Calendar Effects

How does execution delay affect the performance of the Best Value and Weighted versions of the “Simple Asset Class ETF Value Strategy” (SACEVS)? These strategies each month allocate funds to the following asset class exchange-traded funds (ETF) according to valuations of term, credit and equity risk premiums, or to cash if no premiums are undervalued:

3-month Treasury bills (Cash)
iShares 20+ Year Treasury Bond (TLT)
iShares iBoxx $ Investment Grade Corporate Bond (LQD)
SPDR S&P 500 (SPY)

To investigate, we compare 22 variations of each strategy with execution days ranging from end-of-month (EOM) per the baseline strategy to 21 trading days after EOM (EOM+21). For example, an EOM+5 variation computes allocations based on EOM but delays execution until the close five trading days after EOM. We include a benchmark that each month allocates 60% to SPY and 40% to TLT (60-40) to see whether variations are unique to SACEVS. We focus on gross compound annual growth rate (CAGR), maximum drawdown (MaxDD) and annual Sharpe ratio as key performance statistics. Using daily dividend-adjusted closes for the above ETFs from the end of July 2002 through January 2020, we find that:

Keep Reading

Are there any seasonal, technical or fundamental strategies that reliably time the U.S. stock market as proxied by the S&P 500 Total Return Index? In the February 2018 version of his paper entitled “Investing In The S&P 500 Index: Can Anything Beat the Buy-And-Hold Strategy?”, Hubert Dichtl compares excess returns (relative to the U.S. Treasury bill [T-bill] yield) and Sharpe ratios for investment strategies that time the S&P 500 Index monthly based on each of:

• 4,096 seasonality strategies.
• 24 technical strategies (10 slow-fast moving average crossover rules; 8 intrinsic [time series or absolute] momentum rules; and, 6 on-balance volume rules).
• 18 fundamental variable strategies based on a rolling 180-month regression, with 1950-1965 used to generate initial predictions.

In all cases, when not in stocks, the strategies hold T-bills as a proxy for cash. His main out-of-sample test period is 1966-2014, with emphasis on a “crisis” subsample of 2000-2014. He includes extended tests on seasonality and some technical strategies using 1931-2014. He assumes constant stock index-cash switching frictions of 0.25%. He addresses data snooping bias from testing multiple strategies on the same sample by applying Hansen’s test for superior predictive ability. Using monthly S&P 500 Index levels/total returns and U.S. Treasury bill yields since 1931 and values of fundamental variables since January 1950, all through December 2014, he finds that:

Keep Reading

Does the market react most strongly to the earliest quarterly earnings announcements? In their October 2019 paper entitled “Calendar Rotations: A New Approach for Studying the Impact of Timing using Earnings Announcements”, Suzie Noh, Eric So and Rodrigo Verdi study effects of the relative order of U.S. firm quarterly earnings announcements, which vary systematically for some firms according to the day of the week of the first day of a month. Specifically, they qualify firms by identifying those firms that exhibit systematic earnings announcement schedules (such as Friday of the fourth week after quarter ends, sometimes set in firm bylaws) for at least four consecutive same fiscal quarters. They then for each firm each fiscal quarter:

• Calculate EA Order, ranking of earnings announcement date divided by number of firms with the same fiscal quarter-end.
• Compute change in EA Order compared to the same fiscal quarter last year, indicating a calendar acceleration or delay in announcement. Positive (negative) change in EA Order indicates delay (acceleration)
• Examine effects of change in EA Order on media coverage (number of articles), stock trading volume and stock return from one trading day before to one trading day after earnings announcement.

Using sample of 76,622 firm-quarters during 2004 through 2017, they find that: Keep Reading

“Stock Market and the National Election Cycle” examines the behavior of the U.S. stock market across the U.S. presidential term cycle (years 1, 2, 3 or 4) starting in 1950. Is a longer sample informative? To extend the sample period, we use the long run S&P Composite Index of Robert Shiller. The value of this index each month is the average daily level during that month. It is therefore “blurry” compared to a month-end series, but the blurriness is not of much concern over a 4-year cycle. Using monthly S&P Composite Index levels from the end of December 1872 through August 2019 (about 37.5 presidential terms), we find that:

Keep Reading

A subscriber commented and asked:

“Some have suggested that the end-of-the-month effect benefits monthly simple moving average strategies that trade on the last day of the month. Is there an optimal day of the month for long-term SMA calculation and does the end-of-the-month effect explain the optimal day?”

To investigate, we compare 21 variations of a 10-month simple moving average (SMA10) timing strategy generated by shifting the monthly return calculation cycle relative to trading days from the end of the month (EOM). Specifically, the 21 variations represent calculation cycles ranging from 10 trading days before EOM (EOM-10) to 10 trading days after EOM (EOM+10). We apply the strategy to the S&P 500 Index as a proxy for the U.S. stock market. The strategy holds the S&P 500 Index (cash) whenever the index is above (below) its SMA10 as of the most recent monthly calculation. Using daily S&P 500 Index closes and 3-month Treasury bill (T-bill) yields as the return on cash during January 1990 through mid-June 2019, we find that: Keep Reading

“Optimal Monthly Cycle for SACEMS?” investigates whether using a monthly cycle other than end-of-month (EOM) to pick winning assets improves performance of the Simple Asset Class ETF Momentum Strategy (SACEMS). This strategy each month picks winners from the following set of exchange-traded funds (ETF) based on total returns over a specified lookback interval:

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

In response, a subscriber asked whether sticking with an EOM cycle for determining the winner, but delaying signal execution, affects strategy performance. To investigate, we compare 23 variations of SACEMS portfolios that all use EOM to pick winners but shift execution from the contemporaneous EOM to the next open or to closes over the next 21 trading days (about one month). For example, EOM+5 uses an EOM cycle to determine winners but delays execution until the close five trading days after EOM. We focus on gross compound annual growth rate (CAGR) and maximum drawdown (MaxDD) as key performance statistics for the Top 1, equally weighted (EW) Top 2 and EW Top 3 portfolios of monthly winners. Using daily dividend-adjusted opens and closes for the asset class proxies and the yield for Cash during February 2006 (limited by DBC) through January 2019, we find that: Keep Reading

Does persistence in the difference in direction between overnight stock trading and intraday stock trading behaviors (tug of war) predict future returns? In their January 2019 paper entitled “Overnight Returns, Daytime Reversals, and Future Stock Returns: The Risk of Investing in a Tug of War with Noise Traders”, Ferhat Akbas, Ekkehart Boehmer, Chao Jiang and Paul Koch investigate relationships between intensity of the daily tug-of-war between between overnight (noise) and intraday (other) stock traders and future stock returns. They specify tug-of-war intensity as percentage of trading days during a month for which a stock exhibits negative (or positive) daytime reversals divided by average monthly percentage of negative (or positive) reversals over the last 12 months. They then examine whether either negative or positive tug-of-war intensity predicts future stock returns. Using overnight/intraday stock returns for a broad sample of U.S. common stocks, along with monthly returns for widely accepted factors, during May 1993 through December 2017, they find that:

Keep Reading

Do very old data confirm reliability of widely accepted asset return factor premiums? In their January 2019 paper entitled “Global Factor Premiums”, Guido Baltussen, Laurens Swinkels and Pim van Vliet present replication (1981-2011) and out-of-sample (1800-1908 and 2012-2016) tests of six global factor premiums across four asset classes. The asset classes are equity indexes, government bonds, commodities and currencies. The factors are: time series (intrinsic or absolute) momentum, designated as trend; cross-sectional (relative) momentum, designated as momentum; value; carry (long high yields and short low yields); seasonality (rolling “hot” months); and, betting against beta (BAB). They explicitly account for p-hacking (data snooping bias) and further explore economic explanations of global factor premiums. Using monthly global data as available during 1800 through 2016 to construct the six factors and four asset class return series, they find that:

Keep Reading

Does choice of multi-asset portfolio rebalance date(s) materially affect performance? In their October 2018 paper entitled “Rebalance Timing Luck: The Difference Between Hired and Fired”, Corey Hoffstein, Justin Sibears and Nathan Faber investigate effects of varying portfolio rebalance date on performance. Specifically, they quantify noise (luck) from varying annual rebalance date for a 60% S&P 500 Index-40% 5-year constant maturity U.S. Treasury note (60-40) U.S. market portfolio. Using monthly total returns for these two assets during January 1922 through June 2018, they find that: Keep Reading

As a follow-up to “Sunspot Cycle and Stock Market Returns” a reader asked: “Sunspot activity does have a direct relationship to weather. Could one speculate on the agriculture market using the sunspot cycle?” To investigate, we relate sunspot activity to the fairly long U.S. Producer Price Index (PPI) for grains. Using monthly averages of daily sunspot counts and monthly PPI for grains during January 1926 (limited by PPI data) through October 2018, we find that: Keep Reading