No. of Recommendations: 17
rharmelink wrote:
Ditto, if you don't think significantly different results based on varying the starting date would indicate a problem.

Alright, I'm more than open to being educated on how to do an all-starts-test on a timing system.

On December 1st this year I shared this chart of one of my indicators. It is an indicator I developed quite a few years ago, and have been using ever since in my trading.

The strategy that I use to trade this is based upon:

1. Buy on an Extremely Bullish Signal (ignore new signals within one month)
2. Screen for the optimal 3x ETF to buy
3. Set a Limit Order 2.5% below the signal day's close price - if the order fails to be filled buy at market close
4. Sell when a fixed PROFIT LATCH is reached, or three months out from the signal, whichever comes first.

The signals from October of 2011 to present are:

1. 10/4/2011
2. 11/21/2011
3. 5/23//2012
4. 11/16/2012

Here are the results based on a variety of PL levels:

Profit Latch    ROI      CAGR    GSD   Sharpe  Ulcer Index  DDD3     BBW   Signal  % of Time  Drawdown
20.00% 107.67% 84.45% 39.24 2.13 1.98% 0.15% 444.19 100% 59.28% -13.70%
25.00% 144.50% 111.48% 43.52 2.22 2.00% 0.15% 586.38 100% 59.28% -13.70%
30.00% 182.72% 138.84% 46.65 2.37 2.05% 0.15% 728.94 100% 59.28% -13.70%
35.00% 183.39% 139.32% 51 2.24 4.21% 0.28% 390.58 100% 59.28% -13.70%
40.00% 204.75% 154.34% 52.62 2.34 4.25% 0.28% 431.88 100% 59.28% -13.70%
45.00% 226.90% 169.74% 53.35 2.45 4.25% 0.28% 474.96 100% 59.28% -13.70%
50.00% 249.83% 185.50% 53.72 2.57 4.25% 0.28% 519.07 100% 59.28% -13.70%
55.00% 273.54% 201.62% 54.22 2.68 4.25% 0.28% 564.18 100% 59.28% -13.70%
60.00% 298.02% 218.09% 55.53 2.76 4.26% 0.28% 610.27 100% 59.28% -13.70%

For the sake of illustration let me show how this turned out for the first buy signal on 10/4/2011. The 3x ETF selected was DRN. I set a limit order 2.5% below the close price of 10/4 and it was filled on 10/5. With a 60% PL the gain was reached on 10/27 and sold. The PL is also set using a limit order shortly after the order was filled on the day of purchase.

Now, my question to you is how do you do an All Starts Day test on this sort of system?

Now it may be that you'll say this system does not require a rebalance, while the other system does. But again, what you need to realize is that the rebalance makes very little difference in the ETF Pair system. You are simply redistributing the funds equally between two ETFs. What day that happens is meaningless. The fact that there is so little difference between two to eight weeks proves that out.

An All Start Days test is VERY important when you are dealing with fundamental data, but impossible with a test like I'm illustrating in this post, and irrelevant to an ETF Pair test.

More rules. So you reported the results you achieved, but not the actual method you used. So no one else can possibly repeat the test.

I share the details when I see a genuine interest. Plus, as you discovered with simply holding the pairs indefinitely, the actual details of this are also meaningless. Any method would work. Pairs held short do best during hight volatility periods in the market, and not so well during calmer periods. Whatever method you want to use to identify that will work.

What surprises me is that you traded without checking other levels than the VXO at 19. To invest money without checking the brittleness of the cutoff point seems that I am missing something.

Wouldn't the VXO level be a key criteria in the backtest?

Test it and you'll discover that any method will work. I simply chose this one. It is not central to the system.
Print the post  


What was Your Dumbest Investment?
Share it with us -- and learn from others' stories of flubs.
When Life Gives You Lemons
We all have had hardships and made poor decisions. The important thing is how we respond and grow. Read the story of a Fool who started from nothing, and looks to gain everything.
Community Home
Speak Your Mind, Start Your Blog, Rate Your Stocks

Community Team Fools - who are those TMF's?
Contact Us
Contact Customer Service and other Fool departments here.
Work for Fools?
Winner of the Washingtonian great places to work, and Glassdoor #1 Company to Work For 2015! Have access to all of TMF's online and email products for FREE, and be paid for your contributions to TMF! Click the link and start your Fool career.