Bonds are to Fear Early in the Year

Jan 6, 2014 | Uncategorized

The treasury bond market has showed a strong seasonal tendency to perform poorly during the early part of the year.  People often ask me “why” this would be so.  In fact I get that question often enough to make me wish I had a good answer.  Alas, as a proud graduate of “The School of Whatever Works”, I can only repeat our school motto, which is “Whatever!”

Two Early Year Trends in T-Bonds: Part 1

First let’s look at the performance of t-bond futures between the end of the first trading day of the new year and the 14^th trading day of February starting in 1978.  Figure 1 displays the performance achieved by an (extremely stubborn and not terribly astute) investor who held a long position in t-bond futures during this time period every year.

 Figure 1 – Long t-bond futures from on January Trading Day 1 through February Trading Day 14 (1978-present)

All told, the loss came to -$49,511 (excluding any slippage and/or commissions).  Of course, like all seasonal trends there is never any guarantee that the trend will hold true the next time around.  For the record the Jan TD 1 through Feb TD 14 period saw T-bonds:

– Gain 10 times
– Lose 25 times
– Breakeven 1 time

Each point movement in t-bond futures is worth $1,000

– The median gain during up years was +$2,234
– The median loss during down years was -$2,406
– The largest gain was $6,937 in 2000.
– The largest loss was -$15,281 in 1980.

So basically, t-bonds gained 28% of the time, and lost or broke even 72% of the time, and the median loss was slightly greater than the median gain.

Two Early Year Trend in T-Bonds: Part 2
Let’s look next at the net performance for t-bonds during the first four months of the calendar year.  Typically, after bonds sink into mid-February there is a bounce in the second half of February.  But for our test we will just consider the results achieved by holding a long position in t-bonds from December 31^st each year through the end of April.  These results appear in Figure 2.

 Figure 2 – Long t-bond futures December 31^st through April 31st

All told, the loss came to -$66,389 (excluding any slippage and/or commissions).  Of course, like all seasonal trends there is never any guarantee that the trend will hold true the next time around.  For the record the Dec 31^st to Apr 30th period saw T-bonds:

– Gain 16 times
– Lose 20 times

Each point movement in t-bond futures is worth $1,000

– The median gain during up years was +$1,797
– The median loss during down years was -$4,813
– The largest gain was $13,968 in 1986.
– The largest loss was -$11,313 in 1994.

In reality, the January through April time frame has seen t-bonds show a loss only 56% of the time.  So this trend is absolutely by no means a sure thing, so the one thing you should absolutely not do is get it in your head that t-bonds are bound to decline between now and the end of April.

The key thing to note regarding this trend is that the median “down” year has witnessed a decline that is 2.7 times larger than the median gain shown during the “up” years.  So the key is simply to recognize the potential danger.

Summary

With t-bonds presently quite oversold, it is a little difficult to jump on the bearish bandwagon at the moment (in fact, bonds are rallying nicely as I write here on the first trading day of the year).  And as I have tried to make clear, a decline in t-bond prices during either of both of the highlighted periods is by no means a sure thing.  Still, this little bit of history suggests that getting wildly bullish on t-bonds may not be the best strategy.

Jay Kaeppel

Wow, It Really is the Most Wonderful Time of the Year!

Dec 26, 2013 | Uncategorized

Some seasonal trends have shown a tendency to persist through time (hence the use of the word “trend”, I guess).  As it turns out we are at the cusp of one of “those times” right now.  It is sitting there like a wrapped gift under the tree with our name on it – so let’s not waste any time diving in.

December-January Changeover

The period we will look at encompasses the last 4 trading days of December and the first 3 trading days of the following January.  In other words, a contiguous 7 day trading period during which the stock market has showed a tendency to behave in a bullish manner.

Now given the persistence of the recent market run up, many may be a little leery of diving in here.  Which I understand.  Still, the numbers are what they are, so let’s take a look.

The Test

So as not to make it easy on ourselves, this test begins in December 1933, i.e., in the early days off the great Depression.  We will buy the Dow Jones industrials Average at the close of the fifth to last trading day of the year and sell at the close of the third trading day of January. This test assumes no interest is earned while out of the market so that we measure only the performance during the supposedly bullish period.

The Results

Figure 1 displays the growth of $1,000 invested in the Dow every year since 1933 during the seven trading days just described.

Two anecdotal comments from a quick perusal of the graph in Figure 1:

-There is clearly a lower left to upper right trend, which is what we want to see in any equity curve
-It is by no means “perfect”, so a little closer analysis of the numbers may be useful in convincing ourselves that this trend might actually be useful.  So in order to gain some perspective, let’s compare the performance of the Dow during this time period versus Dow performance for all trading days.
A few figures of note:

-System average daily performance is +0.22% versus +0.03% for all trading days (7.53 times greater).
-System median daily performance is +0.17% versus +0.03% for all trading days (4.00 times greater).
-338 out of 560 system trading days showed a gain (60.4%).
-10,946 out of all 20,922 trading days showed a gain (52.3%).
-Average 7-day return only during system days = +1.55%.
-Average 7-day return for all trading days = +0.20%.
-The 7-day system period has showed a gain in 62 of the past 80 years (or 77.5% of the time)
One other thing to note is that returns (and albeit risk) is enhanced by trading leveraged funds such as ticker UDPIX (Profunds UltraDow) or UDOW (ProShares UltraDow30 ETF).

 Summary

So is the Dow destined to be higher at the close on January 6, 2014 than it was at the close on December 24^th, 2013?  Not necessarily.  But that would seem to be the way to bet.

Jay Kaeppel