Optimization moving average crossing strategy

• 05/05/2020 at 4:12 PM #130086
  GraHal

  Maybe we are not realistic when we expect high gains as we see in backtest as they have been curve fitted to waves other than 10 sec TF?

  If we trade DJI at 0.2 size on 10 seconds TF then the strategy should exit at about £10 / 50 points tops, probably more like £6 / 30 points gain?

  My view is that if we trade the 5 sec or 10 sec then we are trading very short term waves of about 5 mins tops?

  Trades on 10 sec TF that stay open for  > 5 mins are probably not really based on a 10 sec TF strategy??

  We need trades to open, ideally straight into profit after making up the spread and then at first sign of a pullback on 10 sec TF we need to exit?

  

  GraHal

  Participant

  Topics: 204

  Replies: 8403

  Been thanked: 2043 times
  05/05/2020 at 4:29 PM #130092
  nonetheless

  You’re probably right. That would be a proper scalping approach. With this one, the TS goes to breakeven at about 17 points, but maybe it could benefit from some other exit condition? I tried that with the longer TFs and there was no gain, but could be a good idea at sub 1m. Either fixed amount in the TP, or any reversal of a v fast MA ?

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
  05/05/2020 at 4:52 PM #130101
  Franro

  Yeah.. Tell me if you guys modify it for this week somehow, I feel its hard.

  

  Franro

  Participant

  Topics: 21

  Replies: 197

  Been thanked: 19 times
  05/05/2020 at 4:52 PM #130102
  Francesco

  About that, i invite you to follow the ScalpMaster topic… news coming in the next days 🙂

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  05/05/2020 at 4:55 PM #130103
  nonetheless

  @francesco — too true. These things (micro-trend) have a perverse appeal but may be total fantasy. TBH, one of the reasons I posted it was to show how easy it is to get those kind of results … for a week. Running this same system (Corona) across different TFs I think the optimum may be 2m, but until we get v11, that’s still only a 1 year back test. For now I’m concentrating on the 5m zone.

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
  05/05/2020 at 5:01 PM #130104
  Francesco

  I’m afraid that v11 only a dream will remain for a long time 🙂

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  05/05/2020 at 5:04 PM #130105
  nonetheless

  Franro wrote:

  I feel its hard

  Actually it’s not … and you’ll have to learn it sooner or later. Read the whole thread so you understand the logic of what it’s doing, esp @Jan’s part which is the heart of it. Then start with the Corona OTD.itf which has all the variables in the optimisation box ready to go. Facilissimo.

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
  05/05/2020 at 5:04 PM #130106
  GraHal

  nonetheless wrote:

  Either fixed amount in the TP, or any reversal of a v fast MA ?

  Both sound like good ideas, esp the very fast MA as that allows for if we hit lucky and we are surfing a bigger wave?? 🙂

  I have been thinking lately about trades only opening at multiples of 5 mins … when conditions are also met of course? A strategy may stand a better chance of going into early profit due to plenty of Algos from the professionals likely running on 5 min TF?

  I’ve noticed an odd phenomenon lately, at 10 to the hour there seems a lot of action / volatility?  Not every hour, but 11:50, 12:50 bit less so, 14:50, 15:50 and 20:50 (not every day, but more often than random).

  I wonder if @Vonasi has ever done one of his excellent comprehensive studies on such tactics??    🙂

  Just a few thoughts anyway.

  

  GraHal

  Participant

  Topics: 204

  Replies: 8403

  Been thanked: 2043 times
  05/05/2020 at 7:18 PM #130124
  Franro

  Maybe putting in a lower sl? I feel like its good as im saying but idk about the big drawdowns. If you geniouses find something out tell me.

  

  Franro

  Participant

  Topics: 21

  Replies: 197

  Been thanked: 19 times
  05/05/2020 at 7:50 PM #130130
  Franro

  Also all losses so far is on short trades if you would want to know.

  

  Franro

  Participant

  Topics: 21

  Replies: 197

  Been thanked: 19 times
  05/06/2020 at 5:38 PM #130288
  Vonasi

  GraHal wrote:

  I wonder if @vonasi has ever done one of his excellent comprehensive studies on such tactics??

  You would have to exactly define ‘a lot of action / volatility’ and then analyzing it is not so difficult. Our main problem is the lack of historical data on fast time frames which means that we are only really analyzing a very small data sample as there are not that many ten to the hours on a five minute chart.

  With arrays in v11 it would be very easy but at we might have to wait for IG to sign up so we can get more than end of day easily.

  Perhaps there is a lot of activity because there are a lot of people who trade hour and 4 hour charts and when it gets near to the close of the candle perhaps they get a bit twitchy and close their trade early to lock in a profit or to stop a loss from being any bigger than it currently is.

  1 user thanked author for this post.

  GraHal

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1810 times
  05/06/2020 at 5:43 PM #130290
  ArnoldB

  I believe we are all waiting for the new features of PRT v11 !! I think that the algorithms that we are developing will be able to make a qualitative leap, in particular with much wider access to historical data.

  

  ArnoldB

  Participant

  Topics: 4

  Replies: 12

  Been thanked: 1 time
  05/06/2020 at 6:30 PM #130299
  GraHal

  Anybody trading the DJ 10s Corona on 10 seconds and getting better results than attached DJ 10s Corona running on 2 min TF?? 🙂

  Pos Size = 0.2 so £10.95 Av Gain per Trade to directly compare with Pos Size = 1

  spread = 4

   

  

  GraHal

  Participant

  Topics: 204

  Replies: 8403

  Been thanked: 2043 times
  05/06/2020 at 7:25 PM #130304
  GraHal

  Re above … I didn’t change anything or optimise for 2 mins TF btw … I just backtested straight off.

  Oh I changed the Timeframe to default that’s all!

  So it might auger well for a more predictable performance?

  

  GraHal

  Participant

  Topics: 204

  Replies: 8403

  Been thanked: 2043 times
  05/06/2020 at 7:31 PM #130308
  nonetheless

  Is that with 6m in the top TF?

  2m is a really good TF for this, i’ve been working on that today and would have something to post if it didn’t keep crashing on me. 🤬

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
• Author
  Posts