robertm

VSA, like most other indicators, is more useful as a marketing gimmick by those who can't turn a $ with their own account to rope in unsuspecting wanna be traders than anything else. But nice try...

I think every trader has this thought at some stage Tim, I know I put the same question to my mentor many years back (he's a former broker, stupid idea he said), but finding a less stressful source of income that lets you be free to trade when you want, and how you want, while paying the bills so your account can compound without you drawing it down, is the best solution. Just stick at it, gaining experience in different market conditions is one of the keys, and only time can open that door for you.

There's actually some really good information in this thread so it's worth bringing it back up the list. Other than dropping volume (if you choose, I know people who use it with FX still) there is no need for any indicator or changes to how you detect the price action you are wanting to trade (unless you are trying to trade a lot of gaps which you won't find). You will need additional filters around such things as time of day to trade, and an appreciation of the fact FX markets range rather than head for the moon, declare bankruptcy or get bought out in a takeover, all of which can effect your money management strategy.

S/R is visible on any timeframe (and most chart types) you consider has the liquidty and brokerage to allow you to trade it. The lower the timeframe the more noise you have to deal with though. If your candles consist of 3-5 tics (H-L), you probably aren't getting a good view of what's going on. Just remember to point yourself in the right direction.

They don't need to be "the same". We work with OHCL references, not trying to match up identical instances in time to the tick. Saying that back-testing is pointless implies that with no prior experience any fool could sit in front of a screen and just pump out the cash. Experience is back-testing in slow motion.

Do you know if your source did a similarly detailed study of when those volatility sweet spots are most prevalent Tams?

Study when others make money using a similar strategy (ie, trend, scalp, whatever) to yourself on a similar timeframe that you wish to trade. You can find this by sitting in on rooms, or viewing results that are published, or looking at when you get the most spam in your inbox as all the scam systems/robots usually make money at the same time (hint hint). By study I mean over several years and market conditions, not a few hours of research.

Agreed. Bots can be approached successfully from two angles. 1. You are working some sort of HF magic they'll probably restrict soon anyway. 2. Having fully grounded your system you are adept enough to translate that perfectly into a rigid rule set to mimic when you trade, what you trade, and how your trade, and everything that you "see" when you look at the market. Thus simple systems are actually easier to build than complex ones in a way. You are just replicating that part of your system that gives you your edge. Unlike much of the marketing hype making money "24hrs a day" in a market probably isn't the best strategy even if you never needed to sleep, eat, or have a life.

Hi guys, So I've coded up my expert in Amibroker with far less pain than anticipated (although I haven't started pushing the hard points on the to do list yet). My next step involves how to handle the alert outputs. I'm running around 20 symbols on various timeframes on RT feed. My early steps involved simply plotting the expert on the chart and tying this to an early warning alert, and a close of bar alert (output goes to alerts list, and I get a nice visual reference for early stages of building and checking the system as well as handy things like changing the chart background when a trade is about to confirm). With around 20 symbols on a single 5m timeframe Amibrokers resource meter is screaming 200% of capacity already, but time wise still processing everything fine, and CPU/Memory load is nothing. Do I need to worry about this? Opening a new Amibroker window for each timeframe seems one solution as the load is window specific as they the load drops as timeframes increase. Even so, loading up 100 or so charts obviously isn't the optimal solution given I just need alerts. Part of doing this is to cut back the number of screens and minimize the information I need to deal with. Thus an alert list telling me what chart to look at is somewhat ideal (ultimate goal is reduce it to smartphone capability to receive/act on alerts, but that's for later). So I'm thinking the following solutions : 1. I run it as a scan set to repeat say every 1 minute as this would be required to set off my alert which is timed for 60 seconds before bar close. Output results to a watchlist but this means keeping a close eye on it, or can I just tell the scan to output alerts as well? 2. Somehow reference the multiple timeframes from the AFL code within a single symbol chart (possibly using TimeFrameCompress function and giving it a hard list of what times to scan what timeframes). This would allow me to keep the graphical display alerts on the chart (color change) and address the trades from the alert list or text output in the chart header telling me what timeframe is confirming. It will substantially increase the AFL code load though. 3. I go to another platform better suited to this. Is anybody doing this successfully on other platforms? 4. Turn it over to something like Strategy Runner and automate it (a probable future side project exploration but not what I want at this point). Suggestions and advice appreciated :-)

I noticed IB's HK site now has a big warning that opening an account with the HK office means falling under US Laws. I asked a rep about it, he didn't even reply, they must be losing large numbers offshore now. There are plenty of Non-US based brokers in Europe you could hunt around though. My rep at a US futures broker I sometime use was trying to get me to swing my FX trading over to them, I bought up they US based broker question, they said they could just swing it via another country instead. Thankfully I'm not a US resident though, the tax laws are out of control. On one hand they do things like this, while freely allowing manipulation of the tax system at other levels (fancy making your board meeting a 3 day boat cruise? or tax deductible dinners and opera tickets? all easy done within the US tax system). The issue with spot FX is it can be used to funnel funds from any point in the globe to any other point and avoid/reduce tax and detection in the process. Thus this could be a result of anti terror laws or some such thing and not just tax related. The land of the free!

I'm interested to know of peoples experience using the IB data feed (primarily for FX) with Amibroker specifically, but also how people have found this in general (SR, NT, whatever). Also is anybody using the execution plugin for IB that is available for Amibroker and how have you found it? Thanks again peeps.

Thanks Kiwi. Yes I've heard a lot of good things about Sierra, downloaded it a few times, but never seem to get around to playing with it. Broker wise I have a few, I don't tie myself to a specific broker, most of the best ones will offer the same tools to get the job done. I noticed Amibroker has a feed plugin for IB data from TWS, and has a plugin for executing with IB from the platform now. Having just explored the latest version I LOVE the Interval linking feature, I've wanted this on previous charting packages like eSignal for years now. I also like Ami's ability to switch layers on/of to keep charts free of clutter. Some things are still a bit clunky like not being able to "edge out" charts by getting rid of the title bar in the chart frame (which eSig does for me). Topping it all off I know several die hard Amibroker users with extensive coding/testing experience so I'm leaning that direction for my explorations at this point before going down an all new path like Metatrader. I'm pulling some bits of code together now and will be back to bug some coders when I hit the roadblocks.

It does, and I love it for end of day work, but I don't like it for multi chart handling. I'm still on 4.8 though so I should check out 5.5 featuring "automatic Walk-Forward Testing, Multi-monitor floating charts, symbol and interval linking", according to their pitch.

Thanks BF, waited to see if any others chipped in before replying. I think I'll have a look at EL via MT for starters due to the popularity & therefore ease finding info/assistance. Is there any difference if I start on MT5 given it's the newer? (ie, will all the public domain info still be applicable or were there major changes between the two code wise?). I realize it's not offered by as many brokers for the moment though, or just use MT4 then migrate later? After proving it it's useful I'll consider any changes in my eSig preference for main charting package.

FX can be traded as futures or spot. Also I know people using NT via IB and GFT offer SR solutions, so they can be used if that is what the solution calls for.

Hi Guys, After years of working out an FX system that suits how I want to trade I'm looking for a platform that can help identify the trades and hoping for some (experienced) suggestions. I am NOT looking to build a robot. What I want to do is simply place an arrow on the confirmation candle and have an audible sound play to alert me to a trade opportunity, on various time frames and pairs. Code wise it'll be pretty simple. X period close above MA with macro MA filter on another time frame, maybe a few indicator filters as well. I had a quick look around on google but mostly found advanced EA robot builders. I use E-signal for data but haven't delved into it's other features (and it's a bit of a dog to program I heard?). I use Dealbook to execute but haven't looked at coding it directly (anybody tried?) and as I'm not planning on trading the signal generation directly I'm happy to look anywhere. Suggestions? MT? NT? SR?? I have enough understanding to manipulate code, or my other half is a programmer and can talk geek if required :-D Thanks in advance.

If you could magically reproduce the exact trend and volatility of the period in which they were backtested they probably work fine. The unfortunately reality is these are ever changing, but a good sales pitch tends to gloss over this point.....

Not necessarily. Having more periods to calculate the average from will smooth out spikes in volatility better. Take FX as an example, news spikes will skew figures when viewed over a small number of periods, and as much of the activity happens during a few hours even though it is a 24hr market this will mess with some of your intraday views on volatility and averages. At the end of day the most important job of the MA is telling us which direction the trend is going allowing you to code around this basis, so being accurate is secondary to simply achieving something that gives you the results you are after.

Because there's money to be made in regulation, which is why an unregulated CFD market operating in the US just isn't allowed Obama is a nice guy but he's looking pretty shaky on the second term if a half decent candidate steps up.

:rofl: :rofl: :rofl: :rofl:

I'll second that :-) There is also another "Ribbon" code to give you a simple on/off that runs along the bottom of Ami or Metastock to tell you if price is above/below your trend setup condition (MA). Gives you a very clean chart to work with and lets you focus on the s/r levels without any MA's getting in the way. I have to say I found visual indicators very helpful with those light bulb moments in the early days of system exploration though :-)

Update, whooooohoooooooooo. Hope this opens the door for CFD providers as they can now hedge their positions. China?s ?Big Step? on Futures May Boost Investments (Update2) - Bloomberg.com

Should be interesting to watch, might wait til the mini becomes available though :-P http://www.bloomberg.com/apps/news?pid=20601087&sid=aYiuYawgXe48&pos=5

I'm going basically with what Kiwi & DD said. Longer term systems on say weekly charts you can be less accurate on the entry as it's a bit like turning around a loaded oil tanker if you are going with the trend - it'll happen eventually but it'll take time. Shorter term getting entries right can be more important if your risk:rewards are lower, and also dependent on how you measure your stop from entry (for which ATR is an excellent measure even if you use other methods). Entries can be exactly the same as exits (ie, x period high/low above/below MA) if you choose, it just depends what works for you Having a clearly defined entry will keep you working within your framework of trading rules more than anything else and stop you jumping into trades just because you have some free heat/risk capital or haven't had a trade signal for a while.

A lot of it will depend on your market/time frame. I know a lot of "FX Hero's" who became full time FX traders late last year. Most of them are back in jobs & haven't made money for 6 months +, markets change, sure you can adapt but some conditions are better than others for pulling in the $$'s. If you rely on big longer term wins on equities then mid 2007-mid 2009 you probably didn't make much. It'll take years for those trades to gather momentum as well. Think of it like being a commission only salesperson working the big end of town. It may take years to close a deal, but then you get a big thumping commission check all in one hit. Markets move in cycles, ensure you understand the cycles of the markets you want to trade (trend, congestion, range, volatility, duration), and what time frame to trade them on. Becoming a scalper isn't necessarily the path to full time trading, and certainly isn't everybody's cup of tea. My own preference is to leave all my money in my trading float unless I need to use it (which doesn't mean it's sitting at a brokers). Risk management takes care of the rest, and living within my means (or as a mentor once put it, sometimes you dine on noodles, other times you dine out, the market will tell you which and when). The more you have to lose, the more conservative you should be.