DMUNCASTER
34 posts
msg #116013
- Ignore DMUNCASTER |
10/17/2013 2:16:20 PM
Sounds like TRO's been watching too many future apocalypse movies.
|
olathegolf
119 posts
msg #116014
- Ignore olathegolf
modified |
10/17/2013 2:24:16 PM
I'm not following. (Surprise, surprise.)
If you average 8 trades per week, then you're roughly investing $40K (in the $5K trade column). You made $2,106 during that timeframe on the 1% stop? How does that correlate to 156%? And why aren't the $10K/trade gains = 2X the $5K/trade gains?
--------
1. Yes, 8 trades per week at $5,000 per trade is $40,000 average per week.
2. Yes, $2,106 total for the 40 week period. This includes $8.95x2=$17.90 trading cost per trade.
3. 156% is simply the addition of all wins and losses (not including the trading cost). 8 trades triggering the 1% stop would be a total of 8%. So, 318 trades at a total of 156% equals about 0.5% per trade. This metric may not be that useful as stated.
4. Increasing to $10K defrays the cost of the trade. It shows the difficulty of overcoming trading costs when you don't have a lot of capital to invest.
|
jimmyjazz
102 posts
msg #116015
- Ignore jimmyjazz |
10/17/2013 5:14:17 PM
Ah . . . I misinterpreted "Total Gain % does not include trade costs" and "Win % is does not include trade costs". I assumed the net gain also didn't include those costs. My bad. Very illustrative -- trade size matters with fixed trade costs.
|
jimmyjazz
102 posts
msg #116023
- Ignore jimmyjazz |
10/18/2013 9:15:31 AM
"It will be interesting to see the longer term results in down markets like late 2008 - early 2009. Note the correlation between target %, win %, and total gain."
Can you backtest that time period? I'm going to guess that the optimal target winning % will be lower in a bear market, but that's just a hunch.
|
Kevin_in_GA
4,599 posts
msg #116027
- Ignore Kevin_in_GA |
10/18/2013 10:09:14 AM
The very first post of this thread can help answer that question:
http://forums.stockfetcher.com/forums/Filter-Exchange/5-A-WEEK-FILTER-BASED-ON-TRO-S-CROCK-POT/81139
This version of the Crock Pot shot for a 5% gain or took a 10% stop loss, investing in only the top stock per week chosen. The return before commission costs was 136% from 10/12/2008 through 10/13/2009. What you see is that during periods of serious drawdown the inverse ETFs rise to the top, and that is what the system tends to be invested in. When the market rebounded in April you see a shift back to long positions in stocks.
|
jimmyjazz
102 posts
msg #116046
- Ignore jimmyjazz |
10/18/2013 2:39:04 PM
Yes Kevin, I use that filter and trade it every week. I don't exactly follow your entry & exit rules, and sometimes I use calls, but the spirit is the same.
Have you considered exiting at 5% OR Friday close, without the 10% stop loss? That would take out the inherent error of assuming 5% gets triggered before -10%. I might take a crack at manually backtesting those trades.
|
olathegolf
119 posts
msg #116049
- Ignore olathegolf |
10/18/2013 3:54:04 PM
Kevin - I had forgotten about that post. Good stuff. Thanks.
|
frsrblch
35 posts
msg #116050
- Ignore frsrblch |
10/18/2013 4:06:08 PM
If you want to resolve trades where the 5% profit stop and and the 10% limit were both triggered in the same week, you can look up the stock on TradingView. They have daily data going back years that you can scroll through to see which limit triggered first.
|
Kevin_in_GA
4,599 posts
msg #116119
- Ignore Kevin_in_GA
modified |
10/20/2013 8:42:27 PM
Trades for tomorrow's open:
SFUN - 74 shares
DUST - 120 shares
CSIQ - 183 shares
QCOR - 65 shares
YOKU - 140 shares
WLT - 280 shares
Based on last week's profit, we will put approximately $4,250 into each trade. These have all hit 13 for 13 so far, so I divided the total equity across the total number of stocks that have gone 13/13.
My manual backtesting of this approach so far has not been all that reassuring. When these stocks do not hit the target, they often crash brutally and wipe out several weeks of gains. I'll continue to document the system trades here and we can see how it does going forward.
|
mahkoh
1,065 posts
msg #116121
- Ignore mahkoh |
10/21/2013 3:28:29 AM
Those planning to hold beyond Friday if the target isn't met may want to reconsider QCOR and WLT. They are scheduled to report earnings next week.
|