Intraday reversal strategy (CL) - comments welcome

[austinp]

You are describing a very common trap-spring price pattern I've traded for a few years now... it's one core component of how markets move based on computer programs clearing stops one direction and then v-turn reversing abruptly the opposite.

The overall setup is what we label "key-reverse" sequences. You will not be able to automate it based on system language alone, because the 1-2-3 structure of the general pattern is too varied. But once you define the parameters on a discretionary basis, it is money $$$

 

[austinp]

Quote from Zr1Trader:

IMO the best "so called edges" can't be automated.... yet. My .02

100% correct. There used to be a common fallacy that any discretionary logic could be written in computerized systematic language. That's a basic assumption of computer geeks who don't know trading.

I have enough systems writing experience to know that much if not most of what the human mind can decipher is too complex in nuances for translation to systems language. If indeed that weren't the case, no veteran discretionary traders would exist... we'd all be fully automated and nowhere near the screens for a living

 

[HurricaneUS]

oh what the hell....I'll throw you a bone...no pun intended...

dump the mechanical systems because they will always fail....you don't have the programming experience or resources necessary for a reliable automated system


Why do mechanical systems ultimately fail for retail traders no matter how much effort they put into them?

Because they will always lack context. You can't really program context into a mechanical system (unless maybe you have deep pockets like goldman sachs).

What is context?
I'll leave you to ponder this question...but i'll get you at least started with some pieces....momentum....intuition......etc...etc..etc

 

[tobbe]

Quote from austinp:

You will not be able to automate it based on system language alone, because the 1-2-3 structure of the general pattern is too varied. But once you define the parameters on a discretionary basis, it is money $$$

Is interesting how you say you can define the parameters on a discretionary basis but not in a programming language. Care to elaborate? What you say is implying that the system can't be taught to someone else either. If you can teach it, you should be able to program it, unless of course there is some hidden way of information transfer between people that would elude a computer algorithm.

 

[Zr1Trader]

Quote from tobbe:

If you can teach it, you should be able to program it,.

False,

Read WRBtraders hammer thread and program that come back with your report thanks! Too many moving parts , context, news, supply /demand zones on various timeframes. Entry , exit. Way too complex. Not saying one day it can't be done, but not today.

WRB is a vet and I'm sure he can explain better if you want to contact him about trying to automate some of his methods. Many have tried and failed miserably yet he trades with only a few loss days a year.

 

[dom993]

Quote from dom993:

I am quite concerned by the amount & the duration of the recent drawdown period. I did MC sim using the trade distribution up to the start of the drawdown, for a number of trades corresponding to what happened since (108 trades) ... the 7,000 drawdown is 4 std-dev away from the mean, and the drawdown duration itself (96 trades) is 8 std-dev away from the mean. But I really lack experience to interpret these figures, and I would appreciate your comments/feedback.

To be clear, I am seeking comments/feedback on the recent drawdown characteristics, if possible comparisons with other traders' systems from that point of view.

I don't really want anyone to waste his/her precious time explaining why they think what I am doing is impossible (I read "Inside the black box", and duly noted the lessons exposed p.129).

 

[bwolinsky]

Quote from dom993:

I identified about a year ago a reversal pattern on CL, which seemed interesting enough for me to start developing an automated system for it.

I won't disclose the specifics of the strategy here, but the idea is to look for for a trend reversal manifesting itself by price making a HH directly off a LL (reversal off a down-trend) - the entry long is in the 1st pullback following that HH (reverse everything for a short after a reversal off an up-trend). I use 2 targets, target-1 is "scalping" and has an average reward of about 85% the average risk, target-2 is "runner" and has an average reward of ~250% the average risk.

This reversal pattern does show-up in all timeframes, and I chose to go as "low" as possible in order to maximize the number of opportunities. I decided to work off a 200-volume chart (CL), which initially seemed a good comprise (I would have used a 100-volume chart, if it wasn't for a limitation of my trading platform which limits the number of bars in a chart to 65,000 - CL can actually go higher than that on a single month on a 100-vol chart).

The real key to this system is pivots identification. I developed a 1st version that I took live mid-September 2011, only to stop trading it early November as the system experienced a severe loss of performance (which really had started early August). That loss of performance did continue into the new year, until around the start of active trading of CLH12 (late January). Meanwhile, I was frantically scrutinizing the charts, looking for clues. But really, I couldn't find any "reason" for that loss of performance, aside from using "bad" pivots.

I then reworked the pivots identification mechanism ... that yielded a significant improvement, manifesting itself not only in the last few months. But it remains that the last quarter of 2011 had the most severe drawdown period of the backtesting history (I currently use tick data since October 2009 for this backtesting, this is 28 months, total number of trades in backtesting is 554).

Attached is the P&L curve out of the latest backtesting. Interestingly, the system recovery -since the beginning of CLH12- is only forward testing.

I am quite concerned by the amount & the duration of the recent drawdown period. I did MC sim using the trade distribution up to the start of the drawdown, for a number of trades corresponding to what happened since (108 trades) ... the 7,000 drawdown is 4 std-dev away from the mean, and the drawdown duration itself (96 trades) is 8 std-dev away from the mean. But I really lack experience to interpret these figures, and I would appreciate your comments/feedback.


To answer some questions I know will be asked:
- the system is all about price action ... it uses no indicator whatsoever, only H/L pivots & projection techniques from these
- system accounts for $5 comms per contract / trade
- system assumes 1-tick slippage on all stop exits
- system entries & targets are LMT, assumed filled only when price trades through the limit
- system P/F = 1.95 on entire 28 months ; it is down from 2.25 on the 1st 24 months of the backtesting period
- system P/F for target-1 is 1.65 on entire 28 months
- system P/F for target-2 is 2.00 on entire 28 months

The 1-tick slippage should be applied in 2 tick increments to all backtesting. It is a mistake to only include 1 tick slippage when they are blanket values to use so that you don't get overtraded results.

Use convert to market orders, and also add 2 ticks of slippage to it.

I have no way of deciphering drawdown from these values, pf for target 1 is obviously a lesser and more frequent target than the others. The system P/F and P/F for target 2 appear to include overnight trades.

Limit order execution guarantees against any validity in this backtest.

 

[bone]

Quote from dom993:

To be clear, I am seeking comments/feedback on the recent drawdown characteristics, if possible comparisons with other traders' systems from that point of view.

I don't really want anyone to waste his/her precious time explaining why they think what I am doing is impossible (I read "Inside the black box", and duly noted the lessons exposed p.129).

Well, the lack of relevant feedback should be very informative to you if you are not blind to the message about your strategy.

Alternatively, pleased be advised that Kase Pivots, Jan Arps, DiNapoli, Andrews Pitchforks, or any the generic swing reversal studies that I am aware of on LIVE intraday CL Nymex data is a shit show.

Linda B. Ratschke uses combination studies for specific timeframes to make it work - provided she believes the market is not trending ( that's a trick ). It takes alot of experience to make it work, because you will get your ass handed to you and give alot/most/all of it back when the market finds a new trading range and value/control area.

My advice would be to use it in conjunction with MP, and not keep going to the well when the market is persistently finding fresh highs or lows.

And you'll still get sodomized.

But go ahead and try.

 

[dom993]

Quote from bwolinsky:

The 1-tick slippage should be applied in 2 tick increments to all backtesting. It is a mistake to only include 1 tick slippage when they are blanket values to use so that you don't get overtraded results.

How about you proof-read before you post?

On another note, stop slippage in backtesting must be reflective of stop slippage in real-time trading. I do trade CL live enough to know that 1-tick slippage in backtesting does reflect the average stop slippage good enough.

Quote from bwolinsky:

Limit order execution guarantees against any validity in this backtest.

This statement guarantees against any validity in your comment.

But in case you had serious concerns about this, note that my system does handle fills at LMT price (when price doesn't get through the limit, ie. lucky fills) in such a way the live performance can only be better than than the official backtesting. Of course, those lucky fills are excluded from the backtesting results (but they are systematically simulated, to verify the trade management in those cases).

 

[austinp]

Quote from tobbe:

Is interesting how you say you can define the parameters on a discretionary basis but not in a programming language. Care to elaborate? What you say is implying that the system can't be taught to someone else either. If you can teach it, you should be able to program it, unless of course there is some hidden way of information transfer between people that would elude a computer algorithm.

what cannot be written by anyone who frequents a public message board is the complex nuances of management for the various scenarios involved with this equation.

the OP has identified a key price pattern behavior in the market. he is trying to curve-fit (pigeonhole) results from a pullback situation following reversal and first peak exhaustion = pause. The challenge there is price can do a vast number of things between peak, pullback and resumption in the particular trade sequences which are positive outcomes.

I have four main price structure patterns in any symbol or timeframe which I trade, and that is one of them. But it's not as simple as if A does B = take action at C logic. The human mind can watch what unfolds, how it unfolds and most importantly, how it is filtered by other key price measurement tools.

The OP has maybe 20% of the parts in place. The remaining 80% cannot be programmed, because of the myriad variables involved. But it's cake to trade discretionary, once one gets the hang of 100% parts for the complete picture.