Hi:
Fri's results:
[ATTACH=CONFIG]19565[/ATTACH]
For Mon, we are tracking:
[ATTACH]19566[/ATTACH]
Shawn
Printable View
Hi:
Fri's results:
[ATTACH=CONFIG]19565[/ATTACH]
For Mon, we are tracking:
[ATTACH]19566[/ATTACH]
Shawn
Hi:
Results for Mon, Aug 19, 2013:
[ATTACH=CONFIG]19570[/ATTACH]
For Tu, Aug 20, 2013:
[ATTACH]19571[/ATTACH]
Shawn
Shawn,
I may be wrong if I ran the calculations, but surprised that Open-Close returns have been better than Close-Close. I would have guessed the other way around. Good news for those that want to follow.
Regards,
Harry
Hi Harry,
In general, you are correct when viewed on a month by month basis.
I have gone through the data published in this thread from May 7 to Aug 14. I did not publish certain days here when I was out on vacation or had other matters, so I have excluded SPY data on those dates too. I will update these stats and post at the end of Aug.
Overall from May 7 to Aug 14: Open-to-Close +10.11% vs Close-to-Close +13.57% vs SPY +4.48%.
For May 7-31 stub period, Open-to-Close +1.60% vs Close-to-Close -0.89% vs SPY +0.99%
For Jun, Open-to-Close -0.56% vs Close-to-Close -0.64% vs SPY -1.46%
For Jul, Open-to-Close +7.30% vs Close-to-Close +13.57% vs SPY +5.01%
For Aug 1-14 stub period, Open-to-Close +1.85% vs Close-to-Close +1.45% vs SPY -0.06%
May 7 to Aug 14 has included a pretty decent 8% pullback from May 22 to Jun 24, and a 10% V-shaped rally from Jun 24 to Aug 2, and Q2 earnings.. So, the sample size has been small, but some nice events have occurred during the period.
I have not done a careful study yet, but much of the Jul out-performance by Close-to-Close appears to be due to LEV Leaders being the *on average* benefactor of earnings surprises which express themselves primarily as earnings gaps which the Open-to-Close did not benefit from as much.
I would also note that Pascal has done some back testing and said that this method works best when the market is on an overall buy signal and not too over-bought.
It makes sense to me that when the 20d MF is on a sell signal (as it is now), IBD is in market correction (as it is now), etc. this methodology will likely not give positive returns and may not even out-perform the SPY since the LEV Leaders are usually high beta names.
More forward-testing data is needed to build our confidence in this work, so I will continue publishing the data.
Take care,
Shawn
Wow Shawn, quite a response! Thank you for the detailed information.
Hi-
20d MF is on a sell, I am 100% cash, but we continue the work...
[ATTACH=CONFIG]19607[/ATTACH]
For Wed,
[ATTACH]19608[/ATTACH]
Shawn
Good evening Shawn,
I don't believe AAPL closed positive O-C today?
Harry
Hi Harry,
Thanks for helping with the QA..
The results did not look right to me this morning, because the C-2-C always goes from most positive to most negative, and when I looked this am, that was not the case.. Had to update the spread sheet.
Thanks,
Shawn
Hi-
I am going to try and report results based on a 5d rolling window. I assume an equal amount is invested each day over the last 5 days in the LEV Leader groups, e.g. $10K per day is bot on the open and the names from Aug 15 would be sold on the close today. Under these assumptions, the overall amount in the market would be about $50K at a time. This is s strategy that reduces the slippage losses due to commissions and unfavorable bid/asks which can occur on the thinner names. I am going to paper trade this strategy at Interactive Brokers to get a better sense of possible real-world returns.
The 5d rolling results are summarized at the bottom of .png file and the overall file is attached below.
Results from Wed. Aug 21, 2013:
[ATTACH=CONFIG]19609[/ATTACH]
Names for Tracking on Th, Aug 22, 2013:
[ATTACH]19610[/ATTACH]
Here is the spreadsheet for the 5d rolling results:
[ATTACH]19611[/ATTACH]
I am traveling tomorrow, but will do my best to keep up with the study here.
Take care,
Shawn
Hi-
Results for Th:
[ATTACH=CONFIG]19634[/ATTACH]
For Fri:
[ATTACH]19635[/ATTACH]
5d rolling:
[ATTACH]19636[/ATTACH]
Shawn