NexusFi: Find Your Edge


Home Menu

 





MultiCharts lag time?


Discussion in MultiCharts

Updated
      Top Posters
    1. looks_one bobbakerr with 63 posts (24 thanks)
    2. looks_two Big Mike with 19 posts (19 thanks)
    3. looks_3 Jura with 11 posts (8 thanks)
    4. looks_4 Bimi with 10 posts (7 thanks)
      Best Posters
    1. looks_one bretter with 4 thanks per post
    2. looks_two RM99 with 2.3 thanks per post
    3. looks_3 Big Mike with 1 thanks per post
    4. looks_4 bobbakerr with 0.4 thanks per post
    1. trending_up 56,350 views
    2. thumb_up 144 thanks given
    3. group 23 followers
    1. forum 171 posts
    2. attach_file 2 attachments




 
Search this Thread

MultiCharts lag time?

  #111 (permalink)
 RM99 
Austin, TX
 
Experience: Advanced
Platform: TradeStation
Trading: Futures
Posts: 839 since Mar 2011
Thanks Given: 124
Thanks Received: 704


bobbakerr View Post
Thank you for your input, RM99. But I don't think this will solve my problem.

First, the average number of R.T.'s using ES (which I am concentrating on now) is only around 150 per day. CL was more than double that. Secondly, Limit orders didn't really help. It was the Lag time between WHEN the signal was given to Enter by MultiCharts and when it SHOULD have given the signal.

And speaking of Limit orders, I always thought a limit order was a limit order was a limit order. How can an Entry be made at a price worse than the limit order?? But it has happened. My last attempt at the live trading used only Limit orders. And I was filled at prices much worse than my limits.

Something was either wrong with my code or with MultiCharts. That's why I have this thread here. There have been some small errors in my code, but I think the bulk of the problem is with MC. I'm hoping against hope that Version 7 will eliminate these bugs. But to tell you the truth, I doubt that it will.

You focused on my second point, but the first point was more cogent. I've been down the path you're on. I gave up.

One of the most significant performance metrics is "netprofit/trade."

it tells you how "efficient" your strategy is with respect to the profit it makes.

I have some old strategies that generated NICE profits and curves, very stable, low drawdown, etc, but the netprofit/trade was low enough to either wipe out most of the profit, or make it very choppy, high drawdown, etc.

I would suggest, when you come to the end of thie evolution you're undergoing (which many of us have) you start seeking trading strategies that trade less frequently, and make more profit/trade.

Secondly, as you progress through automated strategies, you'll find that you begin using fixed P/L amounts. Then over time, when you discover how to successfully trade with variable profit/loss amounts, you'll discover a MUCH HIGHER correlation between backtests and forward or live trading.

For example....if I'm able to successfully identify trend with some degree of certainty (not very difficult) then I have an edge.

I can enter the trades and optimize a backtest using fixed profit and fixed loss amounts. Hopefully, if all goes well, I can profit at a 2:1 winning trade profit/losing trade loss ratio and some win% of 40% or better. (but as you'll discover, this is very difficult to do as it relies heavily upon the entry as the primary edge).

Then, at some point I discover how to identify the strength of the trend or at least get some sense of strong trends and weaker trends. Then, I can begin to vary my positionsizes and profit goals accordingly.

Then, you evolve even further, where you begin to identify turning points in the market. So you're able to identify a trend, enter with the grain...and ride the trend all the way until the market turns, at which point you exit. Maybe that's a big gain, maybe it's a small gain...but it's a gain. You still feature a static stop loss, and maybe you combine it with some second order loss limit (like a tiered stop, hyperbolic trail, etc).

What you discover, at the end of this journey, is that the strategy that identifies turning and pivot points in the markets, and enters and exits on those, is very HIGHLY correlated, because it does a number of things.

1) It almost certainly solidifies that your wins/losses in backtest/historical data are identical to live trading. (you might observe SOME variance if you're using a fixed stoploss, but for the most part, a winner will be a winner and a loser a loser).

2) The variance with respect to slippage and entry/exit efficiency and platform execution, etc....will be what varies. But because you're trading larger trends, these variances are much smaller and have much less effect.

So in essence, if a trade entered upon a market turning point, and ended on a market turning point, the real variance between backtest and live, would be how much (if any) slippage ocurred between the two transactions, and any adverse market effects your order may have created (essentially, any support/resistance a limit order(s) created or any additional movement that a market order may have generated). Again, these issues on a large, voluminous instrument like CL are negligible. Especially during RTH. You REALLY have to be a whale in order to alter the market perceptibly during RTH CL trading. If you're trading a stock with a small share float, or an instrument off hours, then the variance is more pronounced. (that's a 3rd point I chose not to add...that most of the time, traders evolve to where there auto strategies only trade during RTH, because the inherent limitations still exist with respect to "hands free" trading on automated systems, the inherent risks, increased slippage, etc, for off hours trading).

Essentially, this whole process is a learning and growth process by trial and error. You try something, maybe it works partially, you retain the successful part, and seek to replace the failing part with something new.

I'll sum it all up for you.

1) Trade RTH.
2) Use variable profit limits that allow you big or bigger wins (which leads to a P/L ratio of 2:1 or better).
3) Use sophisticated money management techniques like trailing stops, tiered stops, etc...but ALWAYS LIMIT YOUR DOWNSIDE LOSS and keep your upside wins unlimited.
4) Avoid strategies that win by a thousand paper cuts. Start seeking an edge based off larger moves. HFT is for guys that are directly plugged into the exchange or running their own tradeserver/market maker capability.

"A dumb man never learns. A smart man learns from his own failure and success. But a wise man learns from the failure and success of others."
Reply With Quote

Can you help answer these questions
from other members on NexusFi?
The space time continuum and the dynamics of a financial …
Emini and Emicro Index
Build trailing stop for micro index(s)
Psychology and Money Management
ZombieSqueeze
Platforms and Indicators
Online prop firm The Funded Trader (TFT) going under?
Traders Hideout
My NT8 Volume Profile Split by Asian/Euro/Open
NinjaTrader
 
Best Threads (Most Thanked)
in the last 7 days on NexusFi
Get funded firms 2023/2024 - Any recommendations or word …
59 thanks
Funded Trader platforms
37 thanks
NexusFi site changelog and issues/problem reporting
23 thanks
GFIs1 1 DAX trade per day journal
22 thanks
The Program
19 thanks
  #112 (permalink)
 RM99 
Austin, TX
 
Experience: Advanced
Platform: TradeStation
Trading: Futures
Posts: 839 since Mar 2011
Thanks Given: 124
Thanks Received: 704

Some other typical "evolutions."

4) Eventually in your quest to seek larger moves and larger profit/trade strategies, you'll discover (if you haven't already) that regular time based charts are tough. To improve and focus mainly on price action, you'll start using (if you haven't already) advanced bar types like Range bars and Renko bars. The Renko with tails is kinda the holy grail of charting right now and I'm envious of those that have it. You and I on TS/MC will be waiting for sometime I think.

But nontheless, you'll start to learn on this board how to craft strategies that are more "timeless" that withstand market fluctuations. You do this by focusing on volume and price action and you try to remove time altogether.

It's much easier to identify "trends" on advanced bars (and turning points) than regular time based bars. But as I stated in the thread about limit orders...it comes at a price. With TS, there are limitations to how far you can backtest, what type of backtests you can conduct, etc.

"A dumb man never learns. A smart man learns from his own failure and success. But a wise man learns from the failure and success of others."
Reply With Quote
Thanked by:
  #113 (permalink)
 
bobbakerr's Avatar
 bobbakerr 
Riverdale, Idaho, USA
 
Experience: Intermediate
Platform: Optimus Futures + MultiCharts + TradeStation
Broker: Optimus Futures, Rithmic Data, TradeStation
Trading: CL, ES
Posts: 115 since Aug 2010
Thanks Given: 241
Thanks Received: 51


You gave great advice, RM99. Thanks.

I just don't think I can write a program to trade that way. To get the bigger runs, I would think things would have to be done manually. I have traded manually for many years. And I'm tired. I'm looking for a way to trade automatically so that there is no longer the continuous stress throughout the day. Perhaps, like you say, that type of program is only for those who are 'directly plugged into the exchange'. If that's the case, then I may be wasting my time looking for an automated strategy.

But I'll keep looking. And maybe I'll be back to trading manually. At least I get my orders in and out timely that way.

Started this thread Reply With Quote
  #114 (permalink)
 RM99 
Austin, TX
 
Experience: Advanced
Platform: TradeStation
Trading: Futures
Posts: 839 since Mar 2011
Thanks Given: 124
Thanks Received: 704


bobbakerr View Post
You gave great advice, RM99. Thanks.

I just don't think I can write a program to trade that way. To get the bigger runs, I would think things would have to be done manually. I have traded manually for many years. And I'm tired. I'm looking for a way to trade automatically so that there is no longer the continuous stress throughout the day. Perhaps, like you say, that type of program is only for those who are 'directly plugged into the exchange'. If that's the case, then I may be wasting my time looking for an automated strategy.

But I'll keep looking. And maybe I'll be back to trading manually. At least I get my orders in and out timely that way.

If you trade discretionary and it's based on any set of rules, it can be coded and automated. Some rules are more difficult than others and may require some very skilled programmers, but it can be done.

If you're successful at discretionary, why don't you simply try to replicate it and automate it?

"A dumb man never learns. A smart man learns from his own failure and success. But a wise man learns from the failure and success of others."
Reply With Quote
Thanked by:
  #115 (permalink)
 
bobbakerr's Avatar
 bobbakerr 
Riverdale, Idaho, USA
 
Experience: Intermediate
Platform: Optimus Futures + MultiCharts + TradeStation
Broker: Optimus Futures, Rithmic Data, TradeStation
Trading: CL, ES
Posts: 115 since Aug 2010
Thanks Given: 241
Thanks Received: 51

I'm a scalper, and what I've done I'm trying to code (and improve on). However, a lot of trades were done by the 'seat of my pants'. That can't be programmed.

Started this thread Reply With Quote
  #116 (permalink)
 RM99 
Austin, TX
 
Experience: Advanced
Platform: TradeStation
Trading: Futures
Posts: 839 since Mar 2011
Thanks Given: 124
Thanks Received: 704


bobbakerr View Post
I'm a scalper, and what I've done I'm trying to code (and improve on). However, a lot of trades were done by the 'seat of my pants'. That can't be programmed.

1) Yeah, if it's seat of pants, then that's not rules based.

2) You say you're a scalper, but there's no way you were trading nearly the number of round trips a day were you? (not manually/discretionary) that you're system is now trading automated?

"A dumb man never learns. A smart man learns from his own failure and success. But a wise man learns from the failure and success of others."
Reply With Quote
  #117 (permalink)
Bimi
London
 
Posts: 118 since Mar 2010
Thanks Given: 42
Thanks Received: 58


bobbakerr View Post
...
Something was either wrong with my code or with MultiCharts. That's why I have this thread here. There have been some small errors in my code, but I think the bulk of the problem is with MC. I'm hoping against hope that Version 7 will eliminate these bugs. But to tell you the truth, I doubt that it will.

Bashing MultiCharts won't score you any brownie points.

The mechanical process of sending a buy/sell order to the broker is very simple and straight forward. I doubt you will find any bugs there. MC has been around for ~5 years, there are thousands of traders using it everyday, including many from big trading houses. If there is a bug in the order delivery code, you will not be the first one to discover it.

There are bugs in any software; for MC, most of the bugs are in charting and manual trading (new), but I doubt you will find much in trading. Because charting, which is basically presentation, are much more involved then trading.

The complex part of autotrade is not in sending the orders to the broker,
but in the logic that triggers the orders. I will take a look at your simplified code in the next few days, and let you know my findings. What you are trying to do is HFT... Time is of essence. You really need to drill down to the micro time slices and and have all the stars aligned to make it work. You might think your concept is simple, but we all know that human mind is sophisticated, we can process complex concepts with ease and assume the execution is simple.

More later...

Reply With Quote
  #118 (permalink)
 
bobbakerr's Avatar
 bobbakerr 
Riverdale, Idaho, USA
 
Experience: Intermediate
Platform: Optimus Futures + MultiCharts + TradeStation
Broker: Optimus Futures, Rithmic Data, TradeStation
Trading: CL, ES
Posts: 115 since Aug 2010
Thanks Given: 241
Thanks Received: 51


RM99 View Post
2) ... You say you're a scalper, but there's no way you were trading nearly the number of round trips a day were you? (not manually/discretionary) that you're system is now trading automated?


For sure I wasn't! I'd have to 'rest' between trades. And I tried to pick 'winners only'. But that's really impossible. I don't think anybody knows for sure what price is going to do.

And in this programming now, I'm trying to go for longer runs. That's why I let it go from Signal to Signal. I felt it was better to chance it with more losers to get the bigger profits.

Started this thread Reply With Quote
  #119 (permalink)
 
bobbakerr's Avatar
 bobbakerr 
Riverdale, Idaho, USA
 
Experience: Intermediate
Platform: Optimus Futures + MultiCharts + TradeStation
Broker: Optimus Futures, Rithmic Data, TradeStation
Trading: CL, ES
Posts: 115 since Aug 2010
Thanks Given: 241
Thanks Received: 51


Bimi View Post
Bashing MultiCharts won't score you any brownie points.

Bimi, you're really a straight-forward person! And no, I'm not scoring points with MultiCharts. But I sincerely believe that they have major flaws in their system. When I see my signals inserted back in time by as much as a good hour (!), and then these results are included in the backtesting results, then MC deserves to be 'bashed'. If a dummy like me can see these things, then for sure they can. And problems like this should be fixed!

Now you and some others say I should program work-arounds for things like this. That's what I'm trying to do now. However bugs come up even in these! I may not be the world's greatest programmer, but I do have some brains.


Quoting 
The mechanical process of sending a buy/sell order to the broker is very simple and straight forward. I doubt you will find any bugs there.

True. The orders do get sent. But it's the Time Lag that is killing me. This should not be!


Quoting 
The complex part of autotrade is not in sending the orders to the broker,
but in the logic that triggers the orders. I will take a look at your simplified code in the next few days, and let you know my findings. What you are trying to do is HFT... Time is of essence. You really need to drill down to the micro time slices and and have all the stars aligned to make it work. You might think your concept is simple, but we all know that human mind is sophisticated, we can process complex concepts with ease and assume the execution is simple.

More later...

You're exactly right about the complex part being in the logic that triggers the orders. And I sure hope you can find mistakes in my code to make the logic right. I know there's something good here. I just can't get it to work properly with MC.

And by the way. Even with all the bugs that I say MC has, I still think they're the best charting service out there. They can do things that no others can. I just want them perfect. !

Started this thread Reply With Quote
  #120 (permalink)
Bimi
London
 
Posts: 118 since Mar 2010
Thanks Given: 42
Thanks Received: 58



bobbakerr View Post
Bimi, you're really a straight-forward person!

is there another way to approach a problem?


bobbakerr View Post
And no, I'm not scoring points with MultiCharts. But I sincerely believe that they have major flaws in their system. When I see my signals inserted back in time by as much as a good hour (!), and then these results are included in the backtesting results, then MC deserves to be 'bashed'. If a dummy like me can see these things, then for sure they can. And problems like this should be fixed!


bobbakerr View Post
Now you and some others say I should program work-arounds for things like this. That's what I'm trying to do now. However bugs come up even in these! I may not be the world's greatest programmer, but I do have some brains.

you must have confused myself with some other person.


bobbakerr View Post
True. The orders do get sent. But it's the Time Lag that is killing me. This should not be!

I have not looked at your code, so I can't comment on the location of the problem.
For the pages I have read here, I do observe a few things:
1. you do not have a flow chart (ie you do not have a clear view of the whole picture)
2. you have not tested your logic in an indicator. ie. there is no certainty that your logic is firing at the time and places that you expected them to be. So much for the claim that the order was sent an hour late !
3. If I ask for your pseudo code, probably you do not have it either. Not that it is a must, but will show you how you have came to code your logic.
4. If I ask for a variable map, probably you will look at me with a blank face.
5. If I ask for an audit print out, what can you offer me? do you have the audit trail for the order that was an hour late?

I am not saying these to dis you.
I am merely describing the observations -- you have not done your due diligence,
and hope you realize that walking with a big stick is not going to get you anywhere any faster.
You have to start with your code... go through each step with a fine tooth comb.
you have to manually calculate each variable and condition... under all permutations, and to verify their applicability.
Before you have done the above, calling MultiCharts to the carpet will not make you look pretty.

There are debugging step you must make.
There are audit trails you must trace.
If you like, I can take you through them step by step.
It is not particularly difficult to do, but
it is not going to be quick.
You must take your time (and care) to walk through the code
and apply them to all the possible real life scenario.


I am in a good mood today, I did well, so I talked more. Maybe too much.

Later...

Reply With Quote




Last Updated on April 4, 2012


© 2024 NexusFi™, s.a., All Rights Reserved.
Av Ricardo J. Alfaro, Century Tower, Panama City, Panama, Ph: +507 833-9432 (Panama and Intl), +1 888-312-3001 (USA and Canada)
All information is for educational use only and is not investment advice. There is a substantial risk of loss in trading commodity futures, stocks, options and foreign exchange products. Past performance is not indicative of future results.
About Us - Contact Us - Site Rules, Acceptable Use, and Terms and Conditions - Privacy Policy - Downloads - Top
no new posts