Cepren - torgovyy den 'mozhet byt' uspeshnym, no zanimayet mnogo masterstva. takzhe otmetit ', etot forum angliyskiy tol'ko mnogo bol'shinstvo ne mozhet ponyat' vas.

Unfortunately, we rarely see what's really going on, whether we are male or female. Most often we see what we want to see.

As for intraday trading. In my opinion it can be compared with the direction of the wind during high tide or low tide the sea. Try to guess !!!

Day trading is the beggar's attempt to participate in "big guys" markets. If you close your positions before the end of regular trading hours, many brokers allow you to use as low as 1/10th of the exchange's initial margin. This opens doors to "the world of futures trading" to many under- or at-$1000 account holders. There's nothing else to day trading. It's neither art, nor science, it's just stupidity.

James - just curious what the average trade profit loss is? Mark

TradeStation Performance Summary Expand

All Trades

Long Trades

Short Trades

Total Net Profit $69,950.00 $40,217.50 $29,732.50

Gross Profit $100,287.50 $60,222.50 $40,065.00

Gross Loss ($30,337.50) ($20,005.00) ($10,332.50)

Profit Factor 3.31 3.01 3.88

Roll Over Credit $0.00 $0.00 $0.00

Open Position P/L $97.50 $97.50 $0.00

Select Total Net Profit $59,520.00 $33,245.00 $26,275.00

Select Gross Profit $88,940.00 $52,332.50 $36,607.50

Select Gross Loss ($29,420.00) ($19,087.50) ($10,332.50)

Select Profit Factor 3.02 2.74 3.54

Adjusted Total Net Profit $59,283.68 $31,976.91 $22,949.69

Adjusted Gross Profit $92,232.21 $54,139.11 $34,758.26

Adjusted Gross Loss ($32,948.54) ($22,162.19) ($11,808.57)

Adjusted Profit Factor 2.80 2.44 2.94

Total Number of Trades 290 184 106

Percent Profitable 53.45% 53.26% 53.77%

Winning Trades 155 98 57

Losing Trades 135 86 49

Even Trades 0 0 0

Avg. Trade Net Profit $241.21 $218.57 $280.50

Avg. Winning Trade $647.02 $614.52 $702.89

Avg. Losing Trade ($224.72) ($232.62) ($210.87)

Ratio Avg. Win:Avg. Loss 2.88 2.64 3.33

Largest Winning Trade $5,295.00 $5,295.00 $3,457.50

Largest Losing Trade ($917.50) ($917.50) ($892.50)

Largest Winner as % of Gross Profit 5.28% 8.79% 8.63%

Largest Loser as % of Gross Loss 3.02% 4.59% 8.64%

Net Profit as % of Largest Loss 7,623.98% 4,383.38% 3,331.37%

Select Net Profit as % of Largest Loss 6,487.19% 3,623.43% 2,943.98%

Adjusted Net Profit as % of Largest Loss 6,461.44% 3,485.22% 2,571.39%

Max. Consecutive Winning Trades 8 9 5

Max. Consecutive Losing Trades 7 7 6

Avg. Bars in Total Trades 224.79 225.82 223.01

Avg. Bars in Winning Trades 298.63 293.61 307.25

Avg. Bars in Losing Trades 140.01 148.56 125.02

Avg. Bars in Even Trades 0.00 0.00 0.00

Max. Shares/Contracts Held 1 1 1

Total Shares/Contracts Held 291 185 106

Account Size Required $2,677.50 $1,935.00 $1,430.00

Total Slippage $7,250.00 $4,600.00 $2,650.00

Total Commission $1,450.00 $920.00 $530.00

Return on Initial Capital 699.50%

Annual Rate of Return 140.62%

Buy & Hold Return 8.69%

Return on Account 2,612.51%

Avg. Monthly Return $3,891.53

Std. Deviation of Monthly Return $3,366.01

Return Retracement Ratio n/a

RINA Index 3,136.78

Sharpe Ratio n/a

K-Ratio n/a

Trading Period 1 Year, 5 Months, 20 days, 23 hours, 22 Minutes

Percent of Time in the Market 8.45%

Time in the Market 1 Month, 14 days, 14 hours, 31 Minutes

Longest Flat Period 11 days, 17 hours, 50 Minutes

Max. Equity Run-up $73,437.50

Date of Max. Equity Run-up 12/14/2015 10:17

Max. Equity Run-up as % of Initial Capital 734.38%

Max. Drawdown (Intra-day Peak to Valley)

Value ($3,662.50) ($2,770.00) ($2,460.00)

Date 12/22/2015 11:26

as % of Initial Capital 36.63% 27.70% 24.60%

Net Profit as % of Drawdown 1,909.90% 1,451.90% 1,208.64%

Select Net Profit as % of Drawdown 1,625.12% 1,200.18% 1,068.09%

Adjusted Net Profit as % of Drawdown 1,618.67% 1,154.40% 932.91%

Max. Drawdown (Trade Close to Trade Close)

Value ($2,677.50) ($1,935.00) ($1,430.00)

Date 12/21/15 11:17

as % of Initial Capital 26.77% 19.35% 14.30%

Net Profit as % of Drawdown 2,612.51% 2,078.42% 2,079.20%

Select Net Profit as % of Drawdown 2,222.97% 1,718.09% 1,837.41%

Adjusted Net Profit as % of Drawdown 2,214.14% 1,652.55% 1,604.87%

Max. Trade Drawdown ($937.50) ($937.50) ($875.00)

Mark and group. Please understand that I have no need or want to get into a “contest”. The last two post was more or less to try to establish that

day trading can be a style/method of trading the markets.

James - have you traded this system in real time? There is a developer exploit in TS that allows really good results like this in testing but impossible to achieve in real world trading. I would make sure that limit orders, trailing stops floor limits are not to be used.

Also the use of swing bars and other historical look backs will not be obtainable in real time. The average trade net is far too big and the profit factor double what a fantastic system would have. Just be cautious.

But if your good to go on all the above I would love that have a system like that to trade. I would be trading the crap out of it like crazy.

James .. my issue was .. if Day trading is an sience / art/craft .. i asume that ..are same traders witch can trade by hand .. but this are just few .. i belive ... rest of us ..using algorithms :...

Mark, thank you for your very kind comments. I will be the very first to say 99% of automated strategies back test in TS does not perform well in live trading. It takes many years

of live testing algo’s in order to get even close to real world trading. The ES results posted is just part of a portfolio of many instruments the Master Strategy is capable

of trading. When the strategy is setup to trade the portfolio of instruments the ES results change slightly as then there are not any stops/profit targets or money management

per instrument(unless compounding of trade size per net profit) and all entries/exits for all instruments are on the very same bar.

The below results are trading a single contract per instrument of just these instruments in a portfolio environment.

START DATE END DATE SYMBOL OnePriceTick PriceScale NET PROFIT GROSS PROFIT GROSS LOSS P/L RATIO AVG TRADE PERCENT PROFITABLE TOTAL TRADES

* 12/29/2015 @ES.D 0.25 $100.00 $69,720.00 $98,982.50 -$29,262.50 3.38 $252.61 58.33% 276

* 12/29/2015 @TF.D 0.1 $100.00 $94,540.00 $146,020.00 -$51,480.00 2.84 $342.54 60.87% 276

* 12/29/2015 @NQ.D 0.25 $100.00 $67,670.00 $100,205.00 -$32,535.00 3.08 $242.54 63.80% 279

* 12/29/2015 @YM.D 1 $1.00 $55,805.00 $80,555.00 -$24,750.00 3.25 $202.19 59.06% 276

* 12/29/2015 @EMD.D 0.1 $100.00 $103,465.00 $153,225.00 -$49,760.00 3.08 $376.24 59.27% 275

* 12/29/2015 @NK 5 $1.00 $47,625.00 $77,500.00 -$29,875.00 2.59 $173.18 53.82% 275

* 12/29/2015 @VX 0.05 $100.00 $58,395.00 $90,060.00 -$31,665.00 2.84 $212.35 56.73% 275

* 12/29/2015 @FDAX 0.5 $10.00 $158,557.59 $284,619.66 -$126,062.07 2.26 $591.63 58.58% 268

cheers

James - the bouncing tick problem in TradeStation has long been a problem for system testers who entry and exit on the same bar because there is no way for historical testing to know in what order the trades occurred. TradeStation will always choose the most profitable results under those circumstances. TS tried to correct this bug with the look inside a bar feature but many found it lacking still and run their models on smaller time frame intraday charts so that specifically the entry and exit would not be on the same bar. But you sound like you know all this so I just wanted to post for others who might think they have found the holy grail using TS cause it's a hotbed for systems scammers thru the years and TS has been negligent keeping these dirty secrets from their clients.

Mark and group, I am truly sorry that I did not provide a more detailed description of the entry/exits and the relationship of bar time stamp.

There simply can not be a entry/exit on the same bar time stamp in my strategies.

Mark is totally correct in his comments about TS and bouncing ticks and the useless look inside bar feature. There are many of these issues

that have to be address in the code of the strategy using the TS platform. TS is not for an inexperienced developer no matter what they claim.

Back in the 90's I went to enormous lengths to collate historical tick data from all around the world. DTB sent me a big box of 1.44Mb floppies - lol... TSE wanted to charge me US$70k, so I got that data from a mate at Tudor... Much of the data was streaming ticks... such a mess.

I then brought in a statistician to work with my programmers to collate the data and test various intra-bar assumptions. Bottom-line, the bouncing tick idea was useless. That is, its use did not improve, but typically degraded the correlation of the bar simulated trades to the tick simulated trades.

After all the time and money the frustratingly simply answers was that the best way to compensate for intrabar whips was simply to add a cost to the simulation based on how often it generally occurred.

If you wanted to get more sophisticated then you could incorporate a volatility adjustment, in that the smaller the relative range of the day the more likely it is to rotate close to the extremes of the day, but as you go down the rabbit whole you find that the optimal intrabar assumptions then become dependent on your Strategy and how it generates its orders... Enjoy, and say hello to Alice and the Mad Hatter for me while you’re down there.

@ Mark ( i have to go back to the your T/A) My trading strategy / close on Friday ... or max < 3 days .. I dont know if I can say that this day trading strategy as long 89% of the trades close in the same day . . .

I was frustrated before I understand :

1. Tick charts are not disrupted with gaps in trading. Often the gaps can throw your indicators off course creating a discontinuity in the flow of the charting display.

2. Tick charts incorporate both time and price in their bars. This provides a more compressed consistent presentation of the chart patterns currently in formation.

3. This time and price combination allows for a more valid display of momentum when breakouts are about to occur giving an advantage over just time charts.

4. And tick charts allow for a smoother presentation of the key indicators you choose to use in your trading set ups.

@Mark --> swing bars can be use in case you have them /calculated... and use as patern --> inside your algorithm ..think about "Line Break" ..

Hello there, it was interesting to read about incorrect back testing results in TS, is anyone aware of a simular flaw being present using the MT4 strategy tester?

thanks.

Louis - unfortunately there is a dirty secret with almost all back testing platforms. MT4 is widely known for inherent flaws. Most popular software companies leave these bugs which allow better than obtainable results. Statistically they know most users will never actually know or test the platform for bugs and overly optimistic results are what the customers want to see and so the bugs perpetuate. There have been some famous class actions brought against many trading platform companies.

Louis... go as fast you can from MT4.. try to verify your algorithms in two different platform and brokers data feed.. this is the way to verify your system.

By 1999 I was already almost two years into the design of our new proprietary platform. It was too big for us to develop on our own though and then the bubble burst in 2000 and the hopes of VC disappeared in a blink.

My CEO at the time Steve Woodyatt at the time said, "no really knows what a quantitative trading platform is, but everyone wants exchange connectivity." So that became the new business model, and still is today for Object Trading, although it does have a full simulator which SR Labs and Market Prizm don't for example. So once again the market wasn't interested.

These days I look to hard-core matching engines like First Derivatives Delta Algo-Lab or the OneTick streaming thingy, even SMARTS from Nasdaq. All the products have floundered as "Trading Algo Development Environments" and are now being promoted as "Risk Management Platforms" with various canned scripts for risk monitoring... because everyone needs risk monitoring.

Now I have a new prop platform. Sure, it could be commercialized, but is there a market for it? No. It's designed to run fully automated institutional portfolios and institutions still have Investment Committees, Portfolio Managers, Analysts and Traders who then use algos... They just don't get the idea of automating the ENTIRE process... and what else they could do if the did.

That aside, I still like to prototype in commercial packages and I'm talking with NinjaTrader at the moment as their matching engine in their new Beta v8 is still no real improvement over the flawed v7. This is why I suggested in another thread to interface Python to NT8 with ZeroMQ, but I only had one taker on my offer of free source for such.

@James. Sorry for the long way around, but what I'm trying to say is that I agree with you on the importance of the issue, but the truth is most people don't seem to care.

I don't use MT4, but as Mark inferred it appears designed for the brokers to game you. MultiCharts along with TradeStation, WealthLab and NeoTicker all seem pretty solid, but I'd still check some trades manually against the data. NinjaTrader is seriously flawed even with basic order types and you need to be very careful to check its behaviour, but it does have some benefits in live trading over MC and TS and is easier to use than NeoTicker.

Thank you all for the response and advice. I will keep these 'features' in mind.

@Philip and group. My personal opinion is almost all retail strategy developers are not aware of these issues, the importance of them

or what/how to over come them. I have not used any other platform than TS for about 12+ years. I am painfully aware of the first thoughts

of professionals when information is presented in open forums produced with the TS platform.

Basically, I am on a fact finding mission to find out if my “check list is up to date”. If you think I am a system scammer, then have some fun at my

expense, debunk me. Others, if you know of something that I may not know in defeating the exploits of the new OOEL based TS platform

(9.1 thru 9.5), I would very grateful for the information.

thanks in advance,

James.. Check.. TS.. and MC same.. easy language.. / same signal.. you will not have same resaults.. On smaller time frame.... MC I am using for the last 4 year's.. is full of problems.. I solve them.. using my own aplication for for entry /exit /risk management.. same with TS and IB..

James - I certainly did not mean to imply anything towards you, quite the contrary I have real compassion for all my fellow traders. I think what Phillip and I are drawing from is many years of trying to help people and it falls on deaf ears. Aurel has also stated that he has done comparison testing and found differences that are disturbing.

We all have to do our own due diligence, but when you get a group of experienced traders who all say it's a duck then it's a duck. In spite of that truth many deer hunters will swear it's a deer and not a duck because they are deer hunting and that is what they want to see deer not ducks!

TS can not even get the same results between Radar Screen and the same systems when applied to a chart, try that sometime. But that doesn't make it a unusable platform, we just know the limitations of the platform and make sure these exploits are not utilized. There are some amazing exploits in TS.

Aurel, thank you your comments. I have talked with MC some and they have said that the moving my strategy into MC would require some

special dll’s to be coded and they are not will to do this. A lot of the issues that I am already aware of go much deeper than entry/exit of the signals.

I have done extensive testing on a lot(but not all) of the TS reserve words and have isolated some that do not function the exact same way as before

the 9.1 thru 9.5 versions was released.

many thanks

James again I am not berating on you at all - If I was working on a model with TS, I first write it to a low time, low range, low tick bar well you get the idea. It's a balance if you have trades that are 15 minutes in duration say use 1 minute bars or if trades are 1 hour in duration go to a 5 minute chart. if your trades are 1 minute in duration use 25 tick bars, like that. You want to make sure you actually see the price action moving thru the trade.

Do not use trailing stops, limit orders, floor ceiling blah blah, use on buy sell on close or open of next bar same for exits. Build that and you have something for real, cause you can always do that well or better on order entries. Check your entries and exits were you filled on the high or low of the bar, if so do you think you would have gotten the high and low of that bar - no.

These are just a small fraction of guidelines there is so much a book needs to be written.

James - I just thought about something as I was reading about the TS user functions and Dll's etc. The big guys have all been there, it's like MatLab we have all done that and what best advice I can give you in the world. Get a NOTEPAD and build your system on paper, then move it to spreadsheet as a bench mark. Now work to achieve that on a platform, if you can't do it on a spreadsheet and get it to work, probably it's not going to work.

Myself - I went the opposite route, I took a Cray and let it build systems and then I spent years trying to understand what it did. It was a totally opposite approach and costly time consuming but I learned things that no human would ever think of with their brain. If you want to go that route be ready for a voyage of about ten years dead time spending about 1mm a year and in the end thinking why didn't I just get a NOTEPAD and a pencil and go to work.

Day trading is a craft. There are a 1000 ways to skin a cat, but a good start is forming a bias and trading probabilities. One of my favorite day trade probabilities is the overnight stat. In the /NQ there's a 92% probability of taking out overnight high/low during the cash session. Bias, stat, price level, order flow trigger.

Taking entry is like judging the trajectory of a fly ball (in baseball or cricket) and is refined by seeing order flow for the product you trade over the years.

The movie Pi is a must see for all you system traders.

James, I'm not using TS anymore as I prefer to keep my IP on my machine and don't just trade the U.S...

Following on from what Mark was saying, here are the most important parts of the MC wiki if you ask me.

http://www.multicharts.com/trading-software/index.php/How_Scripts_Work#Order_Execution_Priority

http://www.multicharts.com/trading-software/index.php/Intra-bar_Price_Movement_Assumptions

http://www.multicharts.com/trading-software/index.php/Bar_Magnifier

Like I said, TS and MC are pretty solid, but you need to understand the limitations.

The first trade would have you exit the Long trade at 95 as the Open below the midpoint indicates the Low occurred before the High.

Second trade, a short trade triggered at 90 as a single move to the low is assumed.

Third trade, the Stop Loss on the Short trade is triggered at 105 as the assumption is a consistent rally from the low to the high.

Fourth and final trade, a new Long Entry is taken at 110.

Please ignore "Let’s say the Market has a wide range bar with Close above the Open." ... Wrong assumption. Fingers went off without without cognitive function engaged.

Statistic science

Side bar:

Strategy Concept was put on paper in the late 1990’s. Stacks of journals.

Signal generator extensively tested with:

1. second bars (limited to 6 months history)

1. tick bars(limited to 6 months history)

1. minute ( minimum of 500,000 bars in test )

etc..

Portfolio Execution of signals always in 1 minute charts for uniform test comparison.

[IntrabarOrderGeneration = FALSE] always in code

VAR:VOL(0);

VOL = VOLUME; always in code to stop “ Intra-bar Order Generation Optimization with Look inside bar testing”

just in case of forgetting to uncheck....

If BarStatus(1) = 2 THEN BEGIN (bar is closed check)

Buy TRDSIZE CONTRACTS NEXT BAR at MARKET;

please keep the ideas coming

above should have read ten second bars and ten tick bars ??????

http://screencast.com/t/hXzO6Q5E

James - I suggest NEVER use Volume in a trading model as most platforms have yet more bugs in reporting such. I hope your strategy does not center around volume.

If volume worked properly, was dependable and timely, I would the very first to build models using it. Here is where you need a interface from TT or CQG to get proper volume.

I gave away a free multi server for TS2k that would interface multiple data streams into that platform. I also had a link from TS2k to MatLab, oh gets depressing thinking about resources wasted on such projects.

So if you are using volume get it from another source because TS has some real big bugs with volume. As I recall the real time reporting does not match the historical reporting and they summarize the volume numbers as well. They also do this to some degree with tick bars.

Mark, Volume is not used in any algo and I total agree it never should be. It is only put into the code as one of the preventive measures

to stop the "false enhancement of performance reports built into the platform.

James - interesting image of your system, looks like it's long only, intraday only, orb reversion to mean when gap down, breakout buy on gap up, something happening with stops at 2:30 which would suggest some type of time based exit if a new break out high is not achieved in which case the time stop is sidelined, with a mix of profit target and one percentage based optomizable parameter. nice.

Mark, no profit targets or stops used, only incoming signal. Trades both long and short. The F1

percentage is not what you think it is.

Looks nice James. Looks like the signal around 2:30 is momentum based. If its negative then bail. It's time I put some more time into "time of day" issues again.

It was so much simpler in the 80's and 90's when there was only one session (two in Asia as we have a lunch break) and that's where the volume definitely was!

These days you can completely change the look of a day/session bar depending on where you choose to have the market "Open" or "Close." This can play havoc with pattern based algos... and so the journey continues.

Mark and Philip, thank you for your kind comments. My original mission is still the same, to discover additional ideas or methods to stress

test the duck, as well as find any more incompatibility issues between the duck and the new TS platform.

@ James TS -MC ..are the same as all other platforms. . with good and bad .. personal I dont like to lose money because of one bug or error ..but sure are many trader witch can accept bugs and errors .. and losing money .. i stop accepting when I decide (before 4 Y) to develop my own risk / trade management system ...Philip and Mark ..spend more time . . then usual traders ..trying to understand ..why developers /company's ..dont do their job ..

The art of discipline

.. i like this "The art of discipline".. :)

Day trading is an art which we will have to craft with a scientific approach to acheive a magical output 😀

Manual Day Trading = The Scientific Discipline of Trade Execution Automated Day Trading = Fracking Genius

@William, I like that one!

Best :) .. Manual Day Trading = The Scientific Discipline of Trade Execution Automated Day Trading = Fracking Genius.. beautiful

I Love Day Trading!!

Charter Group has a new website and branding - www.chartergroupadmin.com for more details

William .. today Denis show me his way of day trading .. more simple I never made . and your 'definitions' is perfect :)

I am actually trying to design a system witch can /has to be the most simple ever made ..