It looks like from the first (worst) entry I went through a 12 tick MAE. This is within acceptable risk tolerances for the size I was trading.
I suppose I do go through different ways of managing losing trades. Sometimes I don't allow it to hardly print a candle against me before exiting. Sometimes I allow it to print a swing against me as in the trade we're talking about. I guess it depends on market structure, conviction/confidence, etc.
Maybe I would be better off sticking with the tighter method. I don't know... Exits are a funny thing. You can't ever do the right thing 100% of the time. Take profits and it might run, don't take profits and it might turn sharply on you. All you can do is what you feel is best for the given structure of the market. Range = take more profits sooner/at targets, trend= opportunity to let it run. The main deal is never being frustrated with less than perfection, cause it's all we have to work with.
The following user says Thank You to indextrader7 for this post:
Thanks for the stats there. Could you give a bit of a written explanation? I'm having a bit of trouble knowing for certain how to read those tables. I take it that it's a distribution of respective candle lengths in points?
Sure, happy to oblige - it's thanks to you that I've got back into playing with Excel.
Doing tables of swing size distribution is an idea from @FuturesTrader71 (he talks about it in one of his webinars, and has posts about it on his website).
Basically, you take your chart of choice, apply your zigzag of choice, output the data and then paste it into Excel. So the study isn't of candle size but of swing size in points.
Then you just set up the bins for swing size, and count the number of occurences in each. The most frequent size (what would be called the POC in a volume profile) is colored red, green marks the limit of the first standard deviation, and yellow the second. So you have a pretty good idea of what constitutes a normal rotation, what sort of 'tail' you might reasonably anticipate before getting a swing pullback, etc.
As you can see, there's almost a 3 tick difference in average swing size between the 250 tick chart and the 1 minute, and the main distributions shift about 2 ticks.
The following user says Thank You to futuretrader for this post:
Don't want to distract from this thread too much but could you post a "recipe" for doing this? I use NT and would love a list of specific instructions - specifically I don't know how to "output the data". I'd like to do this for some other markets I trade. Feel free to PM but I thought I'd ask on the thread so all can benefit. Thanks in advance!
Seek freedom and become captive of your desires. Seek discipline and find your liberty. - Frank Herbert
Did you limit the time for this? As in, I would like to see the data from only 8:30AM CST to 11:00AM CST. AKA only the relevant swings for my normal trading session. I would think that if you simply did it for all 24 hours, you would get some real scewed data comparing 250tick vs 1min (1 min flattens out much more since it's time based).
Again, would love to see data only from the aforementioned session.
I agree about the overnight, it would be a pointless comparison. Those were stats from 9:30-16:15 EST, with 90 days loaded. My guess is that if you also eliminate the RTH volume extremes, both high and low, they would converge quite a bit.
I'll run the morning session and post the results - maybe I'll even figure out how to automate the Excel part a bit more.....
With some platforms, you can write some lines of code to show the distribution directly within the platform, as an indicator, which allows to avoid the Excel step.
And you can change the parameters (250 ticks --> 2 min), and the distribution is automatically updated.
I have done it with Sierra Chart, but I do not know if it is possible with Ninja Trader.
The idea is to code an indicator.
This indicator will only work on the last bar of the chart.
On this last bar, it will analyze all bars, and count the number of swings in each "bin".
And the indicator will show this distribution.
My point here is just to indicate you this possible way to get rid of Excel in the process.