Kaeppel's Corner: The Best of Times, the Worst of Times for T-Bonds (Part II)
Jay Kaeppel, Optionetics.com
March 8, 2011
Last week I wrote the best and worst months historically for T-Bonds. This week, let’s “drill down” a little deeper and take a look at the individual trading days of the month and see if we can discern anything useful (Plot Warning: We can!).
The Best and Worst Months
Taken from last week’s piece, Figure 1 displays the gross dollar return achieved by holding a long position of one T-Bond futures contract (each point in price movement is worth $1,000 – so if you are long and the price goes up two points then you make $2,000) only during a given month every year since November 1977.
As one would expect, some months show gains while others show losses. No surprise there. The key however, is the consistency with which a given time period shows a gain or a loss. As it turns out, that is true for individual trading days within a month as well as individual months as a whole. But more on that in a moment.
Figure 1 – T-Bond futures Monthly Gain or Loss (1977-2011)
For our purposes we will make the following designations:
-Bullish Months = May, June, August and November
-Bearish Months = January, February, March and April
Now let’s turn our attention actual trading days of the month
The Best and Worst Trading Days of the Month
Note that in this discussion we are talking about “trading” days each month, not just “calendar”. Most months have somewhere between 19 and 21 trading days, with an occasional month with 22 trading days.
For our purposes we will designate three bullish periods each month:
-Trading Days 1, 2 and 3
-Trading Days 10, 11 and 12
-Trading Days 17, 18, 19, 20, 21 and 22
For our purposes we will designate two bearish periods each month:
-Trading Days 4 through 9
-Trading Days 13 through 16
Figure 2 displays the performance achieved by holding a long position in T-Bond futures during each of the five periods above.
Figure 2 – Gain or Loss from long T-Bond futures position during certain trading day of month periods (1977-2011)
A Combined Model (JSBM1)
So now let’s create a unified timing model using the information above. I refer to this model as Jay’s Seasonal Bond Model #1 (or mercifully, JSBM1) for short. To calculate the model we simply add or subtract point values as follows:
-If the current month is May, June, August or November add one point
-If the current month is January through April subtract one point
-If the month is April, July, September or December simply enter a zero
-If today is trading day 1 through 3, 10 through 12 or 17 through 22 for the current month then add one point
-If today is trading day 4 through 9 or 13 through 16 for the current month then subtract one point
As a result of all of these simple calculations, on any given day the model can read -2, -1, 0, +1 or +2. Does today’s JSBM1 model reading really matter? Consider the long-term results.
The Long-Term Results
Figure 3 displays the results achieved by holding a long position in T-Bonds when the model reads -2 or -1.
Figure 3 – T-Bond futures performance (assuming a long position of one futures contract) on days when JSBM1 equals -2 or -1 (1977-2011)
Figure 4 displays the results of holding a long position in T-Bonds when the model reads +2.
Figure 4 – T-Bond futures performance (assuming a long position of one futures contract) on days when JSBM1 equals +2 (1977-2011)
So now let’s take the final step:
-When the model reads +2 we will hold a long position in T-Bond futures
-When the model is less than 0 we will hold a short position in T-Bond futures
The hypothetical results achieved using this approach appear in Figure 5. The results are fairly compelling.
Figure 5 – JSBM1 performance (long when JSBM1 equals +2 and short when JSBM equals -1 or -2) (1977-2011)
Clearly the long term results have been fairly consistent. Investors who are not inclined to trade futures contracts might consider two ETFs - ticker TLT (which tracks the long bond) and ticker TBT (which tracks the inverse of the long bond, i.e., buying TBT is like selling short T-Bonds). Both TLT and TBT are bought and sold like shares of stock in a stock account).
As with any market, on any given day there a plenty of reasons that an individual might choose to justify a bullish or bearish position. Understandably, for most people, reasons like “because it’s the last trading day of May”, isn’t one of them.
Which – based on the results displayed in Figure 5 – would seem to be a pity.
Staff Writer and Author of “Seasonal Stock Market Trends”
Optionetics.com ~ You’re Options Education Site