Trailing ATR stop management and re-adjust stop display

Forums ProRealTime English forum ProOrder support Trailing ATR stop management and re-adjust stop display

Viewing 3 posts - 1 through 3 (of 3 total)
  • #215191

    Recently I have changed my stop management (and TP) to a trailing ATR across large parts of my portfolio with outstanding improvement to my R/R, Win rate and overall profitability.

    For the most part these are performing very well. However during a FTSE short I noticed the vertical lines attached to the trailing stop display looking rather unusual. This has not been the case on other trades with the same management but I saw it again to a lesser extent on the NDQ last night. Both trades worked out fine and stops were honoured as expected.

    This bot has been started with ‘Re-adjust stops’ and I wondered if this could be the cause being an IG broker issue, or is my code too simple, or is there no problem here?

    This particular bot is actually a 1 min TF, I was just watching it on the 5min and believe it should have been stopped out during a flush point or back test of the 20ema.

    Bot trading has really taught me to question everything especially as I’m not a programmer by trade!

     

    Thanks

    #215193

    Hi,

    I have never seen that, but I can imagine it can occur like that. It should be caused by a combination of the TF you set the Trailing Stop in (per TF command) and the TF you observe it (per TF of the chart), and the fact that you repeat the setting each time (thus of the Set Stop pTrailing).
    Notice the latter, because that may be the unexpected for you.

    It is not forbidden to repeat the setting, and you also most certainly should not change that (the strategy works technically fine as it is (your statement)), but you should realize anyway that this command is not made for having it repeated at each TF bar (like 5 minutes in your case).

    As far as I can judge it, the trailing won’t exactly do what you want just after your code hands it to the broker, as in : the first minute it has this distance per a wrong setting in your code, but in minute 2, 3, 4 and 5 it is what you intended (close to the price). Think of a math error in your code (the c7 assignment).

    Your thought that it is caused by the “Re-adjust Stops” feature is a nice one, and it well can be the case. But then it would still be about this “math error” and how that re-adjusting is applied at each 5 minute threshold. Thus, it would then be set all the way up (close to the price where you bought/sold) at initiation (your Set Stop pTrailing command) and when settled (next 1 minute bar) it does what’s intended.

    This last sentence brings me to a possible error in PRT itself, caused by the broker doing odd things with that stop distance. This latter is a known issue and the former is a guess. Anyway it would explain why you’re not always seeing this (because the Stop Distance is a broker issue which occurs regularly, but mostly not).

    I just brought forward your screenshot for the 20th time, and the more I look at it, the more I think that your intended Stop level is at the top line – not the general line close to the current price. The effect would be that the position goes out at a north excursion (for Short) which makes it a mere functional trailing profit (instead of a functional trailing stop). This is how I said “don’t change it”. It is only that you should make it close to the price permanently.  This is probably done by means of setting it each 1 minute instead of each 5 minutes.

    ???
    🙂
    Peter

     

     

    1 user thanked author for this post.
    #215208

    Thanks for having a look!

    The trailing profit as you call it is actually more correct. I no longer run a set target on these trades as they have a time stop. So for the example I gave I don’t actually know if I would have experienced an error had price retraced far enough to hit the profit stop as it locked in 60pts or so. I intended the stop to trail. If I can get a bot that accepts trailing I will choose to trail, because a lot of them are better with a fixed stop.

    It is actually a 1min TF bot and I just watched this trade on 5 min chart, I don’t normally do that. So I think the stop price gets calculated every candle from the atr default[14]. Thus it makes sense that it can be a calculation error as I’ve given a fixed value with no room to move. It didn’t do it every candle, but most candles open and it would show the stop at the entry stop price (or near) about -$38 for just a second or two before reverting to the trail position.

    The beauty of this method is that if you have a volatility spike against you the stop will move a few points back and it’s like dodging bullets rather than being a sitting duck, then the price can get back to making you money. The downside is you can’t precisely calculate how many r’s your risking. So like everything we do there is a trade-off. Tonight I had a mean reversion short on the FTSE that took a bad entry and one of them cumulates orders. I nearly wee’d my pants when I realised how much risk I was wearing to what I thought! Worked out ok though, and now I will adjust my size.

    Other 1 min trades I have use a fraction of the daily atr[1] which can give me near identical back test results but perhaps avoid the problem altogether as they would be calculated once and then trail?

    I won’t use ‘re-adjust stops’ with 1min trailing profit in future, or at least test it without. But I think your right about the calculation error/broker combined. IG has some weird issues that cost me money but after being the recipient of outsized gains like 220pts in 40 mins Feb ’22 (also FTSE) they seem to balance out over time.

    CC

    1 user thanked author for this post.
Viewing 3 posts - 1 through 3 (of 3 total)

Create your free account now and post your request to benefit from the help of the community
Register or Login