Switch Method Adopted
Combining Mostly USA & Worldwide Systems Creates
Low Risk, Higher Returns Than Trading Either Alone
7.9 years June 28, 2002--May 7, 2010
Mostly USA ETFs & Worldwide ETFs Portfolios
Hypothetical--Weekly
Statistical Footnote below
ADOPTION & RULES
The 'Switch' Method is officially adopted as of the market close Friday, May 14, 2010, price 1030.57. Real-time returns will be reported from that date and price forward.
The method was first created and described four weeks ago in the 'Systems Tips' biweekly letter of April 16, 2010.
TABLE OF HISTORICAL SWITCH TRADES
Mostly USA & Worldwide ETFs Portfolios
Hypothetical
Eight Years 2002-2010
Statistics & Notes below
|
switch |
|
% |
|
|
wks |
|
Date |
to |
price |
profit |
%P |
%L |
held |
|
6/28/2002 |
WW |
100 |
16.1 |
16.1 |
|
5 |
|
8/2/2002 |
USA |
116 |
5.6 |
5.6 |
|
3 |
|
8/23/2002 |
WW |
123 |
13.5 |
13.5 |
|
6 |
|
10/4/2002 |
USA |
139 |
10.1 |
10.1 |
|
8 |
|
11/29/2002 |
WW |
153 |
-6.2 |
|
-6.2 |
3 |
|
12/20/2002 |
USA |
144 |
11.5 |
11.5 |
|
2 |
|
1/3/2003 |
WW |
160 |
3.8 |
3.8 |
|
8 |
|
2/28/2003 |
USA |
167 |
82.5 |
82.5 |
|
52 |
|
2/27/2004 |
WW |
304 |
-6.9 |
|
-6.9 |
3 |
|
3/19/2004 |
USA |
283 |
3.5 |
3.5 |
|
3 |
|
4/8/2004 |
WW |
293 |
-8.0 |
|
-8.0 |
5 |
|
5/14/2004 |
USA |
270 |
11.5 |
11.5 |
|
5 |
|
6/18/2004 |
WW |
301 |
1.3 |
1.3 |
|
8 |
|
8/13/2004 |
USA |
305 |
15.9 |
15.9 |
|
7 |
|
10/1/2004 |
WW |
353 |
-1.9 |
|
-1.9 |
2 |
|
10/15/2004 |
USA |
346 |
-5.0 |
|
-5.0 |
11 |
|
12/31/2004 |
WW |
329 |
-2.8 |
|
-2.8 |
2 |
|
1/14/2005 |
USA |
320 |
5.2 |
5.2 |
|
7 |
|
3/4/2005 |
WW |
337 |
-6.0 |
|
-6.0 |
8 |
|
4/29/2005 |
USA |
316 |
17.3 |
17.3 |
|
14 |
|
8/5/2005 |
WW |
371 |
1.3 |
1.3 |
|
3 |
|
8/26/2005 |
USA |
376 |
3.0 |
3.0 |
|
3 |
|
9/16/2005 |
WW |
387 |
0.6 |
0.6 |
|
5 |
|
10/21/2005 |
USA |
390 |
19.7 |
19.7 |
|
15 |
|
2/3/2006 |
WW |
466 |
-3.2 |
|
-3.2 |
1 |
|
2/10/2006 |
USA |
451 |
7.3 |
7.3 |
|
14 |
|
5/19/2006 |
WW |
484 |
-0.5 |
|
-0.5 |
11 |
|
8/4/2006 |
USA |
482 |
1.4 |
1.4 |
|
31 |
|
3/9/2007 |
WW |
488 |
0.0 |
|
0.0 |
4 |
|
4/6/2007 |
USA |
488 |
7.6 |
7.6 |
|
16 |
|
7/27/2007 |
WW |
525 |
7.2 |
7.2 |
|
8 |
|
9/21/2007 |
USA |
563 |
-0.1 |
|
-0.1 |
6 |
|
11/2/2007 |
WW |
563 |
11.6 |
11.6 |
|
24 |
|
4/18/2008 |
USA |
628 |
0.6 |
0.6 |
|
7 |
|
6/6/2008 |
WW |
632 |
1.7 |
1.7 |
|
10 |
|
8/15/2008 |
USA |
642 |
1.3 |
1.3 |
|
3 |
|
9/5/2008 |
WW |
651 |
27.7 |
27.7 |
|
17 |
|
1/2/2009 |
USA |
831 |
6.8 |
6.8 |
|
2 |
|
1/16/2009 |
WW |
887 |
0.7 |
0.7 |
|
11 |
|
4/3/2009 |
USA |
894 |
7.7 |
7.7 |
|
42 |
|
1/22/2010 |
WW |
963 |
-4.6 |
|
-4.6 |
4 |
|
2/19/2010 |
USA |
919 |
12.2 |
12.2 |
|
11 |
|
5/7/2010 |
USA |
1031 |
open |
|
|
|
|
|
|
|
|
|
|
|
|
|
sum |
316.4 |
-45.1 |
409 |
|
|
|
avg |
10.5 |
-3.8 |
10 |
Statistical Footnote
Compound average return 34.6% per year
Position still open as of 5/14/2010
Average holding period 10 weeks
Average profit per trade 10.5%
Average loss per trade -3.8%
Win/Lose $$ ratio 7 to 1
Profit factor 7.0
Biggest win 82.5%
Largest loss -8.0%
Worst drawdown -20%*
Average 5 switches per year
Longest single switch 52 weeks
(2/28/2003, Mostly USA)
*Outcomes are impervious to drawdowns,
but are included here for those who would like
to see how much they would have been worried
had they been on board.
Note the behavior profiles during the two bear-market periods, the first during late 2002-early 2003, the second marked by the magenta bars at the bottom and top of the chart. While the market tanked in each instance,
-56% in the recent crash, the switch-system results curve not only did not drop, it rose strongly and exploded upward in 2008.
If you choose to hold cash during bear markets, your equity-curve profile would look something like the chart here.
PROJECTED WEALTH COMPARISON
Switch System vs
SP500 & Treasury Bills
-- Dollars --
Hypothetical
|
Years |
Tbills |
SP500 |
Switch |
|
start |
10,000 |
10,000 |
10,000 |
|
5 |
11,124 |
11,877 |
44,121 |
|
10 |
12,374 |
14,106 |
194,666 |
|
15 |
13,765 |
16,754 |
858,885 |
|
20 |
15,312 |
19,898 |
3,789,481 |
|
25 |
17,033 |
23,633 |
16,719,552 |
|
30 |
18,948 |
28,069 |
73,768,264 |
Comment. The first two rows in the table are creditable, believably possible. With the rows below, belief wears thin and thinner until it perhaps disappears. More on this in a moment.
The annual rates of return are 2.2, 3.5, and 34.6% for Treasury Bills, the S&P 500, and the Switch system respectively. The multiple-year rows are the results of compounding. Switch doubles money every 2.3 years on average. Its exploding numbers over the advancing years seem amazing, but I think no more so than much of what else goes on in the Universe.
Do you believe the table above can happen? Now there's a question! Before reading further, please give me your answer to the question via email here. Just type in yes or no, and I will let you know the results.
In fact, you may stop right here. The rest is a self-accusatory example of paradox. There is always, it seems, another word to be said about everything. The goal of this website is, with simplicity, brevity, and ease of use, to make a lot of money--with good risk control. And here I am, combining two simple systems into one--more complex than either before
. . . and still wanting to get one more last word in.
OPTIONAL
A RIFF ON HYPOTHETICAL
To pick up where we left off, Do I believe the table above can happen? No, I do not. Nor do I believe that it cannot happen. If you gave a 'yes' or 'no' answer, that was not a trick question (fond of use by poll takers and disseration writers--discussion for another day). There is no yes/no answer. It is not possible.
My own first answer was 'no'. But even the most casual reflection reveals that nobody knows whether the numbers in the table will fall short or will be exceeded or come out exactly the same as recorded here. I know of no human able to prove any single one of the three possibilities.
Financial consumer-protection agencies make haste to oblige financial practitioners and professionals to forewarn customers and clients of the dangers of the 'hypothetical' performance results. With good cause. There is usually no fraud nor intent of fraud. But what follows--for many--is disappointment--sometimes severe enough to seek recourse in litigation.
The shortcoming is inadequate understanding of potential outcomes by clients or customers. I believe a more full understanding is the reponsibility of the vendor or provider. Every discipline has its own back story. The failure lies in an inadequate presentation of what causes the disappointments. Those causes are perhaps nearly always psychological. Examples in a moment.
All of known history is hypothetical. If you were not present as a participant, the event described is srictly hypothetical for you--as it is also for the historian-author who has writtten the history you read.
So it is with any historical record of stock-market trades. The only parties for whom the transaction is not hypothetical are the actual buyer and seller executing the purchase and the sale which together are a single, simulataneous, factual event. Its recordation becomes an artifact with a semireality of its own that enables its analytical manipulation in company with other related artifacts in the neighborhood or the sequence.
One stumbling block is sensitive dependence on initial conditions--that is, a tiny change at the beginning can result in monstrous change at the end. The ending change can be malign or benign.
Initial conditions, as you may know from my other writings, start from a date and a price both of which are random and arbitrary. Given the looseness of beginnings, can by one be surprised by the potential wildness of the endings in either direction?--favorable or unfavorable.
To reduce the psychological effect of monstrosity at the end (there is no end--that's another point worth keeping in mind), one can segregate a single, continous stream of sequential data into clumps each of which has its own beginning and end. Then take an average all the clumps. The smaller the clumps, and the more there are of them, the more singular and apt will be the single number derived to describe the entire sequence of all the given data.
Thus, in the first table beneath the chart above, the standard deviation of the data in the %P column is more than twice the average profit per trade. That's a lot. When I separate all weeks in the sequence (420) into packets of 52 each and roll them forward one week at a time, I wind up with rolling one-year (52 weeks)
performance results 360 times. I average these all together and come up with single average one-year performance number. That's a better, more realistic depiction of the Switch system's past results.
This can be done with any number of benchmark funds or indexes for comparison. I am going to show you what this looks like. In viewing the next chart, keep in mind that if you had owned this portfolio during the past eight years, you would be riding the roller coaster you are about to see. You would have experienced a peak profit of 138% and a worst loss of -9.6%. During one period the Switch underperformed the market for eight consecutive months. The advantage of knowing all this is that you will not be suprised or distressed and can expect the upward winning cycle to recur.
The average rolling one-year return during the entire period
has been 32.7%. This is close enough to validate the performance numbers cited in table and the chart. Viewing the current picture (next) looks like a good time to be getting into the Switch portfolios and out of the market.
HYPOTHETICAL
Conclusion
here
Posted
5/16/2010 10:25 p.m. EDT
|
Back
| Subscribe--45-day trial Free
|
|
tip79a
© 2010 The 2000 Corporation.
All Rights Reserved.
| |