Calendar Effects

Does the U.S. stock market exhibit predictable return and volatility patterns during the trading day? To investigate, we analyze one-minute prices for SPDR S&P 500 (SPY) over two recent years. Specifically, we calculate average cumulative return, average returns for 15-minute intervals and average standard deviation of one-minute returns during 15-minute intervals over the trading day during each of 2007 (bullish year) and 2008 (bearish year). Using a sample of SPY one-minute prices for 9:30-16:15 spanning 2007-2008 (over 203,000 observations), we find that: Keep Reading

Is the monthly stock return reversal effect currently exploitable? In the August 2011 version of their paper entitled “New Evidence on Short-Term Reversals in Monthly Stock Returns: Overreaction or Illiquidity?”, Chris Stivers and Licheng Sun investigate the persistence, size-sensitivity and seasonality of monthly stock return reversal in the context of three competing explanations: (1) investor overreaction to news (exploitable); (2) market illiquidity (perhaps unexploitable); and, (3) large stocks lead small stocks (exploitable). They evaluate simple value-weighted and equal-weighted prior-month loser-minus-winner (LMW) strategies based on a sort of prior-month returns, and five more complex equal-weighted LMW strategies based on double-sorts of prior-month returns and market capitalizations. Using monthly return and market capitalization data for a broad sample of U.S. stocks and 30 industries over the period February 1926 through December 2010, they find that: Keep Reading

Do stocks around the world tend to perform better around the time of annual earnings announcements by respective firms than during the rest of the year? In the June 2011 draft of their paper entitled “The Earnings Announcement Premium Around the Globe”, Brad Barber, Emmanuel De George, Reuven Lehavy and Brett Trueman investigate whether the earnings announcement premium (elevated returns during earnings announcement months) is a global phenomenon or is isolated to U.S. stocks. They employ a hedge portfolio, reformed monthly, that is long (short) stocks of firms expected (not expected) to announce annual earnings during the next month, The long and short sides are equal-weighted, and the stocks within each side are value-weighted. Using roughly 200,000 annual earnings announcements for about 28,000 firms in 46 countries during 1990 through 2009 to estimate announcement months during 1991-2010, and associated monthly stock returns, they find that: Keep Reading

Are there recognizable country and global financial cycles over the past half century? If so, what are their characteristics? In their April 2011 paper entitled “Financial Cycles: What? How? When?”, Stijn Claessens, Ayhan Kose and Marco Terrones employ regression models to investigate cycles in credit, housing and equity markets for developed countries since 1960. They define a downturn as peak to trough and an upturn as trough to level of previous peak (not trough to new peak). They define series peaks and troughs with the constraints that complete cycle duration must be at least five quarters and each upturn and downturn must be at least two quarters. The main characteristics of cyclical phases are their duration, amplitude and slope, with crunches/busts (booms) defined as downturns (upturns) in the bottom (top) fourth of all observations. They examine pre-globalization (1960-1985) and globalization (1986-2007) subperiods, with phases globally synchronized (highly synchronized) when more than 40% (50%) of countries experience the same phase. Using quarterly data for aggregate loans to the private sector, house/land price indexes and value-weighted stock market indexes for 21 developed countries over the period 1960 through 2007, seasonally adjusted as necessary, they find that: Keep Reading

Are there reliable ways to exploit differences in asset class returns under Democratic and Republican U.S. presidents? In his April 2011 paper entitled “Is the 60-40 Stock-Bond Pension Fund Rule Wise?”, William Ziemba examines relationships between the U.S. presidential election cycle and long-run returns for several asset classes. Specifically, he investigates the differential performance of large capitalization stocks, small capitalization stocks and bonds when Democrats and Republicans hold the presidency. Using annual asset class return data for 1998 through 2010 to extend prior calculations for 1937-1997 and 1942-1997, he finds that: Keep Reading

In their February 2011 book chapter entitled “Seasonal Anomalies”, Constantine Dzhabarov and William Ziemba describe, update and assess several published U.S. stock market anomalies, most of which are directly or indirectly calendar-driven. They update using returns for stock index futures as a low-friction approach to exploiting calendar anomalies. They acknowledge the possible materiality of data mining/snooping bias in past findings and the difficulty of proving statistical significance even in large samples due to high return variability. Using returns for S&P 500 Index and Russell 2000 Index futures from February 1993 through December 2010 to update analyses of these anomalies, they find that: Keep Reading

Does recent data suggest any reliable day-of-the-week U.S. stock market return anomalies? To investigate, we examine close-to-close stock market returns for the five trading days of the week, excluding trading days before and after market holidays and the day after the extended market disruption in September 2001. The intent of these exclusions is to suppress holiday effects on stock market returns. Using daily closing prices for the S&P 500 Index for 1981 through 2010 (7,301 daily returns, ranging from 1,382 Fridays to 1,546 Wednesdays), we find that: Keep Reading

Are January returns important to the profitability of short-term reversal, intermediate-term momentum, long-term reversion and 12-month echo trading strategies? In her December 2010 paper entitled “Momentum, Seasonality and January”, Yaqiong Yao investigates the role of  January returns within these previously discovered anomalies. The study’s core methodology is to reform equally weighted hedge portfolios each month that are long/short stocks in extreme tenths (deciles) of  past returns over various intervals.  The one-month reversal strategy is long (short) losers (winners) based on prior month returns. Momentum strategies are long (short) winners (losers) based on past 11-month or 12-month returns, with a skip month before portfolio formation to avoid short-term reversal. The reversion strategy is long (short) losers (winners) based on past four-year returns, with a skip-year before portfolio formation to avoid intermediate-term momentum. The 12-month echo strategy is long (short) winners (losers) based on returns for the same month the prior one, two or three years. Using monthly returns for a broad sample of NYSE/AMEX stocks during 1926 through 2009, she finds that: Keep Reading

What variables are persistently effective in picking equity sectors for tactical (monthly) trading? In their July 2010 paper entitled “Global Tactical Sector Allocation: A Quantitative Approach”, Ronald Doeswijk and Pim van Vliet investigate the effectiveness of seven variables for tactical trading of ten global equity sector indexes. They test effectiveness of these variables separately and in combination, and after their respective publication dates. The seven variables are: one-month return momentum, 12-1 return momentum (over the 11 months prior to the last month), earnings revision trend, long-term return (over the four years prior to the last year) reversion, aggregate dividend yield, Federal Reserve policy (expansive or contractive) and sell-in-May seasonal.  The ten sectors are energy, materials, industrials, consumer discretionary, consumer staples, health care, financials, information technology, telecommunication services and utilities.  Testing consists of monthly construction of equally weighted long-short portfolios based on variable conditions. For the first five variables, portfolios are long (short) the top (bottom) three sectors. The Federal Reserve policy and sell-in-May seasonal variables indicate whether to be long or short cyclical versus defensive sectors. The authors calculate net profitability based on a constant 0.60% round-trip trading friction. Using monthly sector index total returns and values for non-return variables mostly over the period 1970 through 2008, they find that: Keep Reading

How well do calendar-based anomalies, such as the January and Halloween/Sell-in-May effects, hold up for data extending back three centuries? Do any new anomalies emerge from such a data set? In their October 2010 paper entitled “Are Monthly Seasonals Real? A Three Century Perspective”, Ben Jacobsen and Cherry Zhang examine an extremely long record of UK stock returns for evidence of calendar anomalies. Using 317 years of monthly UK stock index returns and risk-free rate proxies spanning 1693 through 2009, they find that: Keep Reading