• Hold these portfolios until the next standard option expiration date.

They further explore out-of-sample use of results via modified mean-variance optimization of a portfolio consisting of the S&P 500 Index, the risk-free asset and equity options with bid-ask spreads no greater than 10% of price. They size individual option positions as a function of underlying stock volatility, variance risk premium and stock return autocorrelation. They assume investor utility derives from constant relative risk aversion level 3. For the frictionless case, they base option returns on the bid-ask midpoint. For the case with frictions, they assume buys (sells) occur at the ask (bid). Using specified stock and options data during January 1996 through December 2017, they find that: Keep Reading

Does crowding of factor investing strategies reliably predict returns for those strategies? In his March 2019 paper entitled “The Impact of Crowding in Alternative Risk Premia Investing”, Nick Baltas explores mechanics of alternative risk (factor) premium crowding and implications of crowding for future performance. He classifies factor premiums as: divergent (such as momentum), inherently destabilizing due to positive feedback loops and lack of fundamental anchors; or, convergent (such as value), having self-correcting negative feedback loops and fundamental anchors. To test crowding effects, he considers the following premiums: equity value (book-to-market), size (market capitalization), momentum (from regression of return from 12 months ago to one month ago versus volatility), quality (return on assets) and low beta (versus the MSCI World Index); commodities momentum (12-month return); and, currencies value (purchasing power parity) and momentum (12-month return). Each premium consists of returns from a hedge portfolio that is each week long (short) the equal-weighted assets with the highest (lowest) expected returns. For equities, he uses top and bottom tenths. For commodities and currencies, he uses top and bottom thirds. His crowding metric (CoMetric) is average pairwise correlation of factor-adjusted returns of assets within the long or short sides of premium portfolios over the last 52 weeks (except 260 weeks for value). He defines the 20% of weeks with the highest (lowest) CoMetrics as most (least) crowded. Using the specified factor and return data for liquid developed market stocks since September 2004, 24 constituents of the S&P GSCI Commodity Index since January 1999, and 26 developed and emerging markets currency pairs versus the U.S. dollar since January 2000, all through May 2018, he finds that:

Keep Reading