Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
STF discretionary spot Forex system development journal
Don't hesitate to make your presence felt--I may have taken this stuff seriously in my misspent youth but am way older than that now
I think backtesting & optimization is necessary but don't put too much faith in it--more or less an exercise in curve fitting--and don't do blind runs (anymore). I write a lot of bots to test simple ideas. To save time, if a bot doesn't optimize the first time around, or performance fails to improve as features are added, it's discarded--scratch that idea. IMO the proof of the pudding is in forward testing. Frankly I'm still trying to grasp what makes a bot robust (i.e., makes it immune to choice of time frame, instrument, session, etc.) and if not robust, what characteristic(s) of the data is grinding its gears, not so much interested in the behaviour of a single bot.
In the beginning I tried to trade the majors and a lot of the minors, including USD/YEN, but eventually decided to focus on a single currency in the hope of mastering its idiosyncrasies before moving to the next. That was a couple of years ago :-/
I don't give it much thought, actually. I use IB (filtered data) since a more "accurate" feed is reduced to what the broker's order fulfillment is using, and IB's Forex universe is only one in a very large number of currency broker universes including the ones that massage their spreads, so it amazes me we speak of the present value of a tick at all.
Hope that helps !
Can you help answer these questions from other members on NexusFi?
This morning looking at discrepancies between NT's accounting of bot behaviour and IB's record of it, as well as weirdities in bot order placement (i.e., exiting a position but not entering a new one when "reverse position" signal occurs), trying to figure out what is going on. So far enough of a distraction to avoid manual trading.
For the record, first item of confusion is that the parameterization of anaSuperTrendU11 shown on the chart with the active strategy is the default, not the parameterization of anaSuperTrendU11 actually being used by the bot--no surprise if initialization events occur before startup events and in the category of a program bug--mea culpa.
However, while this may cause some head scratching it is not behind the serious performance/accounting issues (the main reason I hesitate to post NT performance summaries--I won't trust them until I find out e.g. what is randomly corrupting NT's database). Right now for example NT Strategies tab indicates the bot is short 1 contract on IB's paper account, whereas IB reports it is flat and NT Positions shows no order active. Really, really do not want to get into order management in this bot at the level NT should be capable of doing it--must be something I'm missing.
ETA: Somewhat later, I've quit searching through code, logs and trace records to find some clue about what's up with NT after my eyes glazed over and blood pressure started to rise, now placing manual orders in an effort to calm down (so far mostly stopped out for the usual couple of pips profit, which tends to drive me nuts), confronted now and then by IB's new, infuriating trick of popping up a window after a random number of orders to inquire do I want the present order for 1 contract processed as a Forex (IdealPro) or currency exchange (Ideal) transaction, the difference being a considerable penalty in spread through Ideal, and would I like IB to make the choice for me in the future (yeah, right). W...T....F???
I'm not particularly religious but just now notice somewhere in the back of my brain someone has turned on a radio that is blaring the Battle Hymn of the Republic...
Watching EUR/USD today for the record, probably won't trade, partly to observe behaviour in the London/NY overlap sans NY activity as a matter of curiosity.
Just over $80 net (before taxes), 9 trades of 1.2 contracts split into 2 targets for 89% profitable--one of those kind of days--trading price as it fluctuates between closely bunched activity areas, keeping an eye on momentum (or lack of it) this AM, each trade lasting no more than a few minutes due to erratic price action. Using an ATM strat tuned to the distance between activity modes (won't call them value areas), trying to get the breakeven stop right. No trade in particular worth mentioning. Migrating bots to another stripped down (app noise free) computer in an effort to placate NT, may be done by end of day if I can muster some enthusiasm (recovering from the distraction of end of season family get-togethers over the last couple of weeks)--they are offline at the moment.
ETA: to remind myself to be paying attention at 11 AM AST (10 AM EST) when US ISM numbers come out, as happens fairly often noticing out of the corner of my eye I may be missing the first significant move of the morning (hence some quick & easy profits) while typing in this journal
ETA: Tending to prefer short trades at the moment but momentum gusty while price caught between prior day low (1.25625) and prior day open (1.2574), and expecting volatility in general between 1.2550 - 1.2575, vicinity of 100 day MA. If support at magic number 1.2550 does not hold (and it can take hours for the market to make up it's mind, actual mood obscured by commercial traders' order books, time of day, proximity to end of session, etc), then will probably continue to prefer short trades. In the meantime however prepared for things to turn around as N American market open looms (another hour). About 1.5 hours until ISM numbers.
ETA: In terms of overall strategy since 9 AM AST (8AM EST) going with downtrend, which means going short for 10-15 pips but opening up initial stops (intended for relatively short lived but high momentum unidirectional bursts in activity) to avoid low momentum bouncing around in these narrow S/R channels in the meantime--not necessarily a wise choice since "low momentum bouncing around" can signal a turnaround if there are buyers lurking (perhaps likely) who are waiting for one reason or another to jump in here (1.2565-1.2575, or S1), but mainly (yet again) prevents me clicking buttons. Given lack luster N American futures at the moment and gold price (which I use rightly or wrongly as a short term filter for the dollar index, one-step removed) guess I have faith in the gravitational attraction of 1.2550.
What all of that amounts to:
ETA: by 10:09 AM AST 1 target hit + second stop hit at breakeven + a pip for a few pips more in total when price abruptly reversed to S1. No surprises there. Supposing (hoping? ) price will now meander upward perhaps as high as 1.2590 in the 1st 30 minutes after N Am. market opens if it gets that far before the ISM numbers come out, rather than rattle around down here, so will likely chance a long.
ETA: 10:37 AM AST, 7 minutes into the open.....oops. Why we should trade what our system is telling us rather than what we hope price will do Stopped out for a loss of over $100 a minute later. After that fiasco up $73 so far for the day, 16 trades so far, 81% profitable (i.e. a bunch of mostly small-profit trades and one doozy of a loss). Fighting the urge to punish the market for that indignity (i.e., fighting the urge to compound the error)...time for breakfast...particularly since price will more likely hover around S1 at least until ISM.
ETA: Break having lasted until 12:30 AST, end of session for me, looks like I didn't miss anything--packing it in for the day.
Curious to see if the ECB bond news leak will be enough to bias EUR/USD upward for the day on top of earlier (presumably short covering) gains. Needs to break 1.2600 for that to happen.
ETA: Suppose you can't win it if you're not in it. By 8:30 AST (7:30 EST) Spanish bond yields starting to drop, indicating someone is buying the rumour (so to speak) but remains to be seen what is left once the mills of the FX markets are finished grinding the news.
ETA: 10:13 AST price consolidating ahead of US open between 1.2596 and 1.2611. Wish it had showed a little more enthusiasm...may bail ahead of the open.
ETA: 8 minutes to open still consolidating at 1.2600, momentum going to zero Will often exit in these circumstances (technically ~ 50% probability) but not today since news seems to be taking a while to digest. If the stop is hit where it is (+8 pips before slippage) during the open it will mean waves (noise amplitude, indecision, game playing) bigger than I want to surf. Prefer buyer enthusiasm right off the bat.
ETA: Stopped out at 10:40 AM for +8 pips before commissions. Buncha wusses Coulda, shoulda, woulda, and might if it takes another run at it, otherwise will have to wait for a short setup.
ETA: 10:51 AM EST Long again at 1.25925, first target (3 pips) hit, stop at break even, remaining target at 1.26035 or +11 pips, but while sitting here twiddling my thumbs have time to think about my session choice. Higher level of activity not worth much if everyone is tugging in different directions.
ETA: 11:00 AM EST 2nd target hit for 11 pips before commission. New long limit at 1.26045 at the last activity peak in case price, hitting 1.2619 while making an initial break for it on the hour, behaves like kids leaving home--gung ho at first but then come the 2nd thoughts. Not going to chase it. If you love it, set it free. If it was meant to be it will come back....but hopefully (like kids) eventually it will leave for good
ETA: 11:15 AM EST. Ah Merkel...bull in a china shop. Got to love STF spot currency......remarks just cost me about $90 when price fell through the floor--not sure what that is in pips but around -8. Short more or less at where I went long last time. Must wait for the smoke to clear before doing the sums.
ETA: 11:20 AM EST. Smoke clearing, down about 11 pips after that little song and dance. Consensus may be emerging that Euro will sell off when Draghi doesn't announce bond buying tomorrow, unless it doesn't, but short in the interim (today) in any event. While I type this hear 2 targets hit without much effort. Notice neither NT Accounting nor Performance Summary picked up the last trade even thought it appears on the chart, whereas IB did...WTF now. Do not need the distraction--got to get NT moved to a new machine in case that helps, but also considering using another platform for manual trading since lately I've been spending more time trying to debug NT than writing strats (you know you're getting bored with a platform when second guessing it isn't as much fun as it used to be)--keep NT for when I need a dog that does tricks, cuz right now that dog don't hunt.
ETA: 12:22 AM EST. After another couple of successful long trades that Performance Summary apparently also didn't capture (Generate did not cause the summary to update, trades did not appear in the Trades tab), discovered that if I turn off Performance Summary Advanced display a scroll bar magically appears, as do the trades, at which point Generate will update the Summary. I'm an idiot for assuming NT's Account Summary Trades window scrolled as required. Still scratching my beard over why trades that are not initially displayed are not used in Accounting calculations until they are displayed, however.
NT banished to the doghouse for now--a stripped down 10.6" notebook I take on motorcycle trips (Intel I3 4 core, 3.3 GHz, 4 Gig memory, 450 Gig disk), attached to an 18" TV/LCD monitor for my 3-chart system (200 Tick, 600 Tick and 1800 Tick charts). Not sure how I'm going to post screenshots yet. Still working on the ergonomics of my trading desk with the notebook & cables now part of the clutter. I need reading glasses to see the notebook screen itself and will trade on the paper account today while working out any kinks.
First trade is long 1.2 contracts at the 3rd stddev of the most pronounced overnight activity mode, entry at mid to lower part of the channel (1.26095) just above yesterday's close. In hindsight could have entered at the bottom of the channel (1.2604, just above yesterday's open 1.26015) :-/. Stop at 1.26005, 1 target at 1.26185. Will set up a network share to post screenshots ASAP but right now need another cup of coffee.
At first glance today's session appears to be a minefield of scheduled news events between 7:45 AM EST and 10 AM EST (ECB Interest Rate, ADP Employment Survey, ECB press conference at 8:30 EST, ISM non-manufacturing). In other news, with gold spiking presumably in advance of ECB announcements may need to take some time off today to babysit my gold holdings (miner ETFs for the most part) in case Draghi does not deliver.
ETA: 6:47 AM AST stop moved to 1.26105 (breakeven + 1 pip) with price now at 1.2616.
ETA: 6:54 stopped out at +1 pip for a profit on paper of $4.60. So far 1 hit, 1 run, no error (unless we count the less than ideal entry an error) and no sign of trouble from NT
In the shot below notice the activity sidebar (red horizontal lines on LHS of the chart) reset when I updated the dataseries to print trade markers on the chart. Need VolumeProfile indicator with persistence.
ETA: 8:11 AM AST debating whether to sit on my hands now until the ECB press conference at 9:30 AM AST (8:30 EST) or try to trade the range, keeping in mind that when the music stops (Draghi speaks) one had better be able to find an empty chair. Considering long entry in the vicinity of 1.2600 if price drops that far, but with extreme caution since the acts opening for the main event (i.e., ECB bond & pony show in the center ring; e.g., inevitable leaks, rumours of leaks & ECB interest rate statement leading up to it) are going to get the crowd all worked up. Suspect the range will simply tighten as tension mounts, unfortunately.
ETA: 8:45 AM AST ECB leaves rate unchanged. Crowd goes wild for a moment or 2. Short at 1.26405, stopped out at +1 pip for another $4.60 profit on paper. Story of my life. Short limit again at 1.26315....waiting. Order accepted, target at 1.26215, stop at 1.26365....stop moved to break even + 1 pip ....stopped out. More or less exercising NT rather than trading to see how the new configuration deals with relatively high order frequency and tick bursts; so far so good (IB & NT agree 3 for 3 scalps, up about $30 after commission on paper so far today)
ETA: 8:58 AM AST appears the interest rate statement caused EUR/USD to migrate, establishing a new, higher mode, median roughly 1.26345. IIRC there were some interested sellers around here earlier today. Some think an uptrend will resume if we break above this price, en route to 1.2740, but barring leaks or any reaction to ADP survey at 9:15 AM AST price likely parked here until 9:30 AM AST when the ECB press conference is supposed to begin. In my youth might have bracketed the new range with OCO stop orders, but older than that now :-/
ETA: 9:15 AM AST slight (negative) reaction to (positive) ADP nudging price lower, possibly due to USD pop, expect to equilibrate back to 35. Guessing whatever legs EUR/USD gets at 9:30 will be first due to EUR buying/selling, then to pressure from USD.
ETA: 9:52 AM AST Watching the EUR/USD it seems the balance of "good news" for the US economy and what may be the usual "no news" for the EU may be overall bearish for the pair, calling to mind yesterday's "consensus" that if Draghi simply kicks the can the Euro will suffer. Short at 1.2586, stop just above yesterday's high (38 pip stop loss, well above my pain threshold) on spec (not sure how to define a high probability trade at the moment except perhaps in terms of fundamentals, and wouldn't be trading at all if it weren't paper and thus an opportunity to experiment with bigger picture strategies ). If the pair breaks down through the floor pivot (1.25783) have 1.2565 & 1.2495 pegged as technical support below that (based on 1800 tick chart). Shock yet to come: ISM non-manufacturing at 11 AM AST (10 AM EST), 30 minutes into N American markets open.
While the floor pivot may be a price everyone is looking at, MM 0/8 (ulitmate support) is just below at 1.2573 on all 3 of my charts. Probably not possible to deconvolve the significance of one from the other given their proximity to each other (having a background in geophysics tend to view effect of S/R on price in terms of impulse response, reflections, refractions and interference patterns, hence a deconvolution issue), so we'll chalk it up to coincidence.
ETA: 10:34 AM AST. In other news, a few minutes into the open Canadian resources, energy and bank instruments seem to be responding favourably to the premarket hubbub, but will be keeping an eye on them, and the DOW to see how earnestly it attempts to close the opening gap. My guess is--it won't try very hard, which may not bode well for my short EUR/USD trade (going nowhere at the moment) unless EUR/USD decouples from the N. American market, possibly in response to significant pressure on the Euro by other parties. At the moment there is some sign that may be happening (EUR/USD declines when e.g. DOW declines, does not recover when DOW recovers), but the trading day is young.
Regarding the short trade, stop moved to a pip or 2 above last retrace to minimize the pain to -10 pips if the Euro's plight should attract an angel (coincidentally it did bounce off first support level mentioned above, 1.2565) or ISM non-manufacturing enough to trigger significant risk-on.
ETA: 11:00 AM AST Hands on deck for ISM non-manufacturing. Haven't looked at CNBC (they're still yakking about the ECB, apparently ignoring what's happening in the US) but DOW response suggests good news. Hard to say if decoupling will be enough to save the trade. Turns out ISM was 53.7, 1.2 above expectations (got that from Babypips). Spider sense is saying "bail now, while you have a chance", but we'll do this by the book. Wonder if knots in the stomach are good for the abs.
ETA: 11:30 AM AST Time of death: 11:30 NT & IB report we're down $108 for the day on paper and after commission after that gambit, guessing too much pressure from risk-on (as feared DOW didn't look back, moving upward like a freight train), EUR/USD back in sync with N American markets. Seems my notion of "big picture" trading needs tweaking. Will watch at least until end of session (12:30 AST), maybe try a few "5 energies" trades for practice, console myself with my real money N American stock market holdings
ETA: By 11:57 first "5 energies" trade will have made back all the losses and more, even if (or when) it gets stopped out.
ETA: 12:55 PM AST. Short EUR/USD at 1.2626, stop at 1.2664 (on paper) interpreting latest action as double top and anticipating retest of post Draghi low of 1.2661 at some point--another "big picture" trade, bearing in mind how the last one ended.
Please host images on futures.io (formerly BMT) instead of third party servers. Your images are painfully slow to load, and/or don't load half the time.
The good news is to date NT and IB accounting agree, suggesting NT is so far stable on the laptop. The bad news is they agree trading paper the last few sessions was a total bust--e.g., overnight trade alone stopped out for a loss of 38 pips on 1.2 contracts. I must be batting 2% on "big picture" trades, particularly when held overnight.
It's fun and educational to experiment on paper--admittedly more fun if one wins--in this case the lesson being that my grasp of "big picture" spot currency trading has some holes in it. If we trade an unproven model of the market (trade assumptions rather than facts) sooner or later we are confronted by reality, or as Shakespeare put it more or less, without extensive backtesting we soon discover there are more things in heaven and earth than are dreamt of in our philosophy--in other words those pesky unaccounted for extraneous variables and methodological inconsistencies will get us every time. Additionally one can't trade principles & patterns that may apply over longer time frames first without having proved the principles & patterns are sound; and second, without money management appropriate to longer time frames. Especially don't try to trade champagne patterns (that apply over the kind of time frames hedge funds & central banks are fond of) on a beer budget. So it's down to serious business--rebuilding the paper account to its former glory doing what I know works, a simple matter of earning ca. 150 pips :-/. By the end of it hopefully faith in my methodology and in NT will be restored.
Journal/futures.io (formerly BMT)Journal20120904/TypicalTrade.png)
) price will now meander upward perhaps as high as 1.2590 in the 1st 30 minutes after N Am. market opens if it gets that far before the ISM numbers come out, rather than rattle around down here, so will likely chance a long.Journal/futures.io (formerly BMT)Journal20120904/Oops.png)
Journal/futures.io (formerly BMT)Journal20120905/BomdNews.png)
Journal/futures.io (formerly BMT)Journal20120905/8MinutesToOpen.png)
Journal/futures.io (formerly BMT)Journal20120905/SetItFree.png)
Journal/futures.io (formerly BMT)Journal20120905/FlyingDog.jpg)
