NexusFi: Find Your Edge


Home Menu

 





STF discretionary spot Forex system development journal


Discussion in Trading Journals

Updated
      Top Posters
    1. looks_one bnichols with 249 posts (152 thanks)
    2. looks_two Big Mike with 17 posts (4 thanks)
    3. looks_3 Adamus with 14 posts (3 thanks)
    4. looks_4 NW27 with 6 posts (0 thanks)
      Best Posters
    1. looks_one sam028 with 4 thanks per post
    2. looks_two bnichols with 0.6 thanks per post
    3. looks_3 Big Mike with 0.2 thanks per post
    4. looks_4 Adamus with 0.2 thanks per post
    1. trending_up 54,789 views
    2. thumb_up 167 thanks given
    3. group 18 followers
    1. forum 292 posts
    2. attach_file 182 attachments




 
Search this Thread

STF discretionary spot Forex system development journal

  #271 (permalink)
 
bnichols's Avatar
 bnichols 
Dartmouth NS
 
Experience: Intermediate
Platform: MC, MC.Net, NT, TWS
Broker: IB / IQFeed / Kids
Trading: Forex, stocks
Posts: 637 since Feb 2010
Thanks Given: 64
Thanks Received: 460

For @Xav1029

6EVP + spot EURUSD (FXCM)


Visit my NexusFi Trade Journal Started this thread Reply With Quote

Can you help answer these questions
from other members on NexusFi?
NexusFi Journal Challenge - April 2024
Feedback and Announcements
Exit Strategy
NinjaTrader
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
NT7 Indicator Script Troubleshooting - Camarilla Pivots
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
24 thanks
GFIs1 1 DAX trade per day journal
22 thanks
The Program
19 thanks
  #272 (permalink)
 
bnichols's Avatar
 bnichols 
Dartmouth NS
 
Experience: Intermediate
Platform: MC, MC.Net, NT, TWS
Broker: IB / IQFeed / Kids
Trading: Forex, stocks
Posts: 637 since Feb 2010
Thanks Given: 64
Thanks Received: 460

Another view of EurVP


Visit my NexusFi Trade Journal Started this thread Reply With Quote
  #273 (permalink)
 
bnichols's Avatar
 bnichols 
Dartmouth NS
 
Experience: Intermediate
Platform: MC, MC.Net, NT, TWS
Broker: IB / IQFeed / Kids
Trading: Forex, stocks
Posts: 637 since Feb 2010
Thanks Given: 64
Thanks Received: 460



Visit my NexusFi Trade Journal Started this thread Reply With Quote
  #274 (permalink)
 
bnichols's Avatar
 bnichols 
Dartmouth NS
 
Experience: Intermediate
Platform: MC, MC.Net, NT, TWS
Broker: IB / IQFeed / Kids
Trading: Forex, stocks
Posts: 637 since Feb 2010
Thanks Given: 64
Thanks Received: 460

Preamble

I didn't make any manual trades between 22 February and last week. The time was spent instead meditating upon forward tests of built-in NT strategy SampleMACrossOver and a strategy proposed by one of my sons, which turns out to be along the lines of the "breakout/momentum" strategy developed by @kevinkdog in his recent nexusfi.com (formerly BMT) webinar, applied to a portfolio of 6 currency pairs, which activity (observing forward behaviour of simple bots--the "everyman" of the bot world) was not about the test but like having a radio on in the background while staring at wallpaper.

Since these periods of introspection invariably lead to a deeper appreciation of what I think I know about trading the exercise brings to mind the "10000 hours--really?" debate about how long it takes to learn how to trade and how to measure the hours, what time should be counted if there is such a thing as wasted time and productive time. IMO in fact all we can say about time & how we spend it is that our days are numbered and having time to spend time beats the alternative--it is what it is. In that regard I expended some effort hypothesizing that if a conventional bot were to become self aware in the trillionth of a second a billion times a second it manifests its raison d'être (the time it takes to execute the "if...then" statements comprising its thought process and which presumably underlie its epistemology) what its worldview might be like, including whether it experiences repeated awakenings (each awakening "rounded by a little sleep" according to the famous playwright BotSpeare ) and if so does this sense of eternal recurrence suggest its being is embedded in a higher plane of existence and even the existence of a Programmer whose being nevertheless may be embedded in an even larger plane of existence ad infinitum--a foray into the Flatland of machine consciousness. Hard to say whether this should be categorized as productive time or time wasted if I were inclined to categorize it, which I'm not.

Consequences of the Forward Test

First, I'm more convinced than ever bots of this sort, that generate trades via a few rules applied to a few indicators, are not my cup of tea. I can't help peeking at their progress and the idiotic trades they are prone to during inevitable losing streaks (even while they are operating within spec let alone when they start to fail) still makes me want to bang my head against the wall. Mainly however, while I may imagine I want to lie on a beach with an umbrella drink in hand while a bot does the heavy lifting, in fact I don't. In my case having nothing to obsess about is far more stressful than watching bots behaving badly: if trading is boring not trading is even more boring.

As expected SampleMA does well when out of band price statistics match training set stats, poorly when they do not. When optimized on data reflecting overall stop order market conditions the strategy behaves like a trend trader. For trend trading optimally fast MA period << (very much less than) slow MA period, which leads the strategy to buy when price apparently starts to rise and sell when price starts to fall. When optimized on data reflecting limit order market conditions the strategy behaves like a range trader (ideally fast MA period >> slow MA period), selling when price starts to rise and buying when price starts to fall.

The example shown in following figure illustrates this behaviour for SampleMA trained on trend data when the price action changes to "chop" (cycle period falls below the average period of the training set). In time series analysis the situation could be thought of as the emergence of dominant frequencies higher than the Nyquist frequency giving rise to aliasing in the response:



If watching bots is like having the radio on, watching SampleMA in action is like tuning the radio to a favourite station. MAs still fascinate me more than anything except perhaps Fibonacci patterns--the trading equivalent of cornstarch and water mixtures, slinkies, gyroscopes, etc.

MAs (like all "lagging" indicators) are occasionally dismissed as useless precisely because they lag price action, which IMO is rather like criticizing a leopard for its spots because it doesn't have stripes like a zebra; if one wants something that doesn't lag price then one might consider learning how to trade price. In real life lag is the basis of all confirmation and the only issue is whether to pay the insurance of a few ticks (the cost of confirmation) or to depend on intuition. I suppose MA crossover systems ideally ought to be tuned to a particular market so that crossovers occur precisely at a time and price corresponding to collective point traders en masse are reaching their individual decision points, reminiscent perhaps of Gann's strange geometry, suggesting new avenues of research--new ways to invest those 10,000 hours.

Resources

On the topic of tools I was reminded that if IQFeed did not exist it would be necessary to invent it. IB's historical data is usable if one starts loading data a couple of days ahead of time, but when doing research I'm not always sure initially what instrument, time frame or bar type I'll be using so this often implies loading data before I know I need it and makes me wonder if IB policy was designed by a practitioner of the Welsh art of self defense at least as defined by someone with a jaundiced view of the Welsh** ( LLAP GOCH, the art of thuggery: "attack your enemy before s/he's even aware you exist").

**I should hasten to add personally I have no issues with Welsh, having very much enjoyed time spent years ago at University College of North Wales.

NT was used for the forward test but IMO still surprises with too many WTF moments (in this case oddities arising trying to test with limit orders and more advanced order management). I'm prepared to admit as others have suggested it may be my lack of experience, lack of intelligence and/or poor fashion sense even after several years battling NT, but if life is short from my perspective MC's relative stability makes life at least seem somewhat less brutish.

Further Research

What the exercise did do was strengthen my conviction that while entry is important, trade management is more important. In conventional entry/stop/target trading trade management boils down to risk/money management but in this case meant time spent subsequently live trading paper soley for the experience of managing an average position around price action by decreasing a position as required to lock in P&L and free up capital and by increasing a position to adjust average entry price and lever P&L.

This approach to trading differs in effect not at all from any other approach but was perceived in the sense of keeping the account directly engaged to price action (i.e., always-in) via a trading system that might be modeled as the transmission of some sort of Rube Goldberg machine, operated in the effort to be on the "right side" of price when it is trending between value areas (especially when reacting to value area highs and lows) so that account balance increases whether price increases or decreases. That is, the machine transmission could be thought of as in "forward gear" when balance increases as price increases and in reverse when balance increases as price decreases. From this perspective first results suggest shifting between forward and reverse involves a temporary martingale as trends decay into consolidation and/or reverse, which is where profits and losses occur in the conventional sense (i.e., are locked in), the minimization of which (unlike negative expectancy games) depends ultimately on experience. That is not to say account balance does not fluctuate in terms of conventional unrealized profits and losses as price fluctuates, unrealized losses perhaps equivalent to grinding the gears of the machine.

As an aside, for me the longer term aim of always-in research is a deeper understanding of Fibonacci patterns. In the process I finally experienced what it is to trade for the sake of trading, which is to say not for the money.

Next Steps: Flocking

As mentioned in a previous post in this thread (or perhaps elsewhere) these days my interest in bot development tends toward developing a trading engine based on a model of the processes that express themselves as price action rather than an engine that reacts to price action (i.e., reactions guided by heuristics derived from price action metrics or informed by one pattern classification scheme or another). In other words, e.g. instead of teaching a bot to trade Market or Volume profile (concepts that nod in the direction of auction theory), I'm curious whether a bot can trade auction theory from first principles.

Price action explained in terms of auction theory from the perspective of modelling calls to mind emergent behaviour (collective behaviour of an ensemble of autonomous individuals governed by a simple set of rules; e.g., flocking of birds). Part of last week was therefore spent unearthing where economic & finance theorists went after floating the "efficient markets" lead balloon in the 1970's, in particular whether they had anything to say about flocking.

Like all mathematicians quants seem compelled to sweep the empirical origins of their ideas under the rug, expressing everything in canonical form (for rigour, they say, whereas to the untrained eye canonical form might also make their work seem less mundane than it actually is and themselves seem more important than they actually are). Once I discovered that (in the spirit of dispensing with the origins of an idea) they'd decided to abstract the notion of flocking in terms of networks of "heterogeneous agents" (and "Heterogeneous Agent Models", or "HAM") it became apparent they've been very busy. In contrast to the mostly useless "postulate a spherical cow" approach by most academics to what moves markets, in particular research by Buz Brock and Cars Hommes is refreshingly precise and grounded in reality, reinforces some naive thoughts I had about how flocking might apply to price action (see e.g., https://www.tinbergen.nl/discussionpapers/05056.pdf). If nothing else saves me exploring caverns of thought that have already been thoroughly spelunked by others.

Visit my NexusFi Trade Journal Started this thread Reply With Quote
  #275 (permalink)
 
bnichols's Avatar
 bnichols 
Dartmouth NS
 
Experience: Intermediate
Platform: MC, MC.Net, NT, TWS
Broker: IB / IQFeed / Kids
Trading: Forex, stocks
Posts: 637 since Feb 2010
Thanks Given: 64
Thanks Received: 460

Preamble

I've been trading paper almost exclusively since the last post mainly to experiment with variations on the method for short time frame spot currencies--as usual essentially Barry Burns' 5 energies system augmented occasionally by Wyckoff methods, regression channels, Fib retracement lines, "volume profile" and cumulative delta.

The other reason I've been trading paper is I'm still digesting the death of my father. We were not close. 3 things his death underscored for me were how precarious life is, how fleeting its enjoyment, and how thin a trader's edge.

"Volume profile" is in quotes because I'm referring to profiles obtained by applying MC's VP indicator to Interactive Brokers FX tick data to plot total ticks and up vs down tick counts. While those who like to point things out may point out that IB's tick data is worthless and there is no such thing as FX volume on which to base "volume profile", in fact the resulting profile (more properly a time-price-opportunity profile or TPO perhaps) resembles actual volume profiles obtained for currency futures based on IQFeed tick data closely enough for my purposes (i.e., VAL, VAH and POC are pretty much in total agreement once the front month futures/spot price difference is taken into account).

I want more experience with Wyckoff's concept of reversals but with vacation season upon us it's time to buckle down to pay for a planned month long motorcycle trip across the continent (NS to Oregon and back), especially since my camping days are over and the thought of staying in low cost no-tell motels no longer appeals. I'm also starting to reconstruct an old 17.5 inch reflecting telescope, a little like the Cheshire Cat of Alice In Wonderland of which all that remains is the primary and secondary mirrors. In other words it's time to trade in order to afford to do the things I'd rather be doing.

The last few days trading cash suggests the experience this time around is focused on trading setups, not trying to trade every move price makes (i.e., not so much scalping). For the first time it seems what price does between setups is of no consequence except to the extent that what price does contributes to the setup--I no longer have feelings one way or the other about "missed" moves. I may still favour counter-trades and range/consolidation trading over with-trend trades for the relative excitement (to increase trading frequency and to capitalize on better risk/potential reward during reversals) but even though patience never was my strong suit now seem able to wait patiently for momentum to give a clear signal.

The number of spot currencies I monitor is now 2, (EUR/USD and USD/JPY at the moment) down from 9. Haven't yet come up with a way to monitor than 2 without distracting from the focus and subtracting from the time required with each one to trade enjoyably at my level of experience.

Bots
I continue to forward test / walkforward-optimize a strategy proposed by my oldest son against spot NZD/USD (his most hated instrument--what is a bot for except to do what we hate to do?). IMO its behaviour in walkforward testing is interesting enough to continue to gather data but since it's his bot he has the final say on disclosure.

I've done nothing with flocking. Days are brighter, victims of the northern lifestyle everywhere are turning off their lamps, packing away their snowshoes, summer after all soon upon us, thought eventually turning to things that make less demand on thought.

Software
I've attached 5 recent experimental projects to do with time series analysis and flocking, intended for experienced programmers only. All app executables are in the associated .../bin/debug directory, which means they need to be optimized for real-world (Release) use.

- FFTFormsApp exercises alglib.dll and will compute the Fast Fourier Transform of a real time series and output the amplitude spectrum to a file. Input file format is the same as used in the last project posted. I've used it to compare price spectra with spectra computed by more recent methods (Hilbert transform, e.g., Ehlers methods) but mostly like it because it does not care if you forget to observe powers of 2 when choosing the length of the input time series

- Flocking_Boids implements a bulls/bears versions of the familiar Boids flocking algorithm (2 flocks driven by different aversion criteria)

- EMDV1 is a C# program that calls a .DLL version of the Hilbert–Huang transform translated from a version C++ written by an author whose name escapes me at the moment. After some experimentation IMO the algorithm does not approach the sophistication of the R-Language implementation, but the translation is a start.

- EMDV2 is the C# version of the Hilbert–Huang transform by the same (unspecified author) as a .DLL callable from MC.Net (or NT)

- EMDTest.cs is an NT indicator that calls the DLL from NT. Note that my experience so far with the indicator is that it is useless at the moment--included for completeness (as a template).

Attached Files
Elite Membership required to download: FFTFormsApp.zip
Elite Membership required to download: Flocking_Boids.zip
Elite Membership required to download: EMDV2.zip
Elite Membership required to download: EMDV1.zip
Elite Membership required to download: EMDTest.cs
Visit my NexusFi Trade Journal Started this thread Reply With Quote
  #276 (permalink)
 
bnichols's Avatar
 bnichols 
Dartmouth NS
 
Experience: Intermediate
Platform: MC, MC.Net, NT, TWS
Broker: IB / IQFeed / Kids
Trading: Forex, stocks
Posts: 637 since Feb 2010
Thanks Given: 64
Thanks Received: 460

Yet again I'm struggling to repair my sleeping habits, the body having apparently settled into inconvenient Circadian rhythm from my youth, wanting to lie down at dawn and get up around noon--inappropriate for trading the London market / US overlap. Not sure if the cause is psychological or biological but something has to change.

Since the last post I've been trading using an MC order placement/exit strategy combo similar to NT's ATM to remove some of the discretion from scaling out. At the moment the same exit strategy is automatically applied to any order the moment it fills to place stops, move stops to BE + costs when the first target is hit and trail until either the runner scales us out completely or the trailing stop executes.

Advantages of such an exit strategy (exit bot) in a pseudo discretionary system are the following:

1. stops and targets are canceled automatically when flat (for folks prone to leaving one or another dangling after a trade has ended, especially those prone occasionally to sitting down at the trading desk to discover they've been margin called overnight on a mysterious order they don't remember placing )
2. it reduces the urge to second-guess the trade if IQ sky-rockets as the trade progresses (which causes us to fiddle and/or exit prematurely). If we're able to bow to the authority of the bot and leave stops & targets alone a bot will save us from ourselves more often than not.
3. for better or worse stop and target orders placed at the broker remain in effect as last updated when the platform inevitably crashes
4. mainly it gives us something (targets, initial and trailing stop offset) to analyze and optimize

This doesn't mean that stops and targets placed by the bot can't be adjusted initially to account for specific price architecture, just that we have to estimate the impact of adjustments when analyzing results.

The disadvantages of an automated exit strategy stem from the fact automated strategies in general are idiots (at best idiot savants). While there may be the odd exception, built- in exit strategies in particular think in terms of tick offsets from entry rather than in terms of price geometry and dynamics routinely used in manual trading. Especially they have no knowledge of the trading system in use and therefore, having no knowledge of what conditions got us into the trade, have no clue whether present conditions still justify the position. This is not to say human traders don't suffer from the same deficiency (no clue when to quit), given that trading is about paying to test probabilities and hence the popularity of fixed offset stops & targets. Since a chart (like any optical illusion) can change from bull to bear in the blink of an eye it's probably best to let expectancy unfold, which is what fixed offset orders tend to encourage.

Figure 1. Bull or Bear?


A couple of EUR/USD trades executed between 10:56 EST and 11:14 EST this AM in Figure 2 below illustrate the exit strategy in action: 1st trade jumps the gun with a short entry in anticipation of price turning back down in the vicinity of the floor traders pivot (yellow dashed line)--2 contracts stopped out at -6 ticks (3 pips). While timing in momentum based trades of this sort may be a bit of an art entries should not be as badly managed as this one was, probably a consequence of lack of focus. I was paying attention for the 2nd trade shown, choosing to fade the stop run/bull trap above the pivot. First 2 targets (1/2 position + 1/4 position) were hit, runner eventually trailing stopped out.

As an aside, I recently heard someone suggest it may be best to get rid of anything on the screen that gives rise to conflicting signals. That may be, and the clutter of multi-timeframe charts and indicators in the figure below definitely generates conflicting signals, but it's what works for me and IMO "what works" trumps "what's best". IMO consensus is not necessarily a good thing; ambiguity of price action can be reduced only so far before we start trying to inflict personal belief on the market (and reap the well known outcome of being too sure of ourselves). I find noise generated by the configuration is an excellent indicator of market sentiment, entries tending to coincide with a lull in the cacophony.

Figure 2. Two EUR/USD spot trades each with 4 auto-placed targets, initial & trailing stops

Visit my NexusFi Trade Journal Started this thread Reply With Quote
  #277 (permalink)
 artemiso 
New York, NY
 
Experience: Beginner
Platform: Vanguard 401k
Broker: Yahoo Finance
Trading: Mutual funds
Posts: 1,152 since Jul 2012
Thanks Given: 784
Thanks Received: 2,685


bnichols View Post
Yet again I'm struggling to repair my sleeping habits, the body having apparently settled into inconvenient Circadian rhythm from my youth, wanting to lie down at dawn and get up around noon--inappropriate for trading the London market / US overlap. Not sure if the cause is psychological or biological but something has to change.

Have you been programming or staring at your screen a lot recently? It affects most programmers that I know of, because the brightness of your screen inhibits melatonin production, which in turn regulates sleep. You might also be suffering from SWSD, nacrolepsy or asthenia.

Reply With Quote
  #278 (permalink)
 
bnichols's Avatar
 bnichols 
Dartmouth NS
 
Experience: Intermediate
Platform: MC, MC.Net, NT, TWS
Broker: IB / IQFeed / Kids
Trading: Forex, stocks
Posts: 637 since Feb 2010
Thanks Given: 64
Thanks Received: 460


artemiso View Post
Have you been programming or staring at your screen a lot recently? It affects most programmers that I know of, because the brightness of your screen inhibits melatonin production, which in turn regulates sleep. You might also be suffering from SWSD, nacrolepsy or asthenia.

Screen brightness is a very good point--the brightness is cranked on 2 x 24" LED monitors, probably equivalent to a SAD lamp (seasonal affective disorder) and when the display darkens due to inactivity (i.e., when not trading, specifically when I should be sleeping rather than programming or watching TV/playing Spider Solitaire :-/) I find myself nudging the mouse to bring them back up for the brightness rather than to see the charts, suggesting I may be hooked on it. Just reduced the brightness considerably.

Visit my NexusFi Trade Journal Started this thread Reply With Quote
  #279 (permalink)
 
bnichols's Avatar
 bnichols 
Dartmouth NS
 
Experience: Intermediate
Platform: MC, MC.Net, NT, TWS
Broker: IB / IQFeed / Kids
Trading: Forex, stocks
Posts: 637 since Feb 2010
Thanks Given: 64
Thanks Received: 460

From the poem "To A Mouse, On Turning Her Up In Her Nest With The Plough" by Robert Burns (see e.g. the poem here), à propos tonight for a couple of reasons.

First, because I'm trying to tame a baby mouse that has been nosing around my trading desk while at the same time trying to eradicate its extended family with poison and a trap line set around the house. It's been quite the battle and while I felt initially I had the intellectual upper hand in such a simple matter, as optimism fades I'm beginning to suspect Douglas Adams in "HitchHiker's Guide" may have been right about the relative importance of humans and mice on this planet after all. [We were invaded recently coincidentally after a well-meaning neighbour took it upon himself to obliterate the local feral cat population.]

Second, it turns out at the moment I can't run MC's Master exit strategy mentioned in the last post without bringing MC to its knees--no clue why. Details in a general MC support discussion forum here.

P.S. last stanza of the poem is devoted to mice and trading:


Quoting 
Still thou art blest, compar'd wi' me
The present only toucheth thee:
But, Och! I backward cast my e'e.
On prospects drear!
An' forward, tho' I canna see,
I guess an' fear!


Visit my NexusFi Trade Journal Started this thread Reply With Quote
  #280 (permalink)
 
bnichols's Avatar
 bnichols 
Dartmouth NS
 
Experience: Intermediate
Platform: MC, MC.Net, NT, TWS
Broker: IB / IQFeed / Kids
Trading: Forex, stocks
Posts: 637 since Feb 2010
Thanks Given: 64
Thanks Received: 460


...for posterity.



Continuing....



Shock sets in....


Visit my NexusFi Trade Journal Started this thread Reply With Quote




Last Updated on May 7, 2013


© 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