Hey wow, someone else finally discovered the strategy I was using 4 years ago :) Doesn't really work anymore since the market has become more mature, but congrats!
cryptographic1
⋅
Here's a hint - add lag to the second linreg line, like so:
lag2 = input(lag2)
outslow = linreg(src, len2, offset)[-lag2]
a few ticks of lag will improve your results generally, prevent some false signals. Also, 100/150 settings you are using are generally pretty good on 15 minute ticks, try in the range of 25/35 for hourly ticks. Cheers :)
cryptographic1
⋅
oops,
outslow = linreg(src, len2, offset)[lag2]
cryptographic1
⋅
strange, brackets don't display in comments... but, square brackets instead of parenthesis outslow = outslow(lag2)
overttherainbow
⋅
As bad as you think it is, you still reaped profits. Better than what a lot of people use as their strategy no doubt.
May I ask, what do you think is the best/a better way of using the linear regression line? I was thinking of using it as a trend indicator (based on what direction it's moving, so something like: uptrend = lsma[0] > lsma[1].
What are your thoughts on it's potential uses?
ridvansozen1
⋅
@overttherainbow, Hi sir, I did think if I find the return moment of 2 different future linear expectation, I can generate a signal. However, it is not so systematic. Therefore,I am working on different usage of linear regression. I hope I would make the better version.