Jun

19

On average, in the last 28 years bonds have begun a short-term rally on the 9th trading day of June (6/13 this year). This is a trade I have taught in seminars and taken for the last 10 years. I don't have a clue as to why this happens — maybe readers have some ideas. Perhaps it's just happenstance. But next year I will take it again just as I have in the past.

Kim Zussman replies:

I tested the suggestion of a bond rally from 9th trading day of June using TNX (10 yr bond yield) back to 1980. From the 9th day, holding for 5d, 10d, 20d, here are the results:

One-Sample T:

Variable     N   Mean     StDev        95% CI         T       P
5d           27  0.0006  0.0239  (-0.009, 0.010)  0.13  0.899
10d          27  0.0041  0.0364  (-0.010, 0.019)  0.59  0.561
20d          27  -0.002  0.0500  (-0.021, 0.018) -0.19  0.849

Using TNX, there were not significant gains in 10yr bonds for the three holds. Note that TNX is yield, so increased bond price would be decreased TNX.

Here are the data:

Date        5d     10d     20d
6/13/2006  0.040  0.050  0.028
6/13/2005  0.002 -0.046  0.012
6/14/2004 -0.037 -0.027 -0.080
6/12/2003  0.054  0.114  0.148
6/13/2002 -0.031 -0.018 -0.059
6/13/2001 -0.011 -0.006 -0.006
6/13/2000 -0.015 -0.003 -0.003
6/11/1999 -0.030  0.002 -0.047
6/11/1998  0.011  0.006 -0.004
6/12/1997 -0.012  0.003 -0.037
6/13/1996 -0.006 -0.026 -0.024
6/13/1995  0.003  0.002 -0.005
6/13/1994  0.014  0.014  0.051
6/11/1993  0.000 -0.022 -0.037
6/11/1992 -0.019 -0.026 -0.055
6/13/1991 -0.005 -0.002 -0.010
6/13/1990  0.018  0.014  0.012
6/13/1989  0.007 -0.017 -0.030
6/13/1988  0.012  0.003  0.019
6/11/1987 -0.029 -0.032 -0.023
6/12/1986 -0.036 -0.070 -0.085
6/13/1985 -0.007  0.018  0.002
6/13/1984  0.016  0.024 -0.001
6/13/1983  0.005  0.033  0.062
6/11/1982  0.050  0.060 -0.004
6/11/1981  0.019  0.029  0.047
6/12/1980  0.003  0.036  0.079

Jeff Rollert remarks:

I'd suggest not using TNX, especially when there are auctions — it doesn't handle those well in the series. I use TNX only for checking on general levels when using my cellphone browser at school functions.

Larry Williams replies:

Dr. Zussman's comments on my bond seasonal trade were confusing to many readers, so here are the details: What I tested was buying bond futures on the open of the 9th trading day of June, using a $1,600 stop and exit on the first profitable opening plus two ticks. In the last 21 years there have been 20 winning trades. That's how I've traded it for the last 10 years. Dr. Zussman tested 10 year cash bonds; I trade 30 year nearby futures — apples and oranges. Hope this clarifies.

Larry Williams continues:

My tests, using nearby active contract in Genesis Data, with no stop, exit X days later:

 5 days  52% wins      + 3,270
10         63%              + 6,926
15         79%              + 8,301
20         63%             +16,676
 


Comments

Name

Email

Website

Speak your mind

3 Comments so far

  1. Kelvin on June 19, 2007 2:30 pm

    I never get a chance to thank Larry. I hope he will read this post.
    I bought his books and for the record his exit strategy is well documented in his book. I never went to his lessons but I saw his post here and did a simple backtest and bought and I made money using rydex bond fund (yes I used first profitable close not open) but as Larry said, you got to try out your own variations…

  2. Keith on June 19, 2007 3:15 pm

    Once again, you can get what you want out of statistics; however you want it. My testing is REAL Time Only. Of course I am short time frame trader. Nice Call, Real Time, Larry. your Seasonals supported my other work.

  3. Ed Kelenyi on June 20, 2007 12:36 pm

    I too wish to add a note of thanks and respect to Larry - I have purchased almost all of his books and they have always provided valid ideas and strategies from which to work from.

Archives

Resources & Links

Search