4EverMaAT

the OP would have to advise on what instruments he is trading. But metatrader 4 is common among forex brokers. MBTrading supports metatrader 4.

I know how you feel. I'm not a programmer myself. vWorker.com is your best bet. I know a few people and can provide contacts if you'd like.

What about using single stock futures?

what portion of the backtest data is walk-forward tested? I suppose that would be difficult without tick data, but tradestation does include tick data, correct? I think between 6 months and 2 years? And minute data goes back even further.

One time a few pals of mine pooled together and purchased a robot (I think Gomega) which claimed that if the robot could not trade at least 5% monthly profit for 3 months with default settings, we could get full money back. We followed their instructions 100%, even using a VPS which wasn't a requirement, and provided their requested proof. They never argued with the proof, but instead said "we'll pay you back when we can". Obviously we never got our money back, but I did learn quite a bit from a different EA that was purchased. Lessons learned. Let's take a closer look at this: If you are going to demand that the automated trading system be left completely unattended during the contest period, then likewise, each method and exit must be documented in such a way that an onlooker could see what your rules are, and that the rules are being applied consistently each time. Just like the robot can adjust itself, you would be allowed to adjust your strategy, but it must be a logical adjustment that was already laid out in your original trading strategy prior to contest begin. You couldn't come up with a new rule after the fact. Pretty much, like the original turtle trading system, it would have to be mechanical, transparent, and clearly defined ahead of time. Not lose more than 2% of trading account balance on any losing day. Or not more than 5 losing days per year. I don't know if anyone could guarantee that, meaning that each person's strategy may not work well under that condition. Wouldn't a more effective challenge would be to see how much the person could make by a certain date? Now if you insist that you want to keep your system private, which is perfectly fine, then you must allow the other contestants (robots) to trade completely their own way also. edit: added link for the original turtle trading system rules pdf

or babypips. I did get a chance to follow some of the price action threads and it seems like when I ask some basic questions about entry/exit, etc, the rules always seem to be very subjective "depending on how price moves around the next bar or the bar after, etc..." no specific if-then rules. The irony is that a complete system: - imitates the trading decisions of the human, but allows instructions to be followed unattended, and can be scaled to handle faster market conditions in a more consistent manner than human mouse click. - removes second-guessing. - encourages thought and even improvement, see below. - forces you to be more accountable. Because it would have required a logical set of instructions to begin with. Plus you have the added benefit of observing how your system works outside of yourself. Like sketching a concept on a whiteboard or mindmap. Obviously with a blackbox system you cannot know this. Build your own OR buy a system that is easy to understand and you understand how to control the different "levers".

3 workarounds to doing it: 1) (less precise): use futures contract equivilant fx contracts and calculate proportional volume into its spot FX equivilant. Also CME and other institutions may actually be starting to do some OTC transactions on an institutional level. 2) (broker-specific): measure the price difference per incoming tick. If they offer some sort of level II volume statistics, use it. I think FXOpen ECN, InstaForex (surprisingly), maybe MBTrading, Interactive Brokers, and maybe the newer tradestation feed does offer some sort of level II insight. 3) (broker-specific): see if you can get a feed with as many liquidity providers/brokers as possible / api access and get the actual level 2 volume as they keep track of it. Then you would compare the net price movement and net volume for each tick. Obviously this is the most accurate of the 3 and the most expensive/difficult. The inherent problem with forex is that each individual broker "makes" the market available to its clients. much of the retail volume is aggregated and re-aggregated, plus there is a lot more that will never be shown in the higher-tier interbank markets, unless some of those providers start revealing real-time volume information. Exchange traded products every tick is accounted for; so no problems getting accurate tick data. Forex brokers tend too "filter/smooth" their datafeeds. You'd be surprised how many actual ticks would come through if they didn't smooth the feed. But then there may be additional platform stability issues, etc. And for historical analysis, tick data is lacking from many brokers; forextester is one of few companies offering tick data collection. If metatrader 5 had been what metatrader 4 was now, this data might have been more accessible already with more traders. My APAMI indicator, which measures net distance between the current price and a previous price point, (qualified by retracements) did reveal some interesting aspects about how price really works. More on that hopefully by the end of this week :haha:

You should more clearly define "system". For me, a trading system (or any system) must be completely mechanical; all decisions are accounted for ahead of time. You have your primary objectives, plus the appropriate "if-then" statements for any detours. Here's the trick to using computer automation, or any type of automation/system/etc: the computer can only do what it is told to do. The computer is not a magician, however it can usually compute repetitive tasks much cheaper, quicker, faster, etc and with greater precision, flexibility, etc than with the human labor attempting the same tasks.:missy:

This thread was one of the best that I've read, and bluehorse's skit on fish.....classic:rofl: The central problem is actually the mis-use of the word in general, where the author does not define the terms that they are using in addressing their audience. The audience has to then assume that their own definition of price action is the same definition that the author is referring to. In the ambiguity, some members of the audience will say "oh, that's what I was looking for". Of course, when you put the actual method or strategy of replicating the price action in practice, you may find out just how deadly ASSumptions can be. This can be solved by the author simply defining the terms they will be using. BTW, "risk" is another one of those general terms that are tossed about; very few people actually define what risk means to them and how they calculate it.

In the edit profile section: non-standard linkedin profiles need ability to be added. my linkedin link does not look like the example you give. http://www.linkedin.com/in/traderslaboratory vs http://th.linkedin.com/pub/jon-grah/51/14/895 perhaps the user should be able to add their entire link, and you can just check to see if the linkedin word is contained within the first part of the url?

what market are you trading? I take it futures? Some brokers have their own free charting platform. Use it. If they give you free API access, then you may be able to use NT. forex usually has metatrader 4 with most of the retail brokers, so free demo accounts. wrbtrader has a point, in that using the broker you plan to actually go live with is the best option, if available. If not, then pick 2nd choice, 3rd choice, etc from a list of brokers you plan on going with.

The volume and the agreed price on each volume unit would be what you consider to be the "completed transactions"? Easy to do with exchange-traded products. OTC like forex is a little more tricky, but can be done.

1) Why you smoothed the price was not very clear. 2) I'm still lost on exactly what is "Hysteresis". 3) If the goal was to improve profitability, then the statistics currently available dont seem to prove this. Saying that your trades were more profitable simply because of less trades and an increase of profit per trade doesn't take into account: - the fact that the drawdown did not decrease by the same proportion as the trades became more profitable. - the sharpe ratios and annual rate of return remained the same for SMA cross, Band Cross, and Price Proxy Band Cross - perhaps a more clear definition of what a "winning" and "losing" trade was (where were the TP and SL, what where the based on, etc). From my point of view, the most profitability would be achieved by optimizing the management of the positions while in the trade (position sizing). But that doesn't seem to be explored in this scenario. The OP focus was on improving perceived entries and that by having "better" (more accurate forecasting) entries, the overall profitability increases. If anything, the OP data gathered suggests that entry accuracy alone cannot be the cause of greater profitability.

Koyasan, this is a good question. i tried to address it in the previous post, but I think the OP should elaborate on his logics (sp?) to get a better understanding on how to "Play the odds"

Just to elaborate, a logical method must include precise rules for entry detection, entry, and exit strategies for any condition that the market throws at you. As a former turtle trader eluded to, any decision that is left up to the trader after the fact will cause indecisiveness, and eventual losses. The entire process should be mechanical, so that there is a precise course of action during each aspect of the trade. coincidentally, a fully mechanical system would be 100%, not necessarily 100% winning trades, but 100% consistent in doing whatever it is designed to do, which is ideally to profit

It's the interpretation of the TA or FA that gets traders in trouble. An indicator is just that: indicates or detects a certain market condition.

losses would also be large if the market moves against you. Are you referring to futures or forex. Typically in forex, the major brokers you will see about $3-5 per round turn trade for a standard lot (100,000 of base currency) which every pip move is approximately +/- $ € £ 10.00 . Oh yeah, there is a spread also between 0.8 and 3.5 pips for the majors and cross pairs. Futures the spread is usually 0-1 tick + commission which for a discount broker is $4-5 round turn with all the fees. edit: Jack beat me to it, lol. But looks like I was right on the spot with my answer, and it's been years since I traded futures.

followup: the article point number 2 (Reason 2: Empirical evidence for TA is negligible) confirms what former Turtle Trader Curtis Faith emphasized that, besides ensuring the entire process of trading was mechanical (and therefore objective, testable, measurable, etc), position sizing is the most important aspect that the user has under their control. Not spending countless hours on predicting the best point of entry. I eventually got the hang of this, and developed a spreadsheet to assist in formulating a proper money management strategy; one that can handle all market conditions according to the size of my trading account.

Just different sides of a coin really. Focusing and choosing to prefer once side over the other helps to get at the very root of what deludes most traders. Isn't the ultimate goal to figure out the core (fundamental) aspects of what drives the market, then see what entries and exits (technical) can allow us to best extract profits? We assert preferences to the market, and attach ourselves very heavily to these preferences. As long as you get to the end goal, and can duplicate the results, does it matter what adjective we use to describe the process. edit: see new post

We all would love to keep our trading machine ideas completely in-house, but the reality is that for most people it does take a steep learning curve to incorporate all of the "if-then" aspects of a strategy into a programming language. It's not just machine translation, but requires interpreting the essence of what the system is trying to accomplish, and using the correct functions so that [your ideal charting package here] can display and autotrade your strategy successfully. Even if I can explain the problem accurately (logically), do I have the technical expertise and time to physically code bug-free? If it is a hobby, and I have loads of free time on my hands, fair enough. But as a business? Outsourcing can be wonderful, but it is NOT drag/drop operation. It takes the patience of Job to ensure that what you say - what the programmer understands = flawless execution. Initially, it can take many long days of back and forth deliberation, proof of concept, etc to get on the same page. What if the person bails? What if they run off with your project, etc. (think Facebook). This is not unusual when outsourcing to emerging economies, or your own backyard as was the case with Facebook. That's the risk that you take to save a few dollars. vWorker acknowledged this problem (entrepreneurs not having technical expertise to properly supervise project) and came up with Tech Sharpra. As someone who has gone through the process with MetaTrader 4/5, I can tell you, then end was rewarding....but it was quite a ride. You have a greater appreciation for finished products in every industry; most people never see the back-of-the-house stuff. edit: sierra chart has its own programming language? Wow, things have changed. They used to insist on using Excel to do the heavy lifting. If they still offer free historical tick data (or 1 min data), with some of their packages.....its hard to beat for a new trader. But i do agree with another poster that it may be worth it to invest in a more comprehensive charting package (maybe multicharts?) that has the proper backtesting/debugging modules, plus the charts should do what you want them to do. Hard to say which package is "best"; you just have to trial them all and experience which one(s) fit well.

Sometimes you have to lose it all before you can fully grasp with experience how the market works. I'm not saying it must be this way for everyone, but Scott Shuburt asked a question in one of his videos: "How many times are you prepared to lose your account equity as tuition of grasping the fundamentals of how the market works?"

The systems that we create are often to maintain a sense of predictability in life. The fear comes in the curve fitting we often apply to nature. We get used to our predictability and begin to wrap assumptions around how our lives will play out. Many of these assumptions take place (or are maintained) unconsciously; on-the-fly stereotypes as the mind is conditioned to justify everything it perceives. Two minutes of honest scrutinty and it is a big shock when you peel back the stereotypes and you see yourself and everyone around you lying all the time. It is true that you can only control what is in your power to do so, mainly yourself; your point of view. But the mind usually wants to take responsibility for everything, as the fear of no control (no self) is at the very core.It's realizing that as much control as you like to think you have is real, *poof* it can all be over in a flash. Did any of it really matter? It does while you are alive, right? So risk shouldn't be something that is run away from, but it is acknowledged fully as inseparable part of trading (and life). Most trading plans (particularly of the directional predictive type) don't account for the uncertainty. Traders want their trades to go the way they want them to go. I suppose that only Rande can answer this, but I would suggest practicing "The Four Agreements". That will definetly remove many of the self-limiting agreements that cause a trader to suffer. Until you challenge, or at least acknoledge your core fears, you'll never have the courage to let go fully. Rande could help by providing support pointing out the fears and laying out some template in which can assist the client in removing fear or no longer giving them attention they don't deserve.

The scarcity model of economics has gripped many people, including myself sometimes. But all anyone has to offer you is their story. The truth that most traders are looking for is usually right in front of them. We are so trained to seek recognition and approval from peers/others, it's amazing. The irony of the whole thing is that you must test for yourself anyway any method or strategy to see if it will work for you. There is a shortcut to this testing that I will summarize: figure out what's true about the market first....then you can do what you want.

If account A has $10,000 and you have 1 lot of EUR/USD long at price x, and account B has $10,000 and you have 4 lots of EURUSD long at the same price x, the risk of ruin is greater in account B since it would take a shorter pip distance from price x to wipe out account B than account A (exactly 4 times less distance). It also takes less distance to reach profit objective in acct B than account A This isn't my point of view, its arithmetic. Letting the "chart tell you where the risk is" is excellent. And if the charts tell you where the all the risk is, the lot/position size would be irrelevant. But we know that's not true with the above example. Perhaps what you meant is that you let the charts determine potential entry/exit points. [edit] In which case you make a choice as to what size to enter and how much to scale up or down. What often happens is we have preferences. And this is fine. I'm not blaming anyone for their preferences, but the markets are completely objective. Every lot or position has a price. Gambling has more of a "negative" or "Las Vegas" loose/wild connotation whereas business is usually associated with more careful planning and pre-meditated calculations. Dont get caught up on semantics. Regardless of what adjective you use (gambling, chance, risk, business, etc) Bluehorse just about sums it up the reality of risk. Nothing will ever manifest without taking some action. The idea of "risk free" rewards may make for a good marketing campaign, or a demo account, but in real trading, its about determining the best choices to make and then actually making them.

Definitely! I made a lot of backtest experimenting with how much I could risk and get away with it, 4% was the highest I could go, anything higher would blow the acount, it's hard to say how much you can risk for any given method, I believe it is a very individual number and how good your system is and how good you are as a trader, with some trading methods you could possible risk more or less depending on profitability, win ratio, risk to reward etc, there is so many factors playing, The point was that if your strategy revolves around predicting high probability direction movement, you should be betting considerably more than lower probability signals. This is consistent with the oft repeated (but never explained) "Let your winners run and cut your losses short" In order to let winners run, you either need to put it all in on initial entry, or space it out (pillar, pyramid) as the direction is favorable. This is where position sizing is critical, because it spreads out your risk and you can adjust profit targets to bank larger profits. I invite people to use the ProfitKeeper equity tool to manage their exits. It forces you to be profitable if you use position sizing that matches your strategy. Then you can lock in profits depending on how much you risk. How much one chooses to risk is up to them. But yes, the more positions you put on for the same account balance, the greater the risk of ruin. The good news is that you can apply betting strategies to minimize the risk. One example, if you were betting in 20% per opportunity to trade (either a single trade or series of trades), you can have a profit target of 60%. If you were betting in only 5%, the profit reward might only be 10%. As demonstrated in profitkeeper, you can auto-adjust the amount of profit you want to take depending on your strategy. And backtesting, beyond testing the fundamental aspects of the system, is worthless IMO. Walk-forward portions, particularly forward testing is useful so that you can visualize if the core aspect has merit. A side benefit (with historical data) is that you can step test (optimize) particular parameters to see what would have worked better in certain market conditions. Without tick data, this would be difficult to accomplish accurately.