StockFetcher Forums · View by Author: (61 messages)  [ Display By: Date / Subject ]<< 1 2 3 4 5 ... 7 >> 
Backtesting Support · Very frustrated with incorrect backtest results
tombrown1
msg #108396
10/16/2012 12:23:12 PM

Ok. Thanks. Ill try that. This must be a change from the old software. It was never necessary to type select by before.

Best,

Tom

Backtesting Support · Very frustrated with incorrect backtest results
tombrown1
msg #108392
10/16/2012 12:47:47 AM

Hello. I try the backtest with 4 different selection criteria:

1. volume descending
2. volume ascending
3. price descending
4. leave the box blank.


They all achieve the same results.

They also yield a stock a couple of months ago that never showed up in my daily filter emails (identical filter).

I haven't done a thorough check for validity yet - didn't think it was necessary.


Here is the backtest:

Stocks are not OTCBB
Bollinger Width Oscillator(5,2) below -50 in last day
Lower Acceleration Band(5) dropped more than 3.5 Percent
Lower Keltner Band(5)dropped more than 3 percent
Day Position(0.5,1)below day position 2 days ago
do not draw day position (-1.00,5)
do not draw day position
average Volume(4) > 120000
close > 1.20
IMI(4) below 9



Entry Filter
levamit old 2/3/2 hld 7 days or rsi>55 (saved filter))
Basic Setup
Name: levamit old 2/3/2 hld 7 days or rsi>55 - #4 (Long)
Approach Type: Long
Start Date: 9/26/2011
End Date: 9/25/2012
Benchmark Symbol: ^SPX
Cash and Holdings
Starting Cash: 100,000.00
Commission Per Trade: -
Trade Size:
Allow Negative Balance:
Partial Shares:
Exit Setup
Stop Loss: -
Profit Stop: -
Trailing Stop Loss: -
Minimum Holding Days: -
Maximum Holding Days: 7
Exit Test #1:
Rsi(2) crosses above 55
Extra Indicators
Entry Columns:
Show Performance After: after 2 days
after 5 days
after 10 days
after 25 days
after 40 days
Advanced Options
Selection Method: volume ascending
Shuffle Entry:
Entry Price: Open
Conditional Entry:
Use Match Date Close:
Exit Price: open
Maximum Trades Per Day: 2
Maximum Open Positions: 3
Maximum Selected Stocks: 2
Close OPEN Positions at End:
Allow Multiple Positions Per Symbol:
Force Same Day Trade:


Backtesting Support · "Maximum selected stocks"
tombrown1
msg #108197
9/27/2012 11:56:07 AM

Ahhhhh...now I get it. Very enlightening. Thanks!

Backtesting Support · "Maximum selected stocks"
tombrown1
msg #108178
9/26/2012 10:20:34 AM

Bump

Filter Exchange · STOCHASTIC CROSSOVER SYSTEM FOR SDS AND SSO
tombrown1
msg #107632
8/15/2012 12:04:31 AM

I guess I'm not as excited about the curve as you are. The method went two years without making a dime from septemberish 2009 to septemberish 2011. It made a ton of money in two separate one year periods and did nothing the rest of the time. Would you continue to trade this system for the next two years straight if you didn't make a dime? Seems unlikely.

Sorry to be such a downer - I do plan on doing some further back testing on this because it shows promise - but I just don't see it as a high probability consistent winner in its current form.

Best,

Tom

Filter Exchange · STOCHASTIC CROSSOVER SYSTEM FOR SDS AND SSO
tombrown1
msg #107601
8/13/2012 11:03:40 PM

OK, so all these new posts got me a bit excited again. I backtested 8/8/10-8/8/11 (the one that was awful in my previous bt) but this time I only took trades for SDS that had a positive 20sma/50sma cross on SPX; and I only took the trades for SSO that had a negative 20sma/50sma cross on SPX.

I am only testing one stock at a time for the year. First SDS, then SSO. I realize this isn't exactly how you trade it, but I think it's close, and it's simpler to backtest this way.

SSO lost 1% for the year while SDS lost 7%.

I was hoping this filter would help things out, but still I'm getting nothing.

There are filters that have made money every year for the last ten years - this one hasn't. I'm still interested in ideas to improve it - I still haven't tried stops - but I'm a bit skeptical because of the large swaths of time it lost money.

I think if you would have traded real money with this thing over the last 5 years or so, you would have given up at some point or another. Kevin, what are you going on that you are trading this with real money? And not even following the method at that. It may be that you are trading it better than the way it was backtested - can you codify the way you are actually trading it?

Thanks again for the ideas guys - I think this one may have hope - I just haven't figured it out yet.

Best,

Tom



Filter Exchange · STOCHASTIC CROSSOVER SYSTEM FOR SDS AND SSO
tombrown1
msg #107541
8/11/2012 6:00:36 PM

Kevin, what criteria are you using for your trades? The trades you're describing don't seem to coincide with the system you outlined.

TB

Filter Exchange · STOCHASTIC CROSSOVER SYSTEM FOR SDS AND SSO
tombrown1
msg #107510
8/9/2012 10:44:10 PM

Aug. 8, 2010-Aug. 8, 2011

SSO gained 7%
SDS lost 12%


Sorry, but this causes me to abandon this backtest. Those are unacceptable results for an entire year's worth of trading regardless of what else it did in previous years.

Too bad because I had some hope for this strategy. If I find some time maybe I'll try it with stops.

Best,

Tom

Filter Exchange · STOCHASTIC CROSSOVER SYSTEM FOR SDS AND SSO
tombrown1
msg #107480
8/8/2012 10:37:49 PM

Thanks Nick.

Filter Exchange · STOCHASTIC CROSSOVER SYSTEM FOR SDS AND SSO
tombrown1
msg #107479
modified
8/8/2012 10:36:54 PM

Ok Kevin, you have my attention. I've started a manual backtest using this for the first line:

- 5 for the k period, 1 for the d period, and 5 for the slowing period

and this for the second line:

- 5 for the k period, 1 for the d period, and 10 for the slowing period

I'm looking for the fast line to cross under the slow line. When the trading day closes with the cross, I buy at the open of the next day. When the trading day closes with the fast line above the slow line, I sell the next open. I used no stops (can't stand them).

I tested one etf at a time. I just finished August 8, 2011 to August 8, 2012.

SSO gained 112%
SDS gained 31%.

After testing the crash, I have a feeling I will have to retest everything with stops, but for now I'll leave them out.

More to come!

Thanks again for sharing Kevin.

TB


Edit: Oops, I made a small mistake in my calculations - backtest numbers are now correct.

StockFetcher Forums ·  · << 1 2 3 4 5 ... 7 >>

*** Disclaimer *** StockFetcher.com does not endorse or suggest any of the securities which are returned in any of the searches or filters. They are provided purely for informational and research purposes. StockFetcher.com does not recommend particular securities. StockFetcher.com, Vestyl Software, L.L.C. and involved content providers shall not be liable for any errors or delays in the content, or for any actions taken based on the content.


Copyright 2018 - Vestyl Software L.L.C.Terms of Service | License | Questions or comments? Contact Us
EOD Data sources: DDFPlus & CSI Data Quotes delayed during active market hours. Delay times are at least 15 mins for NASDAQ, 20 mins for NYSE and Amex. Delayed intraday data provided by DDFPlus


This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.