OmniTrader Professional Forum
OmniTrader Pro General Discussion
RhinoHMA Strategy - Is it Possible?
6 replies, 697 viewings
|
OmniTrader Professional Forum
OmniTrader Pro General Discussion
RhinoHMA Strategy - Is it Possible?
|
Last Activity 5/30/2020 8:23 PM 6 replies, 697 viewings |
|
|
|
| Printer friendly version |
Jeff Brown Member  Posts: 27 Joined: 1/29/2008  User Profile |
I thought, "This looks like something you could create a strategy around and potentially make it a mechanical strategy". Has anyone tried to do this? Here's what my first attempt looks like. I can't say it works because I don't see any signals on the 39-minute Renko chart. But I used the Profile from the Bash file, set the Period to Range - 1ATR(14)/39 Minute. You will see the order of all the data periods at the bottom of the screen shot. Next, I created a strategy using the included sysRhinoHMA system with it using the Current time period. I then added confirmation of the 39 minute chart with the 20 minute chart (Current - 1) in Confirm1. I allow it to look back 4 bars for a signal. I added in a second system for confirmation against the Daily timeframe (Current - 2). This is a custom system I use that creates a kind of volatility indicator and issues signals when the value crosses over a moving average. Here's also a look at one of the confirm blocks. Next, the Filter Block. I kind of combined Al's method with Ron lewis's confirmation method (also from the BASH) where Ron was using information from the WoodiesCCI and MACD_HIST. When I looked at those against the points Al's method would say "enter the trade", they looked like they were good confirmation signals and avoided a few spots where things just moved sideways. So, for a mechanical system, I added them (for now). I simply wanted to see that they were both either above zero or if below zero, they were at least moving higher. Finally the TradePlan block. I started with the Long Option plan and added a profit target stop since Al would take profits at $1,000 (on 10 ITM Calls on the SPY). The best I could do there was say exit on 50% profit since there isn't a $ profit target that I found. My reasoning was that if you buy 10 calls at $2, then @ 50% profit, you have $1,000. I know, not perfect, but it is a starting point. You will also notice that I set the Option Search Settings to be a Fixed Loss Stop on the position instead of the underlying because the video mentioned Al uses a "tight stop". That means to me, he is only watching the value of the Option, not the SPY. Now for a few questions: 1) Has anyone created a strategy that uses multiple timeframes for confirmation similar to this? Any advice on how to do it better or correctly? 2) Any thoughts on the Trade Plan and how to make it use an absolute profit target? 3) Any experience in using Renko timeframes with the multiple timeframes? Is this really possible? 4) Any thoughts, recommendations, things to explore...? I've attached the Strategy in case anyone is interested. You'll have to delete the VIXFlip system and confirmation block -- I'm sure that will cause you problems. [Edited by Jeff Brown on 1/21/2020 10:47 AM]  Attached file : 2-Rhino-Strategy.png (27KB - 611 downloads) Attached file : 3-Rhino-Confirm.png (59KB - 624 downloads) Attached file : 4-Rhino-Filter.png (11KB - 630 downloads) Attached file : 5-Rhino-TP.png (64KB - 611 downloads) Attached file : 6-Rhino-TP-FPS.png (46KB - 619 downloads) Attached file : RhinoHMA.ots (1636KB - 33 downloads) Attached file : 1-Rhino-Timeframes.png (184KB - 564 downloads) | ||
| Jeff Brown Member Posts: 27 Joined: 1/29/2008 User Profile |
1) Simplified Strategy -- 3 blocks, no confirmation: 2) I added the sysRhinoHMA system to the chart. You will also see that I added it to the 20 minute chart and I added my VIXFlip to the daily chart. Since I removed the confirmation and filter blocks, those really don't matter. 3) I see in the top right the 39 minute chart shows a signal from sysRhinoHMA but it doesn't show up in the Vote section (on any of the charts). It is active... In fact I have 4 active -- different versions with different blocks and filters active or not and different order -- trying to see a signal actually generated from the strategy, not just plotted from the system. I haven't created a real time strategy before. Any suggestions as to what I might be doing wrong for the Real Time Profile and Strategies? [Edited by Jeff Brown on 1/21/2020 10:37 AM] Attached file : 1-Rhino-Basic.png (6KB - 569 downloads) Attached file : 2-Rhino-Signal.png (60KB - 585 downloads) Attached file : 3-Signal.png (87KB - 561 downloads) Attached file : 4-Strat-List.png (8KB - 563 downloads) | ||
| Jeff Brown Member Posts: 27 Joined: 1/29/2008 User Profile |
Simply checking it enabled the 3 different time periods to display votes. Now that I have votes showing for different versions of the strategy, I can move to the next challenges -- figuring out confirmation and the right level of fixed profit. [Edited by Jeff Brown on 1/21/2020 3:31 PM] Attached file : 1-DataPeriods-orig.png (8KB - 535 downloads) Attached file : 2-DataPeriods-use.png (8KB - 542 downloads) Attached file : 3-Voting.png (77KB - 542 downloads) | ||
| Jeff Brown Member Posts: 27 Joined: 1/29/2008 User Profile |
First image: Each chart is in a different time frame. The top Left is in sync with the active time frame (Range 1ATR(14)/39 Minutes). In the chart you will see a red line (Position Fixed Stop Loss) and a gray line (Position Trailing Stop). The Trailing Stop is acting as the exit. There is also a stop in the TP to exit before expiration. The results of the one signal shown on the vote line of the upper right chart. More work needed to have the Option Search improve the selection of which Option to use. A look at the strategy with multiple time frames. In talking with Barry, the Systems block must be using the Current time period. That's why I've set the Active Time Period (bottom of the first image) to Range 1ATR(14)/39 Minutes. Then, I used 2 confirm blocks -- each with a different time frame (Current - 1 is the Renko 20 Minute and Current - 2 is Daily). Filtering with a quick confirmation that Woodies CCI and MACD Hist are both either positive or if negative, they are moving toward zero. A quick look at the Trade Plan -- the trailing stop is using a percentage of the ATR to act as the cushion and follow the trade up. Final image is Auto Trade. Barry just surprised me with news that AT can handle Real Time trade automation. I've only used it for EOD. The cool part is, you can set the start and end trading time. Al started at 11:00 EST. I'll open up that window initially to shake it out. The one aspect that AT is lacking is it doesn't have a fixed quantity to trade, only a $ or %. I'll try those out to see how it reacts and how many contracts are being generated with an order. I'm guessing I may need to set a low % but round to 10 contracts. If anyone has experience with this to configure it so it trades a specific number of options, I'd love to hear from you. Final image -- the Trade Calculator. I have it set to manually assign a quantity of 1000. I don't know if this is used or not with AT, but if you manually open the trade, you will see just above the Trade Calculator window that 1000 = 10 option contracts. Barry tells me that this is because TC was originally created for stocks and that is why. Just ignore the lower right portion of the Trade Calculator that says Final Size: 1000 Shares\Contracts. It really means shares. I'm hoping TC comes into play with AT but I'm going to guess it will ignore TC. While this isn't exactly the way Al Fyfe described his system at the BASH, it looks to be coming close to entering at about the right time, selecting a good SPY option to trade and then exiting with a hansom profit. I'll post more after I let it run for awhile. [Edited by Jeff Brown on 1/22/2020 4:13 PM] Attached file : 1-Rhinov4.png (77KB - 503 downloads) Attached file : 2-Advisor.png (70KB - 498 downloads) Attached file : 3-Stratv4.png (7KB - 494 downloads) Attached file : 4-TP.png (62KB - 499 downloads) Attached file : 5-AT.png (19KB - 496 downloads) Attached file : 6-TradeCalc.png (62KB - 499 downloads) | ||
| Jim Dean Sage    Posts: 3298 Joined: 3/13/2006 Location: L'ville, GA User Profile |
| ||
| Ryan Olson Veteran Posts: 151 Joined: 12/29/2006 Location: Austin, Tx User Profile |
I just wanted to say wow - I had no idea this was out here and that you all had put so much work into improving it. I didn't have notifications set for this forum but you all are incredible. | ||
| Buffalo Legend  Posts: 556 Joined: 7/11/2007 Location: stafford, va User Profile |
FYI - Need to be careful with your selection of multiple data periods - the 20min does not go into the 39min evenly. This means during Backtesting there is a 1 min overlap where essentially OT can see into the future. I'm not describing this well, but not using evenly divisible bars can let OT "cheat" during BTs so it knows the outcome of the confirming timeframe bars before the signal bar starts, so it obv uses that info that wouldn't be there at the same time if they were evenly divided - the confirm bar would be closed, confirming or not, BEFORE the signal starts, but in this case it's not. Yeah not doing a good job making this clear sorry. In this case I would use 13min I made this mistake once and what looked great during BT failed horribly - OT can't cheat at the HRE - when really trading. I don't know how this 1 min overlap will effect your stuff but it's a risk if you use it. JS. To trade in 100 share increments so options will work when using % or $ allocation go into Port sim - simulation settings - trading parameters - bottom checkbox - check the box that says to "trade in increments of 100 shares". This will have OT buy the max number of options for the $ or % you select. It works in my ATM3 Options method - it didn't till I checked that! [Edited by Buffalo on 5/30/2020 8:36 PM] |
|
|
|
|
Administrator
Forum Moderator |
Registered User
Unregistered User |
 E-Mail this thread to a friend |
Toggle e-mail notification |
