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)
Profitpilgrim's Blend of CT Scalps & Trend Trailing (free ATMs)
Its options expiration Friday. A respected professional hedge fund trader indicates that option market makers need the ES to stay below themid 30s and above 2300 so bias is down this morning.
Can you help answer these questions from other members on NexusFi?
If you've tracked this journal you know I am attempting an ideal, that is to scalp small reversions against the immediate direction of price by reversing as it reaches proprietary S/R levels from which it is expected to yield 3 to 5 ticks in ES (10 in NQ and YM. If one's S/R levels are well defined and executions are without the slightest delay then such scalping can be very profitable, despite relatively high overhead from commissions. What wipes out reversion scalpers are trending days. Assume a day is trending up. It plows through multiple resistance levels, each of which cost the scalper loss, and it rarely gives a pullback to a resistance turned support so the scalper shorts all day and takes a huge loss. The more extended it gets the more his gut relies on conventional day norms, which normally revert after strong runs. The gut persuades the trader to short again and again. An aggressive revenge trader will ultimately double up near the end of his demise and blow his account out, as I did on the 15th (wrongly dated 2-14 on the previous P/L posting of a $3023.02 loss, which would have exceeded TopStepTrader 150k Combine max daily loss limit. (Hence I reset my sim to $150,000 as if buying a new combine.)
As explained in more detail elsewhere, in order to prevent my guts "cognitive bias" from mismanaging trades, I've developed Trend and CounterTrend ATMs for Ninjatrader 8 to grab the scalping profit yet hold a few remaining contracts which keep their original tight stops but allow the stop to widen and trail so that if the day trends some of those scalps on the right side of the trend will leave remaining contracts that profit enough from the trend to offset losses of trades in the wrong direction.
As today's title ask, "Can it be done?" Most would say "No a scalp is a scalp and a swing is a swing. Trying to turn scalps into swings is trying to use a short term signal for a long term target." My answer would be "That's true unless you scalp enough profits from the scalped contracts in a multi-lot trade to leave a few riding long and short contracts in the market, one set of which get stopped at their tight stop while the other set enjoy a profitable swing trend toward major S/R levels or into end of day. If the day doesn't trend that's fine too, because each trade entry set is net profitable and the trailing contracts offset each other."
If it is possible the math must be correct and executions precise. On the 17th I have skewed the 5 lot ES Trend ATM to leave 2 (not 1) trailing contracts while scalping the other 3 at 5 ticks. I also skewed the ESCtrend ATM to take 2 (not 1) scalp at 3 ticks, 2 (not 3) at 5 ticks, with 1 trailed contract.
Results will be forthcoming.
Prosper with Disciplined Cheer!
(post thanks HERE)
Given another hard loss trending day, it would appear that I must impose on myself more stringent rules than the "Decisive Decision" in my 4th post, which imposed walking away from my computer for 5 minutes after 3 consecitive losses. Because despite my self-awareness of a tendency to continue taking counter trend sell signals into a hard rally (and vica versa), and given that statistical odds favor the continuation of trend, this week
1) I will not allow myself to countertrend enter more than 2 losing trades in a row without taking a trade in the trending direction.
2) I will also not allow myself to leverage more contracts in any one instrument than the initial 5 lot if I have lost the previous 2 trade sets.
3) Only twice a day at major S/R levels will I allow myself to use the ES Trend ATM when the trade is counter to the trend.
I regard these restrictions not as punitive but protective respect for trend following statistics too often overruled by my discretionary countertrend opinions about expected market direction.
A losing day Monday would disqualify a $150k account as beyond Max Drawdown. Let's see if this precautionary respect for trend direction recovers the account and turns aggressive experimental losses into real, long-lasting success.
Prosper with Disciplined Cheer!
(post thanks HERE)
The sixties pop song said "I've looked at love from both sides now, from win and lose, and still somehow, its love's illusions I recall I really don't know love at all." The same could be said of trading. Having written a rigorously tested advanced 3000 line algo in low level language that dynamically adjusts to volatility 3 different ways, reads order flow, applies limited martingale, dynamically adjusts targets and stops, and uses different parameters for longs and shorts, etc, I've looked at trading both ways -- as a professional algo writer, and as this journal indicates, as a discretionary retail trader inventing a low risk, high reward strategy of aggressively scalping while accumulating profitable swing contracts left over from the scalps in order to ride the prevailing trend to sizable profits, a bold and unconventional undertaking that, despite conventional wisdom, I hope proves very profitable.
The difference between the programmatic versus discretionary "love" of trading, could be likened to the difference between carefully anchoring a mechanical robot to a bucking bronco with the hope it doesn't break, and riding it yourself, which requires momentary adaptation to constantly changing perceptions which, reciprocally, strongly influence and are strongly influenced by psychological emotions, to wit, fear, pride, excitement, and the ill-advised aggression into which those emotions can easily sublimate.
Because this journal chronicles experimental adaptations being made to discipline my counterproductive discretionary trading impulses, one of which is not more favorably regarding the direction of the trend and placing too many countertrend trades, I detailed 3 new rules in my previous post to obligate more trading with the trend. But how should the trend be defined? What timeframe is appropriate? It depends on volatility. The higher the volatility the higher timeframe should be referenced to determine trend.
Because I very seriously intend to succeed at this strategy, yesterday I spoke with a more experienced Ninjatrader programmer with plans to hire him to code drawing support and resistance levels automatically instead of manually. But I want those support and resistance levels to be drawn on the prevailing cycle's timeframe and I also want to determine the trend from that prevailing timeframe, not from a statically chosen timeframe. By so doing I can favor trading off of those S/R levels with partiality to a properly defined trend based on the changing prevailing cycle strengths as measured by volatility. As my equity curve shows over the last several hard trending days, a scalper that tries to roll with a short term oscillations during powerful one way trends needs to adjust to larger timeframe analysis or will be badly burned. I have a solution, but first the problem needs to be clearly defined.
The Timeframe Problem:
The problem with S/R levels is that they are drawn on static, chart-defined dataseries of one bar type or another. However, the bar size on the chart may or may not be in sync with the dominant cycle sizes of the hour. By that I mean when markets are more volatile and are making larger excursion one should be trading trends based on larger timeframes. If the period parameter for the moving average you follow to determine the trend is too small you will lose money when your too-short-term MA indicates a reversal in trend, while the prevailing cycle and its related timeframe overrule the trend your MA tells you exists. Conversely when markets consolidate into smaller bars trapped in a range, an MA with too high a static period will not detect the tradeable changes in trend occurring on the smaller, then dominant, timeframe.
The "range2period"Solution:
This is not original thinking but rather my restatement of a creative insight put forth by Lawrence Chan, a more astute trader by far than I. If trend were determined by the slope of a moving average but the period of the moving average dynamically changed based on current volatility ranges then in higher volatility price action the MA trend would be determined based on a longer MA period and not whipsaw so much. In order to dynamically and automatically adjust period for moving averages or for any other indicators that use a period, a trader can use indicator on indicator (MA on range2period) to effectively dynamically adjust period on the fly to match range. Google range2period to learn more about this concept. Range2period is an indicator in NeoTicker that outputs a period size based on volatility. An MA set to a period defined by a range2period indicator to which it is attached and whose changing output resets the correct period, automatically adjusts its period to adapt to changing volatility so the slope and direction of the MA always reflects the dominant timeframe, regardless of the chart's timeframe. If anyone has or knows of a Ninjatrader indicator that does range2period I'm interested in not having to recreate it in Ninjatrader.
Having looked at "love [of trading] from both sides now" I intend to use both sides. Yet knowing how too true it is that "it's love's illusions I recall" I'm working to escape both the programmatic illusions of careless or optimized backtesting, and the discetionary illusions induced through the vicious cycles of counterproductive emotions to which discretionary trading is so vulnerable, requiring self-imposed restrictions based on known odds. If you would like to code range2period in Ninjatrader let me know. I can help.
(BTW, "love" of trading is at best a sacrifice of time and life for the good we can return to others with the profits we recapture from banksters in our Robin Hood hunting expeditions in their king's illegitimately claimed financial forest, otherwise its a selfish indulgence in gaming to serve pride and excess, which is indeed and fatally "love's illusion.")
Prosper with Disciplined Cheer!
(thanks for the thanks)
2-21-17 P/L ($2445.8)
Cum P/L ($143131.82)
Exceeded TST max drawdown so reset again.
2-22-17 P/L ($1209.64) Long $51.46, Short $(1261.10)
Cum P/L
You may have noticed that the progression of the ATM refinement for catching scalps while carrying the remaining contracts to ride whichever trend prevails has been to recognize from analysis that initial ES stops should be not 5 but 7 or 8 ticks with or against the trend respectively, and that profit taking should be more aggressive taking more contracts off at 1st target and exiting at BE if initial move goes 5 ticks against trade.
Yesterday failed to capture short profits. Accordingly today we have made the ATM's even more protective, taking 3 tick profit on 4 of 5 contracts and trailing 1 if in a countertrend trade and 5 ticks on 3 if with the trend, trailing 2 of 5. I like to push envelopes when testing then progressively return toward the center of the envelope until profits are stabilized, otherwise one could not find the most profitable strategies, which are always outside the norms of conventional market wisdom such as reward should be twice the risk. Algo's know that's how retail traders think. I created a successful algo that took twice the risk of the reward, which is more appropriate to countertrend trading which has the inevitable blow by around reversal points.
One of the problems I've had is not resetting the strategy before starting a new trade set, which causes the new stop and target to default to the existing levels from the prev trade. To prevent that I've set the properties of the DOM to Display selected ATM strategy only, the used the position tab in the control panel to monitor overall position from the combined trailing and current positions.
Prosper with Disciplined Cheer!
(thanks for the thanks on the challenge page)
In reviewing my experiment at trading a simulated $150 TST Combine, with 5 contract entries attempting to scalp some profit and always leave a riding contract of 1 or 2 to catch a trend, thus far the combination of ATM's I've invented would appear to be profitable were the buys all sells and the sells all buys. I've burned through 3 hypothetical Combines incurring a $16,410.86 loss. Attached is the trade performance summary as of 12:24 Friday. As described in one of my first few posts, and despite my awareness of my "cognitive" short bias, 2/3 of my trades were nevertheless shorts yet profitable shorts were only 37% profitable with total net short losses of 16,868.82, while longs were net profitable at $457.96, 56% of which were profitable.
The markets created the perfect month to test this method of trading because only 2 of 18 trading days in Feb. closed below its 9:30 open while the ES staged an explosive rally from the low of 2264.50 on Feb 2 to its current high yesterday of 2367.50, or 103 points. The core concept was to attempt to create ATM's that were net profitable scalping in a strong trending market.
The beauty of this test is that not a real penny was lost. Could it be that among the illusions of countertrend trading historic Support and Resistance levels, we the the successful extreme highs and lows from which reversions occur, but not the series of continuations through less extreme highs and lows. Imagine doing just the opposite of what I've demonstrated by instead going long when hitting resistance and going short when hitting support. Do bankster algo's know how to work those expectations of reversal and stop out the typical 2 points stops of retail traders>
Admittedly taking a few ticks profit when commissions burn up 1/3 of a tick both in entering and existing creates a high overhead that has to be overcome with precise trade entry and management.
I'm going to keep trying to perfect this strategy but not with real money until I prove it works. To those of you who attempt reversionary scalping I would say, filters have to be found to prevent some entries to develop a higher level of specificity (choosing specifically the winning trades), and if it goes against you get outfast on a return to near your entry. Even waiting at break even did not work for me. Better to give up a tick and avoid a greater loss.
There is a risk in my conclusions that Feb was an exceptionally strong month and that my strategy would have been more successful had Feb not been so exceptional. One vendor I follow says this weekly pattern of late has never occurred historically, unless you include searches for the inverted pattern of this rally. Nevertheless, I'm glad to have attempted this in the worst of circumstances as I am interested in building a trading system so robust that it can survive and profit in the most harsh market environment.
I hope you have found value in this systematic effort. I still believe that elite traders are able to trade very successfully without waiting for the immediate price action to go there way, but it takes an intimate knowledge of how price acts around S/R levels because I suspect MIT grads have algos to shake out weak knees at the best turning points in the markets.
I may do some serious development before continuing such intense journal commentary but will keep you posted as I learn more.
Hope the journal contest advanced your understanding as it has mine. All the best and thanks for your thanks. IF you have insights I"m interested!
I think they may count the thanks on the journal contest page so if you haven't thanked me there please do.
As the next step in analyzing how to exploit the exploiters whose algos wiggle the wealth out of retail traders, I'm going to use Bloodhound to backtest my ATM algos to see to what extent my distorted psychological "cognitive bias" has intereferred with the equity curve that mechanical entries would produce. This will be a good comparision of purely mechanical, backtested systems versus discretionary entries managed mechanically by Ninjatrader's ATM (automated trade management). As you see in my prev post I burned through 3 hypothetical $150k Combines during the course of this experiment in February incurring a $16,410.86 loss. Has anyone beat that during the contest? If so my congratulations if you did it while systematically pushing the envelope of possibilities in order to explore the feasible limits of unusual strategies where other's rarely go in search of profits.
I regard these experimental losses as instructive revelations on the way to refining a system on the way toward enjoying sizeable stable victories. I'm not interested in being a retail trader that makes a few hundred a day given the sacrifices necessary in becoming a successful competitor in the world's most competitive environment, but rather have my sites set on making thousands a day, hence my interest in the $150k Combine and mastering an aggressive way of successfully trading it. In a sense I'm spying on short term algo's to scalp and on longer term players for the trend.
If I win a $150k combine I will keep you posted on how well the ATM's are working and keep them updated for you here.
Please don't limit your thanks to this thread. If you find value in my continuing experiment in how to aggressively beat the banksters by low risk scalping into trends please give your thanks HERE, which is where they count the number of thanks for contest winners.
