Momentum Investing

What determines whether a stock will exhibit short-term momentum or short-term reversal? In their May 2022 paper entitled “Short-term Relative-Strength Strategies, Turnover, and the Connection between Winner Returns and the 52-week High”, building upon prior research, Chen Chen, Chris Stivers and Licheng Sun investigate interactions among proximity to 52-week high, share turnover and 1-month return momentum/reversal behaviors for U.S. stocks.  Specifically, at the end of each month t, they form 125 portfolios by:

1. Sorting stocks into fifths (quintiles) based on return during month t.
2. Further sorting these quintiles stocks into sub-quintiles based on ratio of price at the end of month t-1 to highest price over the preceding 52 weeks.
3. Further sorting the sub-quintiles into sub-sub-quintiles based on share turnover during month t.

They then use month t+1 value-weighted returns of the resulting 125 portfolios to evaluate short-term momentum/reversal strategies in multiple ways: buying winners and shorting losers (momentum); buying losers and shorting winners (reversal); and, winners-only or losers-only strategies based on 52-week high proximity. Using the specified trading data for a broad sample of U.S. common stocks priced at least $1 during July 1963 to December 2020, they find that:

Keep Reading

A subscriber suggested review of a Follow the Leader (FTL) strategy that, in simplest form, each month holds iShares MSCI Emerging Markets ETF (EEM) when prior-month SPDR S&P 500 ETF Trust (SPY) return is positive and iShares 20+ Year Treasury Bond ETF (TLT) when prior-month SPY return is negative. To investigate, we run this simplest FTL strategy over the available history of EEM. We use buy-and-hold SPY as a benchmark. Using monthly dividend-adjusted returns for SPY since March 2003 and for EEM and TLT since April 2003, all through April 2022, we find that: Keep Reading

Do intraday (open-to-close, trading-driven/technical) and overnight (close-to-open, news-driven) stock returns play different roles in signaling short-term reversal, intermediate-term momentum and long-term reversion? In their March 2022 paper entitled “What Drives Momentum and Reversal? Evidence from Day and Night Signals”, Yashar Barardehi, Vincent Bogousslavsky and Dmitriy Muravyev investigate by relating past cumulative daily, intraday and dividend-adjusted overnight returns over various lookback intervals to future returns. The lookback intervals they consider are:

• For reversal: last month.
• For momentum: from seven or 12 months ago to one month ago.
• For reversion: from 36, 48 or 60 months ago to 12 months ago.

They measure reversal, momentum and reversion effects via monthly gross 3-factor (market, size, book-to-market) alphas of hedge portfolios that are each month long (short) the extreme tenth, or decile, of stocks with the highest (lowest) past returns. Using the specified inputs for a narrow sample of U.S. common stocks during January 1926 through December 1962, and for a broad sample of U.S. common stocks during January 1963 through December 2019, they find that: Keep Reading

A subscriber proposed substituting SPDR S&P Global Natural Resources ETF (GNR) for Invesco DB Commodity Index Tracking Fund (DBC) as a proxy for commodities in the Simple Asset Class ETF Momentum Strategy (SACEMS). GNR holds stocks of relatively large firms engaged in natural resources and commodities businesses. DBC holds a range of commodity futures contracts. To investigate, we run SACEMS since June 2006 but substitute GNR for DBC as soon as GNR becomes available in September 2010 (dovetailing with older data). Using dividend-adjusted closing prices for SACEMS asset class proxies and GNR and the yield for Cash during February 2006 (per tracked SACEMS) through April 2022, we find that: Keep Reading

Can investors improve the performance of stock momentum portfolios by isolating stocks that “hold” their momentum? In their April 2022 paper entitled “Enduring Momentum”, Hui Zeng, Ben Marshall, Nhut Nguyen and Nuttawat Visaltanachoti exploit firm characteristics to identify stocks that continue to be winners or losers after selection as momentum stocks (stocks with enduring momentum). They measure momentum by each month ranking stocks into equal-weighted tenths, or deciles, based on past 6-month returns, with the top (bottom) decile designated winners (losers). They then develop a model that uses information from 37 firm characteristics to estimate each month the probability that each winner or loser stock will continue as a winner or loser during each of the next six months. They verify that the model reasonably predicts momentum persistence and proceed to test the economic value of the predictions by each month reforming an enduring momentum hedge portfolio that is long (short) the 10 equal-weighted winner (loser) stocks with the highest probabilities of remaining winners (losers) and holding the portfolio for six months. They compare the performance of this portfolio to that of a conventional momentum portfolio that is each month long the entire winner decile and short the entire loser decile, also held for six months. Using returns for a broad sample of U.S. common stocks priced over $1.00 and 37 associated firm characteristics during January 1980 through December 2018, they find that: Keep Reading

Do U.S. stock momentum exchange-traded funds (ETF) deliver attractive performance? In their February 2022 paper entitled “A Look Under the Hood of Momentum Funds”, Ayelen Banegas and Carlo Rosa examine the performance of U.S. stock momentum funds. To measure performance uncontaminated by discretionary stock picking and portfolio construction methods, they focus on mechanical (passive) momentum ETFs. They segment selected ETFs into large-cap and small-cap groups to avoid any confounding size effect. Using monthly net returns and assets under management (AUM) for live and dead U.S. equity momentum ETFs with AUM greater than $50 million during January 2006 through October 2021, they find that: Keep Reading

How should investors think about surprises in economic data? In their March 2022 paper entitled “Caught by Surprise: How Markets Respond to Macroeconomic News”, Guido Baltussen and Amar Soebhag devise and investigate a real-time aggregate measure of surprises in economic (not financial) variables around the world. Each measurement for each variable consists of release date/time, initial as-released value, associated consensus (median) forecast, number and standard deviation of individual forecasts and any revision to the previous as-released value across U.S., UK, the Eurozone and Japan markets from the Bloomberg Economic Calendar. They classify variables as either growth-related or inflation-related. They apply recursive principal component analysis to aggregate individual variable surprises separately into daily nowcasts of initial growth-related and inflation-related announcement surprises and associated revision surprises. They investigate the time series behaviors of these nowcasts and then examine their interactions with returns for four asset classes:

1. Stocks via prices of front-month futures contracts rolled the day before expiration for S&P 500, FTSE 100, Nikkei 225 and Eurostoxx 50 indexes.
2. Government bonds via prices of front-month futures contracts rolled the day before first notice on U.S., UK, Europe and Japan 10-year bonds.
3. Credit via returns on 5-year credit default swaps for U.S. and Europe investment grade and high yield corporate bond indexes.
4. Commodities via excess returns for the Bloomberg Commodity Index.

Specifically, they test an investment strategy that takes a position equal to the 1-day lagged value of the growth surprise nowcast or the inflation surprise nowcast on the last trading day of each month. They pool regions within an asset class by equally weighting regional markets. Using daily as-released data for 191 economic variables across global regions and the specified monthly asset class price inputs during March 1997 through December 2019, they find that: Keep Reading

A subscriber suggested comparing intrinsic momentum (IM), also called absolute momentum and time series momentum, to simple moving average (SMA) as alternative signals for equity market entry and exit. To investigate across a wide variety of economic and market conditions, we measure the long run performances of entry and exit signals from IMs over past intervals of one to 12 months (IM1 through IM12) and SMAs ranging from 2 to 12 months (SMA2 through SMA12). We consider two cases for IM signals and one case for SMA signals, as applied to the S&P 500 Index as a proxy for the stock market and the 3-month U.S. Treasury bill (T-bill) as a proxy for cash (the risk-free rate). The three rule types are therefore:

1. IMs Case 1 – in stocks (cash) when past index return is positive (negative).
2. IMs Case 2 – in stocks (cash) when average monthly past index return is above (below) average monthly T-bill yield over the same interval.
3. SMAs – in stocks (cash) when the index is above (below) the SMA.

We estimate S&P 500 Index monthly total returns using quarterly dividend yield calculated from Shiller data for March, June, September and December. This estimation does not affect index timing signals. We focus on net compound annual growth rate (CAGR), maximum drawdown (MaxDD) and annual Sharpe ratio as key performance metrics, with baseline stocks-cash switching frictions 0.2%. We use buying and holding the S&P 500 Index (B&H) as a benchmark. Using monthly closes of the S&P 500 Index during December 1927 through February 2022 (94 years), and contemporaneous monthly index dividend and T-bill yields, we find that:

Keep Reading

Which of the explanations offered in past research best explains, and potentially justifies belief in persistence of, stock return momentum? In the February 2022 revision of their paper entitled “What Explains Momentum? A Perspective From International Data”, Amit Goyal, Narasimhan Jegadeesh and Avanidhar Subrahmanyam test alternative explanations for momentum as hypothesized in earlier studies on a recent international dataset. They each month calculate momentum for each stock as its return over the last 12 months excluding the most recent month, standardized by subtracting the average momentum for all stocks of the same country. They then rank stocks into tenths (deciles) based on country-standardized momentum to form a value-weighted hedge portfolio that is long the winner decile and short the loser decile. They employ hedge portfolio returns to test explanatory powers of measurable proxies for the following rationales:

• Overconfidence that escalates with returns.
• Generally slow diffusion of news among investors.
• Anchoring bias, specified as ratio of current price to 52-week high.
• Frog-in-the-pan (investors underreact to small bits of news that arrive gradually due to limited attention, but react appropriately to shocking news).
• Stock risk that varies with past returns (high for winners and low for losers).

They also examine whether hedge portfolio profitability varies across broad market past return and volatility states. Using data groomed to exclude obvious errors for both listed and delisted stocks across 22 non-U.S. developed markets and 27 emerging markets during 1993 through 2020, they find that: Keep Reading

“SACEMS with Momentum Breadth Crash Protection” evaluates in depth the potential of a simple momentum breadth rule to improve performance of the Simple Asset Class ETF Momentum Strategy (SACEMS). This rule forces the model to all cash when fewer than some threshold of the non-cash SACEMS assets have positive returns over a specified lookback interval. Do major findings of that evaluation still hold? To update, we repeat some of the analyses with the minor changes since made to SACEMS plus recent data. We focus on compound annual growth rates (CAGR) and maximum drawdowns (MaxDD) for the Top 1, equal-weighted (EW) Top 2 and EW Top 3 SACEMS portfolios. We look at all possible momentum breadth thresholds for the baseline SACEMS lookback interval. We then consider lookback intervals ranging from one to 12 months for a specific momentum breadth threshold. Using monthly dividend-adjusted closing prices for SACEMS assets and the T-bill yield during February 2006 through February 2022, we find that:

Keep Reading