Staying In My Lane - A Read-Only Journ(al/ey) | Page 13 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

The Journ(ey)al of a Lazy Trader 0 replies

mt4 read only password question 2 replies

"Read only view" on Metatrader 10 replies

Attachments: Staying In My Lane - A Read-Only Journ(al/ey)
Exit Attachments

Staying In My Lane - A Read-Only Journ(al/ey)

  • #241
  • Mar 13, 2019 2:51am Mar 13, 2019 2:51am
Quoting deltaone
hi there, IF you are trying to get an algo run sufficently on indices, i think itīs a good idea to use a time filter. it might help your algo to be only active when indices are driven by high volume. f.e. trade dax only when fdax on eurex is traded while xetra is open with good volume (f.e. 09:00 - 10:00 cet) maybe itīs a good idea to stay out of market when liquidity is low ( f.e. during 11:00 - 14:30 cet) never allow your algo to open orders on fdax when xetra is closed (index-arbitrage as a hint). never allow your algo to open new orders letīs...
Thanks Delta. To me, as you probably know from most of my bookmarked post.... TIME ACTION IS EVERYTHING. The relationship between Time and volatility/liquidity is strong and PREDICTABLE. Folks chase and value price. I VALUE Time and nothing else... in trading as well as in real life. Time is the only currency I understand :-)
Please continue to share your wisdom bro.
Staying in my lane...
  • #242
  • Mar 13, 2019 2:56am Mar 13, 2019 2:56am
Quoting alphadude
{quote} Nothing beats the Alienware(Dell) laptop. It's a machine with strong edges and protective technology. Built for gaming. I have one and only use it during travel, or as a secondary machine for backtesting. Kids and wife are not allowed to touch my laptop or my PC, they got their own; which they broken the keys in the keyboard several times. Even attached keyboards did not last with the kids.
Thanks. I also have a Dell laptop with the exact same config as MSi but with not the latest intel chipset (I think it ends in U). Soon, both laptops will be returned as CQG does not allow me to run backtests in parallel using multiple laptops. Most likely this brings my backtesting to closure in order to meet my timelines. We shall see. I never like to alter my plan so still TBD. I will respond to some other posts to cover this topic in detail.
Staying in my lane...
  • #243
  • Mar 13, 2019 3:00am Mar 13, 2019 3:00am
Quoting Divergence
{quote} Vee all I can say is WOW. You've taken a simple thing like trading (buy or sell) and complicated it so much in your mind that I am afraid you will never put on a live trade ever just saying
As I have said repeatedly, I can go live tomorrow and come out ahead with relative ease with or without my current state automation. Automation is a guidance tool to ignore when I am 100% focused and to enable when I run into work life distractions. It looks complicated. It sounds mumbo jumbo but at the crux of each individual strategy is nothing but simplicity. Perhaps it is on me to make it "sound" overly complex but it is not. I will share more soon why I say this.
Staying in my lane...
  • #244
  • Mar 13, 2019 3:43am Mar 13, 2019 3:43am
Quoting alphadude
{quote} I agree to some degree that Vee tends to complicate the algos little bit. i.e. you should keep each strategy simple and do one thing only. like range trading, pattern trading, breakouts, etc. once you get one strategy working profitably, launch it, and in your spare time, research more strategies. it will be easier to maintain, diagnose problems, improve bits and pieces, and throw away dead strategies over time. kind of strategy factory. dont try to put them all in one piece.
You can absolutely say that about my past FX strategy. That was beyond complex BUT it was designed to be a highly scalable, redundant and load balanced principles with active/passive failover built right within the model of correlated pairs that is NOT visible to the naked eye. EVERYTHING is do is based on sound and proven underlying principles or I would not touch it at all. That's the level of standard I use to gauge my own involvement.
With that said, let's do a quick comparison between MT4 and Sierra. if you are running say 8 reversion strategies on a intraday basis, each strategy written as it's own EA to turn on/off as you say so. Let's assume each EA have 15 global parameters, 5 input parameters (20 in all) so 20X8=160 parameters. In Sierra there is no such thing as "global parameters". Only studies. Just to display the text fields you see on my chart with values, it requires 2 studies just to display it. One where the actual value is set and another one to display that value on the chart. The Display study screenshot shows I have 37 such display items so 37+37=74 studies of 200 just to show the darn values on the chart (NOT required but I need the associated IDs to assist me with my formulas)
Next comes the "Tools" category, draw a line or stepladder or angle or rectangle or HHLL oe EMA(1) to use in studies or spread or volume or whatever... each and every tiny thing requires it's own study (study = indicator in MT4, study collection = Template in MT4). I have 25 studies categories as "tools". Even the darn clock you see on my screen requires a study to be added in Sierra.

Next category is "Account" which are also series of standard studies to keep track of trading performance. I got 8 of those. So 74+25+8=107 studies out of 200 that have not much to do with any specific strategy!
Next is OC specific market conditions... I got only 8 studies for that to capture ALL market conditions using nothing more than a simple moving average study. Mind you, Sierra provides 100s of delivered studies. they are all junk and I have ruled most of them out via the rigor of "process of Elimination"

Next is Strategies....As everything is done via Alerts and formula that triggers the alert condition, each "trade generating signal" requires 3 studies for Buy Entry/SL/TP and 3 studies for Sell so 6 studies per signal or 6*8=48 studies across 8 strategies. That brings the count to 155

Next is Trade Management. Since all my entries have default attached order settings, I apply brute force to my attached orders (something I am not sure if my broker would like... time will tell) I have one of SL/TP per DFM and one per RTM so 4 in all. Total Count = 159
The remainder of 200-159=41 are just miscellaneous and some related to time classifications, time stops, flatten, reverse and many are just required for backtesting only in relation to when bars are filtered in Sierra under Chart settings

So why all this shit is not complex? Here's an example of an excel type formula that triggers my entry condition. Even a high school kid can understand what I have do if he is familiar with Excel. It appears complex. I make it sound complex perhaps but the strategy themselves are extremely rudimentary and "simple"

Example:

=AND(ID148.SG1<>0,ID73.SG1=0,ID69.SG5>333,ID182.SG51<>2
,OR(ID61.SG1=0,AND(ID61.SG1<>0,OR(AND(ID80.SG1<>0,ID61.SG1<ID80.SG1),AND(ID92.SG1<>0,ID61.SG1<ID23.SG1*ID92.SG1),AND(ID84.SG1<>0,ID61.SG1>ID84.SG1),
AND(ID103.SG1<>0,ID61.SG1>ID23.SG1*ID103.SG1))))
,BARTIME<=TIME(08,0,0)
,AND(C>ID13.SG1,C>ID32.SG1,C>ID35.SG1,ID139.SG1>1.5*ID139.SG1[-1]))

Color Legend:
Strategy Trigger Condition
Daily PnL target met Condition... this mimics "Stop Trading" and required only for Backtesting. For live, I could use delivered Sierra function.
Basic Time Filter
Entry Setup >>> This one line is my entire Strategy! How is this complicated ?

Again, I focus on following the correct protocol essential for long term growth and success. "retail trader" mindset is very narrow focused. Mine is not. My focus is what happens to my algo when I am not around :-)
Sorry for the long post as I feel this was required for everyone to stop thinking I am building some space ship to travel to Mars. As I said, I have spent more than 70% of my time related to issues with Sierra and very little on actual strategy development as I knew exactly what I wanted to implement before the decision was made to use Sierra as my platform of choice. It's all about "Vision" and "Strategy" to me. Now, after doing everything that I know is proven to work in real life practice outside of trading, will I still succeed? I have no idea :-)
Staying in my lane...
  • #245
  • Mar 13, 2019 3:54am Mar 13, 2019 3:54am
Quoting VEEFX
{quote} Thanks Delta. To me, as you probably know from most of my bookmarked post.... TIME ACTION IS EVERYTHING. The relationship between Time and volatility/liquidity is strong and PREDICTABLE. Folks chase and value price. I VALUE Time and nothing else... in trading as well as in real life. Time is the only currency I understand :-) Please continue to share your wisdom bro.
youīre welcome. good luck on your path. will drop by from time to time
fortis fortuna adiuvat
  • #246
  • Mar 13, 2019 3:58am Mar 13, 2019 3:58am
Quoting hilmy83
hey vee just wanted to say thanks for emphasizing the Opportunity cycle theory. It really does resonate with me as I have observed this with my own trading. It's actually a good risk management tool since I have seen how my strategy breaks down after that initial profitable cycle. OC is integrated into my strategy now.
Glad you got it. A guy who can solve rubix cube has to get this":-) I did mention once that you are doing this shit every day in your trading and not realize it. Even Harold does it every day with his one hour OC and not willing to accept it. Actually, Harold is successful because of OC because, unknowingly, he knows WHEN TO STOP.... and it is not 100% based on his PnL. He knows what he knows that comes from deep experience that he is unwilling to accept. If one can decipher his trading stats, OC principles are written all over it... just my guess. I will also say, the recency+Decay principle does not apply to micro scalps which is why I feel scalping offers the Best risk to reward. I just don't like to wait for setup.... but that is on me and my psychological makeup profile. Scalping+Momentum offers the best recency/decay equation imo which is why the best scalps also occur during the first hour and Harold knows this also (unknowingly). I am just good at connecting the dots via spatial analysis and transforming that into core data points and then run interpretation models on it towards quantification. that is all there really is to this madness.
And, I will say this... OC is fractal by it's very nature because everything around her can be broken down into discrete dimensions of scale. what I have shared can be done on a single candle! what I have shared can be done even on your MAE/MFE on a trade by trade basis, on your vwap, on your poc, during asian/ny or even at the point of entry when a trade is executed. Even the Rubix cube has an inbuilt OC i.e. if you keep on aimlessly try to rotate the block, you will go nowhere. this is the very reason why some kid can solve the cube without even looking at it. he has mastered the "Opportunity Cycle Theorem", same theory also applies to poker or blackjack or any betting framework. Will you bet your house if the opportunity to win from your opponents is only a car? Doing the mumbo jumbo again so I will stop here
Staying in my lane...
  • #247
  • Mar 13, 2019 3:59am Mar 13, 2019 3:59am
Quoting profitfarmer
{quote} the process is (especially with Veefx) a rigorous, scientific, all detail oriented, and nearly never ending backtest, chasing the illusion of perfection. checkmark. for some (me) trading is about the money. not the perfection. not even really when to take a signal or when not too. more about if a trade taken how to manage the situation for best. think of a drought, or flood, neither part of the "process"...and both part of the farming. not really much in textbooks about how to try to salvage the crop from either, when teaching you about...
It's 2am here so hopefully my recent posts will shed further light.
Staying in my lane...
  • #248
  • Mar 13, 2019 4:09am Mar 13, 2019 4:09am
One of my bucket list is going to come to fruition soon. I have booked my first cruise to the Bahamas. Should be fun so looking forward to it or I will jump off the darn ship if I am surrounded by 80 year old grannies. We opted against Barcelona as our daughter is already there for spring break partying with her friends :-

Since I turned 50, goal is to make 3 international vacations every year. One down, second one is in May after our daughter's graduation ceremony (probably Iceland or Florida) and then hopefully China or Brazil in Nov/Dec. Travel is the only reason why I am involved in this trading endeavor to rely on when I retire. Travel is the only reason why I have taken the algo route. Freedom is the ultimate form of getting high for me.
Staying in my lane...
  • #249
  • Mar 13, 2019 4:54am Mar 13, 2019 4:54am
Quoting profitfarmer
{quote} the process is (especially with Veefx) a rigorous, scientific, all detail oriented, and nearly never ending backtest, chasing the illusion of perfection. checkmark. for some (me) trading is about the money. not the perfection. not even really when to take a signal or when not too. more about if a trade taken how to manage the situation for best. think of a drought, or flood, neither part of the "process"...and both part of the farming. not really much in textbooks about how to try to salvage the crop from either, when teaching you about...
Not trying to get combative, especially not in this thread. Smart people hang out here. That would include you. So, here is my retort. Drought and flood actually ARE in the plan for ag science, at least here in the states. We had an event called the Dust Bowl in our history. Farmers today rotate crops because of this. In flooding, especially along the Mississippi, farmers do the best they can to protect their crops. There are algorithmic traders and discretionary traders. Divergence and hilmy are both discretionary traders. They have enough seat time where they know what a good bar looks like by feel. Some folks do not trust this, myself included. I have gone through FF and quantified every worthy system out there. And I still to this day trade small. In my case however, those small trades will accumulate quickly, as I am an bar by bar trader and will start buying or selling closes after a certain amount of time. I can safely say that VEE will post his rigor and we will benefit from it. I will say again PF, trading is a process. The money comes from following the process. It is a process for you too, but you have sublimated it.
If you are good at something, never do it for free--Joker
  • #250
  • Mar 13, 2019 12:39pm Mar 13, 2019 12:39pm
Quoting VEEFX
{quote} You can absolutely say that about my past FX strategy.....
Good stuff Vee, I feel your pain. been there with MT4, and got sick of it. For simple strategy is ok, but when you try to manage the risk across strategies, or run a portfolio of strategies it gets messy; even with global variables. The other thing is debugging. (nowadays MT4 have added debugging feature however).

In the world of automated trading, the beast is in the automation, and bugs, lots of bugs, that can cost $$$.

if you know object oriented programming; your best chance is to write a backtester yourself; and that way you can have any feature you like. Just do a simple reading of tick by tick from a flat file; and go through the strategy logic each time. It's actually simple. the next phase is to connect to market data feed, and to order placement api, and there you go.
  • #251
  • Mar 13, 2019 1:23pm Mar 13, 2019 1:23pm
Quoting alphadude
{quote} Good stuff Vee, I feel your pain. been there with MT4, and got sick of it. For simple strategy is ok, but when you try to manage the risk across strategies, or run a portfolio of strategies it gets messy; even with global variables. The other thing is debugging. (nowadays MT4 have added debugging feature however). In the world of automated trading, the beast is in the automation, and bugs, lots of bugs, that can cost $$$. if you know object oriented programming; your best chance is to write a backtester yourself; and that way you can have...
I have a tremendous amount of ideas and innovations in my bucket list of things to achieve during my early retirement. My mind will go rot in a day of nothingness so all this is part of my strategic planning during the first 11 days of the start of each decade of my life cycle :- "The best is yet to come" is what keeps me motivated lol

For those who really like to see "simple charts", heres one from today's session. I was working until 3am and could not get up in time so market replay will do for now.

EDIT: Further analysis of this chart reveals the brute force nature of the application of multiple long-only strategies during the excursion of the coast line i.e. SF up north PCH towards portland. The nook and crannies and the potholes and the devil slides and the tunnels etc etc prevented any of my strategies to trigger.... i think or perhaps it is a bug. God knows! Once the Short only design works, all that is required (I hope) is a simply assessment of HHHL to trigger long only and LLLH to trigger short-only strategies. Then will work will reap some long term fruits atleast in the first 2 hrs of each session (OC)
Attached Image (click to enlarge)
Click to Enlarge Name: Image2.png Size: 211 KB
Staying in my lane...
  • #252
  • Mar 13, 2019 1:41pm Mar 13, 2019 1:41pm
Quoting alphadude
{quote} Good stuff Vee, I feel your pain. been there with MT4, and got sick of it. For simple strategy is ok, but when you try to manage the risk across strategies, or run a portfolio of strategies it gets messy; even with global variables. The other thing is debugging. (nowadays MT4 have added debugging feature however). In the world of automated trading, the beast is in the automation, and bugs, lots of bugs, that can cost $$$. if you know object oriented programming; your best chance is to write a backtester yourself; and that way you can have...
+ Sierra is designed to backtest in market depth so perhaps it is my installation back from November that has bloated some log files etc causing backtesting to slow down. I just created a new instance within Sierra and that seems to have some improved performance. Frankly, in my design, backtest results are absolutely meaningless. I am using it only to test/debug my strategies. Let me explain more since you are equipped to understand imo.

Even though I am using M1 (which in my opinion is the only universal candle... instead of the well known H1 candle), everything is tied to tick. When I say opportunity cycle, I do NOT mean let's look at the past 10 day ATR blah blah... it is all 100% tied to the current tick by tick resulting in the opportunity which creates the cycle to tap into. When I say, I want to target 10% of OC. I mean 10% of the range of the OC created by the last tick since the start of OC (which in my case, for simplicity, is one minute from start of NYSE). When Hilmy says, he targets 20% of range, there is probably some current or last day ATR involved. My design will kick off at the first tick of OC and not give 2 shits about what price did yesterday or during the active session. It's all in the moment. Every decision is made on the last tick which conforms to the OC Framework. If the range grows or expands, ideally, my target also expands with it based on the % multiple setting. but not perfect on slow trending ranges but still, the foundational elements of design is to squeeze the max opportunity from any cycle. And because there is ALWAYS a decay as OC progresses (this is universal to any OC!), you are forced to take one side of the cycle and then grab what you can and get the fuck out. For that, I have a fixed daily TP goal and testing shows 90% of days will give me atleast 50 ticks BUT those 10 days will drain the account as well.... which also confroms with the universal truth told by CP that "Any strategy that takes every long and short signal is a failed strategy over a long period of time".

Sorry... I get carried away :-)
Staying in my lane...
  • #253
  • Mar 13, 2019 2:32pm Mar 13, 2019 2:32pm
Quoting VEEFX
{quote} + Sierra is designed to backtest in market depth so perhaps it is my installation back from November that has bloated some log files etc causing backtesting to slow down. I just created a new instance within Sierra and that seems to have some improved performance. Frankly, in my design, backtest results are absolutely meaningless. I am using it only to test/debug my strategies. Let me explain more since you are equipped to understand imo. Even though I am using M1 (which in my opinion is the only universal candle... instead of the well known...
Nice. These ideas are similar to CP's "Edge in a single candlestick"
  • #254
  • Mar 13, 2019 2:36pm Mar 13, 2019 2:36pm
Quoting alphadude
{quote} Nice. These ideas are similar to CP's "Edge in a single candlestick"
yep....It replicates my own understanding of his 6 dimensions, buy/sell zone etc. I have a stop loss at median and then the edges of OC. Everything inbetween is considered part of a low/mid/high strength dimension of varying scales.
Staying in my lane...
  • #255
  • Mar 13, 2019 7:00pm Mar 13, 2019 7:00pm
Unbelievable !!!!! My new instance of Sierra using File > "New Instance" is now super fast. The "Calculate at Every Tick" backtesting mode is awesome taking only few minutes for one full NYSE session worth of backtest. The chart scrolls smoothly without freezing at 240 replay speed (which would freeze at just 5 speed in the primary instance! Per support, Freezing of charts is normal but not for me... I need to see lol)

I am back in business baby! Phew - very excited now to run all my test in real backtest instead of fake standard replay. Quite the blessing as far as timing of this goes :-

The 64-bit CPU prioritization setting in MSi gaming laptop + running in "Turbo" mode is awesome. Keeping this laptop for now and getting rid of Dell as Dell is only 14inch...I have gotten used to the big monitor now.
Staying in my lane...
  • #256
  • Mar 15, 2019 3:14am Mar 15, 2019 3:14am
It happened again! I lost another 2-page long post on my next 3 research projects by hitting the browser refresh by mistake. I am pissed !

Oh well.. it happened for a reason :-(
Staying in my lane...
  • #257
  • Edited 2:26am Mar 20, 2019 1:43am | Edited 2:26am
I shall prevail in developing a strategy that will test the end of time. It will handle any market conditions. It will accommodate all trading styles. It will accommodate all shock absorbers during new releases. It will handle ranging markets. It will handle trending markets. It will incur drawdown and it will recover from it during start until the end of each OCF ending in positive expectancy... All with only one trade and one contract open at a time with no regard to RM/MM/TM/PM using no influence other than pure price vs time (and volume if necessary)

Objective: Net Positive "End of OC" PnL @ 80+% frequency across each accounting period (per day/week/month/quarter/year) with RoMAD of atleast 2 exploiting atleast 50-80% of range across OCF or 50ticks

Longer term Research Project/Ideas:

  1. Using OCF and conformity to OC Principle as a "Edge Discovery Tool"
  2. Using OCF and conformity to OC Principle to evaluate any system strategy for long term positive expectancy
  3. Using OCF to pinpoint transition between Trending and Ranging market conditions
  4. Mandlebrot, Tesla, Bill Williams/Chaos. Is Leonardo Da Vinci next ? or Quantum Physics (waiting for my son to get to text book). I need to quantify the measurement of speed.

Staying in my lane...
  • #258
  • Mar 24, 2019 4:41am Mar 24, 2019 4:41am
Longer term Research Projects:

  1. Using OCF and conformity to OC Principle as a "Edge Discovery Tool" - Complete
  2. Using OCF and conformity to OC Principle to evaluate any system strategy for long term positive expectancy - Complete (my beliefs are confirmed)
  3. Using OCF to pinpoint transition between Trending and Ranging market conditions (In Progress - This is tough!)
  4. Mandlebrot, Tesla, Bill Williams/Chaos. Is Leonardo Da Vinci next ? or Quantum Physics (waiting for my son to get to text book). I need to quantify the measurement of speed. No need to read more. (Complete) I am continuing my path to form/develop my own theories without any bias from what is out there!
  5. How many trades does it take to always end up in profit at the end of OC (Criteria: only one contract at a time, only one trade open at a time, any combination of SL/TP/BE/nothing at all, no scale-in or scale out (i.e no adding to losers or winners)
  6. How many OCs does it take to generate 100% of $$ profit using only one contract per $2000 of trading capital and using same rules as #5 above.
  7. Coastal to Inland non-stop traversal of price thru time (maximum positive expectancy using maximum number of opportunity cycles)

Staying in my lane...
  • #259
  • Mar 24, 2019 5:16am Mar 24, 2019 5:16am
Interesting (but old) post from Sierra Charts' POV on the future of decentralized futures trading.

https://www.sierrachart.com/SupportB...139357#P139357
Staying in my lane...
  • #260
  • Mar 24, 2019 10:29am Mar 24, 2019 10:29am
Great reading this thread. Ive been doing something very similar to what you are doing here. I’m taking my own path but my work has many parallels to yours. The process is very challenging. I have a love hate relationship with break even (I usually disable).
0 traders viewing now