The figure below showing charts for spot AUD.USD summarizes the current look & feel of my trading setup before it gets pruned to whatever form it will take -- very noisy.

One recent change was to replace the 1 day chart with the 240 minute chart (far right).

Setups are done by eyeballing the 240 minute, 1800- and 600-tick charts. Entry chart is the 200-tick (chart 2nd from the right of the 4 shown), confirmation on the 600-tick chart (2nd from left).

Like many, many others I trade primarily reversals, which tend to deliver the most bang for the buck, and trends. Range and chop (scalping) not so much anymore because I find them higher stress.

The setups are pretty much identical to what many, many others use --some sort of momentum/cycle divergence followed by momentum/cycle/price reversal confirmation. In the case of the trade underway in the figure below weak divergences can be seen in 1800-, 600- and 200-tick charts between MACD/Stochs and price prior to entering the trade. For better or worse I didn't wait either for price confirmation or for 240-minute stochs to turn this time before entering.

If signals support it I'll double up on the entry if it initially goes against me (happened here).

Study continues into how to manage the trailing stop

Regarding trailing stops, or even reversing the reversal trade if always-in, the figure below illustrates the dilemma ---twiddling thumbs and tying up capital while waiting for price to negotiate a spot of bother and either resume direction and hit the 3rd and final target (1/2 contract runner just above the floor pivot at 1.04295) or the trailing stop (at the 15EMA + 1 ATR on the 600-tick chart):



Edited to add: coming from a scalping background and almost 3 HOURS INTO THIS TRADE can't get rid of the song "Hot Rod Lincoln" that started playing in my head some time ago...in this case probably the part about "drive me to drinking"

Bottom line is that 1/2 contract, if hit, will pay something over $250 on top of what's already in the bank from the previous targets, whereas I could take about $50 now. May boil down to what our time is worth. In any event time to apply the balance of the account to potential opportunities in other pairs and let this trade fend for itself.

Another one just like the other one. Finally ran out of patience and closed AUD.USD for 24 ticks or so, life being tooooo short, jumping from the frying pan into the fire trying to trade a measured move on USD.JPY. Fortunately USD.JPY finally found its fingers.

1st figure shows the trade on the 200-tick chart (2 contracts, 4 targets).



2nd figure shows the bigger picture (measured move setup on 600-tick chart)



Edited to add: I noticed long after I peremptorily exited the AUD.USD trade that it would have been stopped out anyway. Nothing more maddening than waiting hours for a trade to complete (on principle, say) only to have price reverse and take out the stop. This raises the oft-debated question of whether one should close automatically after a given period elapses. I'm of 2 minds but agree a trade should be closed before one passes out from unmitigated tedium (either snoring loudly, head over the back of the chair, mouth agape and drooling like those unfortunate denizens of intercontinental redeye flights, or forehead on the keyboard sending an incessant string of characters to a bewildered trading app which finally crashes, having struggled in vain to decipher what you mean by ".......ttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttt....")

For future reference: GC front month vs cumulative delta @Silver Dragon

I placed 26 orders today, starting late (slept in until 5:30 AM EST) with an accidental order before I'd finished a first cup of coffee to sell 1 unit (as opposed to 1 contract, or 100,000 units) of USD.JPY. As explained below it took a while to realize the order had even been placed.

Things improved somewhat after that but overall it was an inauspicious day.

TWS's trade summary shows 400,001 units of USD.JPY were bought and sold in 7 transactions while managing that first unit traded. Total Yen value traded was ¥36,512,091.27, which grossed ¥6,250.02 (Cdn $68.70) or about $51.20 after commissions. The high number of transactions was partly due to the fact it took me a while to realize the beta version of MC I'm using was not acknowledging entries even though it was sending the orders to IB and I had managed to create a sizable position by repeating the entry a number of times more or less blind. It takes a while to shut MC down and restart it in an orderly fashion so I spent some time maneuvering the position using TWS to close it without a loss.

MC unfortunately has no memory of the incident-- no historical order indication on the charts and no Trading Performance Report for USD.JPY today.

TWS also shows trades in AUD.USD (the first trade a very public loss detailed in the home page chat box) and GBP.USD. AUD.USD trades (14 orders!) grossed -$1.50 or -$34 after commissions. GBP.USD grossed $108.65 in 2 trades (4 orders) or $97.30 after commission.

I cut the session short at 10:15 AM EST up $148.55 after commission because it was becoming clear I was in no shape to trade, and spent the rest of the day reading material on futures.io (formerly BMT).

Later I attended a 2 hour webinar sponsored by the CBOE featuring Jan Arps ("Scanning for Profitable Setups in Futures, Forex and Stocks"), Alla Peters ("Fibonacci Price Waves"), Justin Weinraub & Dave Johnson ("Relative Aggression Bars", a system built around what seems to be an interesting way to paint bars as a function of volume delta and a velocity metric) and Kevin Davey who talked about 5 unarguably effective ways to improve trading if one is not already practicing them. I like Kevin's definitions of Do It Yourself'ers and Expert Followers

I went through a stage (about the time I was learning my first system) where I couldn't bear to read about other traders' methods. The differences were a distressing distraction from principles I was labouring to grasp. These days all I see is similarities between systems and can't get enough of them, but solely for the purpose of putting my own system in perspective. For that purpose IMO I don't have to look farther than futures.io (formerly BMT) to learn all I need to know.

Started late again today, first trade between 9:14 & 9:20 EST (trend reversal trade, 1 contract GBP.USD, stopped out & lost $66.00 after commission, Figure 1 below), second trade 9:34 - 10:32 EST (momentum based range trade, 1 contract EUR.GBP, hit both targets and made $233 after commission, Figure 2 below).

Nursing a Cdn gold miner ETF since N. American stock market open put the kibosh on more currency trading; presume I'll multitask better once I trust the ETF trading system (such as it is) and am able to relax.

The interesting thing about the 2nd trade, aside from the fact it was never in trouble and made money, was I had time to add horizontal lines at what seemed to be pivot-based potential resistance levels that might be encountered on the way up. While I know some folks feel such lines are pointless, at this stage in my development it's educational to watch price action in their vicinity--especially allows me to maintain equanimity when price otherwise (without the lines) out of the blue seems to run into a brick wall, if only temporarily

Figure 1: GBP.USD stopped out


Figure 2: EUR.GBP 2 targets

MC .NET ES CD

VP/CD influenced spot trade following employment news, long AUD.USD on MC while watching @AD# on MC .Net. VP/CD shown below, actual trade in progress below that (trade was subsequently stopped out at 15ema - 1 x 15ATR)

VP/CD



Trade:

MC's version of ES at the close today

I slipped back into "research" mode over the weekend, undoing sleep patterns in the process. As a consequence no trading today.

The main endeavor was to expand the wave counting algorithm (TDWave MC indicator) to monitor momentum and cycle divergences. In the process a new MC indicator derived from TDWave was prototyped that combines wave counting and divergence ("TDDivergenceWave") and draws arrows on the chart where it concludes there is some chance of a long or short setup (or target) occurring. The good news is the code as it has progressed so far appears to be be bug free, can ferret out divergences I would never spot (remains to be seen how useful these obscure divergences are) and when it's right it's dead on. It also appears to be able to handle "limit order markets" (i.e., where bars are overlapping, characteristic of trading ranges including chop) and I'm hoping it will support the only kind of breakout trading I'm interested in; namely, a limit order reversal (or counter trend) trade that keeps going (through the roof or the floor), whereby one begins to take profits about the time "breakout specialists" are entering and hence accelerating the move.

Bad news is it doesn't yet assign a quality factor (or probability) to its deductions and generates a relatively high number of low probability signals based on a single, questionable divergence, which may fall yet into the category of "take profit" signals as opposed to entry signals. Some of the low probability signals are related to a suspect part of the algorithm that for the time being has been assigned responsibilities beyond its pay grade, so to speak (it is making strategy level decisions based on indicator level information for the purposes of this evaluation).

The screenshot below shows the prototype running on a 200 tick USD.JPY chart, for now large cyan numbers associated with arrows indicating the nature of the divergence (i.e., short or long /enter or exit conditions based on stochs low/high, macd high/low divergence or some combination). The pair [2 short => 10 long] highlighted by the blue box illustrates what I mean by the sort of breakout trading I favour--short entry at the top of the chop, exiting only once momentum decays. The alert for a subsequent decision point [2 short] can be seen in the lower right hand corner of the screenshot.

Other work over the weekend involved comparing COT reports for currency futures with price data over the last 13 months to study how reported positions correlate with price action and metrics of price action regularly used in trading (from conventional indicators to volume profile and cumulative delta), in particular if any conclusions can be drawn about price action ensuing from flat positions. Some effort was made to model Fib retracement and extension in terms of reported positions, for example given that in theory institutions that are always-in (with-trend in the case of producers/users/consumers of the underlying versus counter-trend for pretty much everyone else) whether continuously buying or continuously selling maintain identical break-even price point equal to 1/2 the maximum price excursion (e.g., the 50% Fib line). That work is ongoing.

Divergence indicator on current 200 tick USD.JPY chart