Calendar Effects

“Intraday U.S. Stock Market Behavior” examines behavior of the S&P 500 Index at 15-minute intervals over the trading day during each of 2007 (bullish year) and 2008 (bearish year), finding slight tendencies for market weakness during mid-afternoon and market volatility at the beginning and the end of the trading day. Does recent data confirm these findings? To investigate, we calculate average cumulative returns and standard deviations of returns for both the S&P 500 Index and SPDR S&P 500 (SPY) measured at 5-minute intervals during the trading day over the last six months. Using 5-minute levels/prices for the S&P 500 Index and for SPY during 9:30-16:00 over the period August 2012 through September 2014, we find that: Keep Reading

Is the Turn-of-the-Month (TOTM) effect globally ubiquitous and persistent? In his August 2014 paper entitled “The Turn-of-The-Month-Effect: Evidence from Periodic Generalized Autoregressive Conditional Heteroskedasticity (PGARCH) Model”, Eleftherios Giovanis examines the TOTM effect in 20 country stock markets spanning the Americas, Australia, Europe and Asia. He defines TOTM as the interval including the last trading day of each calendar month through the third trading day of the next calendar month. He applies complex techniques to account for potential autocorrelation, heteroskedasticity and volatility clustering in daily market returns. His samples vary in start date by country, from as early as January 1950 (for the U.S.) to as late as January 2001 (for Australia). He considers full samples from the beginning of each country series through 2013 and two subsamples: (1) from the beginning of each country sample through 2007; and, (2) the financial crisis of 2008 through 2009. Using daily closes for the 20 country stock market indexes as described, he finds that: Keep Reading

Does the stock momentum anomaly interact with the quarterly financial cycle? In his August 2014 paper entitled “Seasonal Patterns in Momentum and Reversal in the U.S. Stock Market: The Consequences of Tax-Loss Sales and Window Dressing”, David Brown examines whether tax-loss selling and window dressing at the ends of calendar quarters affect U.S. stock momentum strategy returns. Each month, he ranks stocks by returns over the last 12 months, skipping the last month to avoid reversal, and then forms a momentum hedge portfolio that is long (short) the capitalization-weighted tenth of stocks with the highest (lowest) past returns, making the long and short sides of the portfolio equal in magnitude. He then measures how this portfolio performs by calendar month to check for end-of-quarter effects. He also investigates whether the level of capital losses among stocks in the portfolio affects performance. Using monthly returns for NYSE, AMEX and NASDAQ common stocks, along with contemporaneous risk-free rates and Fama-French model risk factor returns, during January 1927 through December 2013, he finds that: Keep Reading

In response to “Optimal Monthly Cycle for Simple Asset Class ETF Momentum Strategy?”, a subscriber asked about the optimal monthly cycle for “Simple Sector ETF Momentum Strategy”, which each month allocates all funds to the one of the following nine Select Sector Standard & Poor’s Depository Receipts (SPDR) exchange-traded funds (ETF) with the highest total return over the past six months :

Materials Select Sector SPDR (XLB)
Energy Select Sector SPDR (XLE)
Financial Select Sector SPDR (XLF)
Industrial Select Sector SPDR (XLI)
Technology Select Sector SPDR (XLK)
Consumer Staples Select Sector SPDR (XLP)
Utilities Select Sector SPDR (XLU)
Health Care Select Sector SPDR (XLV)
Consumer Discretionary Select SPDR (XLY)

To investigate, we compare 21 variations of the strategy based on shifting the monthly return calculation cycle relative to trading days from the end of the month (EOM). For example, an EOM+5 cycle ranks assets based on closing prices five trading days after EOM each month. Using daily dividend-adjusted closes for the sector ETFs from mid-January 1999 through mid-July 2014 (about 186 months), we find that:

Keep Reading

Does intensity of firm quarterly earnings releases affect stock market behaviors? A reader proposed the following stock market timing strategy based on a strictly calendar-based definition of earnings season: go short (long) the market at the close at the end of the first full week (sixth full week) of each calendar quarter, representing the beginning (end) of earnings season. The hypothesis is that the broad stock market performs poorly during earnings season and well outside of earnings season. Using weekly closes for the S&P 500 Index since January 1950 and for the S&P 500 Implied Volatility Index (VIX) since January 1990, both through June 2014, we find that: Keep Reading

Do returns for segments of the normal U.S. stock market trading day (9:30 AM to 4:00 PM Eastern time) exhibit exploitable interactions? In the May 2014 version of their paper entitled “Intraday Momentum: The First Half-Hour Return Predicts the Last Half-Hour Return”, Lei Gao, Yufeng Han and Guofu Zhou investigate intraday U.S. stock market predictability based on half-hour segments. They focus on interaction between returns for the first and last half-hour segments. Using half-hour returns for SPDR S&P 500 (SPY) since January 1999 and for PowerShares QQQ (QQQ) since March 1999 and contemporaneous release dates for major economic statistics through December 2012, they find that: Keep Reading

Does the Sell-in-May/Halloween effect hold in recent data? In their April 2014 paper entitled “Sell in May and Go Away: Still Good Advice for Investors?”, Hubert Dichtl and Wolfgang Drobetz explore whether holding one of several stock indexes (cash) during November-April (May-October) beats buying and holding the index. They focus on sample periods since: (1) liquid index proxies are readily available for each index to both institutional and individual investors; and, (2) first publication in a top academic journal confirming the Halloween effect. They use both conventional regressions and bootstrap simulations. They consider six mostly total return indexes: S&P 500, DAX 30, FTSE 10, CAC 40, EuroStoxx 50 (not total return) and MSCI Emerging Markets (EM). They use a one-month interest rate for the return on cash. They apply a range of switching frictions to assess sensitivity of results to trading costs. Using monthly returns for the specified indexes from the first available January for each through December 2012 (so that they always work with full calendar years), they find that: Keep Reading

Five years ago, a reader noted and asked: “CNBC’s Fast Money cited a ‘seasonal’ strategy described in Barron’s, as follows: Go long the market from Wal-Mart’s (WMT) earnings release until Alcoa’s (AA) earnings release and short the market from Alcoa’s earnings release until Wal-Mart’s earnings release (earnings season). Over the last six years, the market has been up nicely during the former period and down an average 8% during the latter. Any testing on this?” To test this strategy, we assemble AA earnings release dates and WMT earnings release dates since the beginning of 1997 (the earliest available for AA), estimating the date for one missing WMT release. This sample period is much longer than that cited. Using these earnings release dates, daily dividend-adjusted closes for S&P 500 SPDR (SPY) and ProShares Short S&P500 (SH) and the daily 13-week Treasury bill (T-bill) yields over the period 2/25/97 (6/21/06 for SH) through 10/8/13 (67 quarters), we find that: Keep Reading

Does the Halloween effect (sell in May) still hold? In the June 2013 version of their paper entitled “Are Stock Markets Really so Inefficient? The Case of the ‘Halloween Indicator'”, Hubert Dichtl and Wolfgang Drobetz investigate whether true out-of-sample results confirm that the Halloween effect persists for five total return indexes: S&P 500, DAX 30, FTSE 100, CAC 40 and Euro Stoxx 50. They consider both regression tests to compare average monthly returns and simulated trading strategies based on bootstrapping. They consider three sample periods for each index: (1) the total available sample period; (2) the subperiod since availability of a liquid investment proxy (such as a mutual fund or exchange-traded fund) to exploit the Halloween effect; and, (3) the subperiod starting January 2003 (after publication of the seminal international study of the effect). For simulated trading strategies, they assume switches from stocks to cash at the end of April and cash to stocks at the end of October incur trading friction of 0.5%. Using monthly total returns of the specified indexes as available (starting years ranging from 1965 to 1988) through December 2012, and contemporaneous one-month local interest rates as the return on cash, they find that: Keep Reading

Are there worldwide anomalies with regard to equity market returns by calendar month? In his June 2013 paper entitled “Stock Market Performance: High and Low Months”, Vichet Sum examines stock market performance in 70 countries to determine which months generate relatively high and low returns. He weights country stock markets equally in calculating worldwide statistics. Using monthly returns as available (with many series beginning in the 1980s and 1990s) mostly through May 2012, he finds that: Keep Reading