Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
Hi Wizard I wrote a little strategy to test going long/short when all 4 signals line up. I couldn't get historical tick data so I had to use minute data. I tried 1m, 2m, 5m and the results were all negative. I'm wondering if I'm missing something. I don't expect backtesting to be perfect but in general what I look for is an idea that is positive in backtesting and then I trade it manually.
I'm attaching my code. Comment out the log(...) calls, that's a little function I wrote that's in a partial class to print a timestamp with each Print.
Hi Velocity, your template looks real neat. It would be helpful if you post this template file. That would save real time individualy tweaking all the indicator settings... thanks
I'm not surprised that the minute data gives crummy results.
Since you have no tick data try range bars of 2, 3 or 4. Or if the lack of tick data prevents you making range bars I'd just give up testing the minute data for this system.
Wizard,
I checked the LBR310 setting again.
According to LBR group's lecture, the fast line is 3/10 oscillator, the slow line is 16ma of 3/10 oscillator. The setting should be fast = 3; longLine = 16; slow = 10 in our indicator.
So you are right, I am wrong. LBR310 setting showed on indicator should be 3,16,10; not 3,10,16.
I tested my strategy on Minute, Range, Renko & Tick & determined the signals are clearer & less choppy on Tick charts. On auto-trading: Over the past 2 years, I've tried coding various strategies with mixed results. I don't like Ninja's auto-trader function. I've found it *always* takes the trade 1 bar too late. I see the cross over or - whatever - on my chart, where I would take the trade & in the strategy - nothing happens! 1 bar later - the trading strategy kicks in & places the trade! I won't trust my hard-earned $$ to Ninja's auto-trading strategy function. I tried to imagine myself turning a strategy on & then going to the movies. I can't see that happening. It can't be trusted. I'd worry myself sick. What about if you lose the internet connection (& I do, sometimes) & it has to re-connect. Too much risk.
After all, it's what you learn AFTER you know it all, that counts!
I only write strategies to test out ideas. I used to autotrade several strategies, I did it for months and they worked and I didn't have any problems. It was cool when we were on vacation we went to Disneyland and I came home and I had made $700 that day on my automated strategy. So I can vouch for the automation. I have ran automated strategies on and off since December and never had a series problem. Experienced a couple design flaws but I quickly coded around them. I also set it up to email me on every execution so I can know what's going on when I'm away.
All that said, most of my strategies stopped working (they couldn't adapt to the constantly changing markets) and a few that still work are less profitable than my other strategies that I do manually so I stopped them.
Anyway, it's great to test out ideas but I now do everything manually. One thing I use strategies for is developing indicators. I've developed lots of indicators but when I test them only a handful test positive. It's nice having "proof" that an indicator works, eyeballing it is often misleading.
Now that I'm back from vacation I can test out the strategy I wrote with tick data. Will post the results.