Days before Thanksgiving last year, fixed-income strategists at Morgan Stanley noticed something intriguing.
Treasurys were rallying even though their yields, which typically move in the opposite direction, were broadly trending higher. The strategists’ theory was that traders who had bet against Treasurys were covering their so-called shorts ahead of the long weekend, as holding on to them could be costly if there were a major market-moving event.
And so observing this price action, they studied Treasury futures every Thanksgiving and around every major US holiday.
“We thought there might be scope to combine observed seasonality around each holiday into a potentially uncorrelated, rules-based annual trading strategy,” Tony Small and Matthew Hornbach said in a note on Tuesday. “Such a strategy could either act as a timing tool or a standalone supplement to discretionary trading.”
The cliché but absolutely necessary caveat here is that past returns don’t guarantee future results.
“While history seldom repeats itself, it may rhyme enough for this strategy to offer investors an idea about how to tactically trade around US holidays,” Small and Hornbach said. “In addition, this study opens the door for further analysis around non-US holidays given price patterns may exist in overseas bond futures markets as well as other asset classes.”
They studied all US holidays during which the bond market is closed and most investors are away from their desks, including Memorial Day, Independence Day, Labor Day, Thanksgiving, Christmas, and New Year’s Day.
Using 10-year futures prices in the five business days before and after each holiday, the strategists initiated and closed long and short strategies on each of those days and compared them with one another — for example, a long strategy opened five days before the holiday and closed three days after, or one opened four days before and closed a day after.
From this, they selected strategies with a win-to-loss ratio of at least 60%. All trades occurred at 3 p.m. ET or earlier if the market closed early. It assumed the same number of futures contracts for every holiday every year.
The strategy returned 2.22% annually on average since its inception in 1983 and was positive every year after 1987. Its best year was 2008, at 7.33%.
From 1995 to 1996, the strategy lost money around six consecutive holidays, the longest streak.
Below is a breakdown of the trading rules:
The strategists acknowledged that market drivers changed over time but offered some ideas on why these rules worked with some consistency.
For example, it has been best to short 10-year Treasury futures before Christmas. It also so happens that central banks, among the biggest buyers of Treasurys, tend to shut down their purchases before the end of the year and resume early in January. Also, investors may be selling Treasurys to fulfill year-end liabilities. And as buying resumes in full force in January, the model shows it has been profitable to go long after New Year’s Day.
Morgan Stanley recommends buying after Memorial Day, saying that has returned 1% or more every time the holiday fell on May 28 (which it will this year) or later. The strategists suggested that month-end rebalancing or reinvestments of large quarterly refunding coupons may be working in the market’s favor.
“Given the risk of data mining, investors should accept this analysis with some degree of caution,” Small and Hornbach said. “There is no guarantee that historical seasonality and past price performance will continue in the future.”
Leave a Reply
Leave a Reply
This site uses Akismet to reduce spam. Learn how your comment data is processed.
