What works, and what doesn't

someone with 20/80 can be more profitable than who does 90/10. those rates are meaningless

You misunderstand.... I'm talking about better than 50:50 win/lose pip ratio.

Sure, one can lose 66% of trades and still take a profit:-)...

Don't fall for the popular myths from the trading books!
Strategies with low win rate tend to experience deep and long periods of drawdown and the equity curves look like roller coaster with huge ups and downs. Because of that, you have no choice but to trade small size in order to prevent total loss of capital.(you constantly have to reduce size to escape "Gambler's ruin" https://en.wikipedia.org/wiki/Gambler%27s_ruin )
It's always better to have strategy with higher win rate (over 50%) because with smoother equity curve and minimum drawdown you can use more leverage and ultimately make more money in shorter period of time. It's much more efficient, and easier to grow the capital exponentially. This things have been proven by mathematics and statistics long time ago......

Alphaomega, I agree:-)

The point in my original post is really that the 'Edge' is provided by the traders own instinct and experience. Most of the indicators and methods merely provide comfort and peace of mind. Mathematically most trading systems, when traded mechanically, with no insight from the trader, will lose money in the long term due to trading costs.

yes, it is called art of taking losses, it is great and one of the most precious skills trade can master. the other side of the story is holding on to winning positions which will cover all the losses and generate high profit. running after delusional things like high win rate eventually cause disaster, nurtures big egos and feeds the urge to be right all the time. i would rather be profitable than being right.

Yeah, that's 100% true. The only thing working all the time is constant adaptation.

The technical stuff and statistics is just a flavour of trading that appeals to the more analytical type of people. In the end its just a set of rules that buys when price is going up and sells when price is going down.

If you have a pair of eyes and absolute inner calm you can just do this without the statistical stuff. Though I will always recommend mechanical systems because emotions always find their way to creep in and instill bias.

On the topic of win/loss ratio; maximizing returns normally involves losing more trades than you win.

This is where I have a slight difference of opinion. Human intervention when expressed as 'discetionary trading' is frequently the reason for 'overtrading'. I appreciate that perhaps there are a few traders with the commensurate skills to achieve a positive edge in the long term, yet there will never be any solid (auditable) way to test this opinion as the definition of discretionary trading itself means that the technique itself is not standardised and is applied at the whim of the trader.

You can certainly teach some skills to others, but there is no guarantee you can transfer these skills with sufficient fidelity to exactly reproduce results. There is therefore the necessity for 'belief' in discretionary trading and this is where humans run into so much trouble as lying within a belief system is human bias that makes trading such a subjective 'art form'.

I actually believe that limited human intervention (as we inevitably make wrong decisions when dealing with highly unpredictable systems) with freedom for price to move within as wide as possible broad directional constraints is the only way to achieve a long term sustainable edge in this market. At the end of the day is your performance a result of good profitable trades, or a result from avoiding bad possible trades? So much attention is placed on the former with only scant attention placed on the latter and I actually feel with enough testing now under my belt that 'less trading is more' by virtue of the second principle. Think of a trend trader who is dependent on an income who is faced with the difficult task of trading a single instrument. How many times does price move sufficiently for them to obtain an income. Overtrading kills a trend trader and that is why they need to diversify. Each instrument themselves may only deliver once or twice a year. This puts psychological pressure on the discertionary trader to attempt to trade a trend as soon as the market twitches. With a diversified portfolio you can be so much more relaxed and only take a position when it 'literally' screams out at you....not when a butterfly flaps it's wings. Is it human bias that is forcing you to trade or the sounds of the market?

With this opinion in mind then the more robust strategies are therefore the simpler strategies that maximise degrees of freedom by having limited 'constraint' variables and can be applied to any liquid trading instrument. Having this opinion therefore makes you less interested in those 'short term' techniques that deliver huge unsustainable returns and simply focus on those very simple measures that can be quantifiably demonstrated to work across market conditions without blowing up an account and deliver modest but sustainable returns.

The aim I am seeking to achieve with the deployment of simple durable strategies as a part of my system development is to recognise that to defeat randomness, a baseline strategy needs to simply achieve a return that compensates the frictional costs of trading with a small profit margin. The primary intent of the baseline strategy is to deal with market risk and turn the house odds in your favour. It is only through full automation that you can actually quantify if this is being achieved. In more detail what I am referring to are simple strategies that simply attempt to capture 'levy flight' or statistical price bias in otherwise unpredictable price movement which hopefully is sufficient to overcome the frictional costs of trading (spread, negative swap, slippage). I have yet to find tools that can definately interrogate a price series and with 100% confidence seperate that which is random movement versus non-random price bias, but there are statistical tools that have a good crack at it in statistical terms with of course an associated confidence interval. It is therefore those tools that I use to develop 'simple' robust strategies.

Once you achieve this aim in the underlying technique, then that's when money management and diversification can be harnessed to accelerate performance through additional phases of system development. By limiting the 'power' of the actual underlying technique you are naturally reducing risk inherent in the technique itself, it is a bit like using a 'slow cooker' rather than a 'pressure cooker' to cook your meat. You are not trying to coerce (force) the market to your system and simply letting the market do it's thing with an end intent of simply achieving above 50% house odds to get into a position of strength. Once achieved you can then turn up the heat using other safer methods of performance attribution such as diversification or the tweaking of variables to respond to current market conditions without altering the underlying small edge of the system.

Most traders focus on the technique itself as the primary performance measure whereas to me the prime purpose of the underlying technique is using principles of risk management (to survive in all market conditions through managing unpredictability) as it's basis and only adding sufficient bias to turn the house odds in your favour. Like a casino, you then use other methods to supplement performance that does not interfere with this underlying principle.

For example:
Stage 1 - Develop baseline strategy (using a selection drawn from widely available simple strategies) tested across an equal weighted (using position sizing) array of low-correlated instruments drawn from highly liquid asset classes (forex, commodities, treasuries, indices etc.) that is demonstrated say over the past (10 years say or XXXX trades of sufficiently high sample size) to be reflective of the performance of each asset class itself and can achieve a net return of say 10% PA (which includes costs of trading) and an associated worst drawdown of say 30%. Test this on out of sample data to ensure it is not 'curve fitted'.

Stage 2 - Develop additional systems/ or modify portfolio allocations that seek to fill the drawdown gap of Stage 1 and do not compromise the overall return of the strategy (eg. flatten the equity curve) to achieve a result over the same period of say 10% CAGR and max drawdown of say 15-20%. Once again test the conclusions against out of sample data.

Stage 3 - Play with the equity curve using different instrument allocations to understand the strengths/weaknesses of the strategy and identify valid quantifiable limits of the system that can be used as a basis to determine if the system is over-performing or is 'broke' according to you personal risk tolerance capacity. Also play with system variables as sensitivity tests and undertake Monte Carlo simulations. By the end of Stage 3 you should know your system inside out and can in a very confident way conclude whether it is robust or not.

Stage 4 - Once fine tuned, then use Position Sizing methods as a basis to multiply out baseline statistics. For example, use Van Tharpe's position sizing methods to optimise performance efficiency by system 'scaling'.

Stage 5 - Now that you have a robust system that has been comprehensively tested across a range of conditions and appropriately scaled look at simple ways to further accelerate performance without compromising the 'risk management' aspects of the underlying system. For example look at the impacts of account compounding on performance (re-investing profits). This keeps 'relative' drawdowns at the same level but accelerates performance. Eg. Perfomance now 15% PA and drawdown of say 20% over back-test period.

Stage 6 - Then further accelerate performance in an adaptive manner when deciding to take the strategy live. Optimise the system when going live for current market conditions and at defined intervals use adaptive walk-forward testing and retest these assumptions against your back-tested benchmark results enabling you to 'turbo boost performance'. The intention of this phase is to accelerate performance beyond your back-tested benchmark of say 15% PA without impacting drawdown and this process needs to be actively managed. Hopefully the result can perhaps add another few percent to overall performance.

Note that the above steps does not relate to any particular technique and can be applied to 'any technique' that is sufficiently simple in harvesting the 'known knowns' of market uncertainty in a complex system. The technique applied in this interpretation is of far lesser importance than the actual process of testing the validity of the technique in a quantifiable manner. What these processes do is force you away from subjective bias (a traders worst enemy IMO) and into the very real world of 'reality'. For those who think this is philosophical....well, you are then perhaps missing the whole point of reality.

It is only by playing the same game as the banking system, the insurance industry or the casino that you can survive in this market for the long term. You can transfer this technique with fidelity provided they understand the game you are playing and apply it correctly. Namely to beat the probabilistic market to death with statistics which is never a certainty, only one backed up with statistical validity. The same can be said for the game of life or any game that applies in a complex system where certainty is never an enduring feature.

In my humble view, what works is having a reference point for your trades. Each bar that passes you know what to do on the next bar. Buy, sell, SAR, hold, or exit the market. I think as the day goes along, traders get lost. What works is having a map to tell you in advance what to do. You should not be thinking that much during the trading session.

In my opinion (and experience) trading is more art than science. It's performance oriented discipline, where the inner game is the most important thing. The quest for certainty and proof is very dangerous thing in trading! Trying to beat the market only with statistics and equations often leads to dead end. There are brilliant mathematicians and many smart people who can't make a dime in the markets and at the same time there are people who doesn't appear to be smart in any way, yet they make millions (by discretionary trading). So there is something more then just numbers. When the shit hits the fan, all statistical proof becomes useless. Then only the best traders can evaluate the situation correctly. Don't underestimate the human brain! It is still the most awesome computer/machine. We are very good at pattern recognition, creativity, learning new things quickly and coming up with solutions to the problems on the fly. When you spend enough time in this game (or any game) the intuition and the experience just starts to speaks to you. It's hard to explain. You just feel the flow and 9 out of 10 times you see the opportunity before it happens and you can clearly sense the danger when it's coming. The big problem is reading these emotions correctly and then acting fast based on them. It's very hard to observe yourself in the heat of the battle and to be objective about your own emotions. So that's why after we master all the technical stuff ( the tools) we have to work on the inner game. There is nothing else out there (externally) The "holy grail" is already in our heads by default. We just don't know how to use it. Without deliberate practice and desire to develop in this area we only use small fraction of the potential.

No probs. I can appreciate your viewpoint and recognise that there 'probably' are some excellent discretionary traders out there. But I would tend to count them on my hand rather than assume it applies generally.


This is actually where I disagree. Where the 'shit hits the fan' relates to 'alpha crisis events' which to those at least seeking some non-random behaviour to gain and edge, can capitalise on these events where market behaviour in crisis generates predictable behaviour. A number of heavy duty quantitative studies of market behaviour that are publicly available substantiate the principle that the best chance for non-random market behaviour lies in crisis events that are defined as the 'fat tails' of a gaussian distribution. The problem is in trying to survive long enough in this market to participate in these events. Furthermore there is definitive proof in the form of statistics that non-random drift exists in extended price series that cannot be visually described as distrinct from non-predictable behaviour, but simply serves to extend price directionality in price action behaviour. You can trade the statistics by placing your entries and stops in zones at the extremal locations of randomness (noise) in an associated price bias direction and hence slightly tilt the house odds in your favour not by virtue of where price lies today, but in a knowledge of where it rarely is (the zones surrounding price movement) and where it is likely to be tomorrow. This means you need to focus on the constraints within which price moves as your guide to trade and stop placement positioning and forecasting.


I certainly don't underestimate that amazing piece of grey matter, but I recognise it's ability to blind judgement and critical thinking.

I can appreciate where you are coming from in this but just differ in opinion as I believe a process can make a good trader. It is a nurture versus nature argument that probably best describes my argument. Of course there is skill in system development, but this can be taught just as in the same why we can fly our planes in the sky with a degree of certainty without attempting to flap the arms.

Todays trade.. weak run up in euro... So I was looking for short.
Took profit a bit early due to reaching my daily goal.. wish I had held it a bit longer..
When seeing a quick run up... you should be asking .. .where do I short!

Drop was weak and need to be filled right? So where to go long now?

Currently in a trade on the Cad.

Lookin to TP around 3054 -3056 if i see some resistance there, otherwise gonna tp around 3048.


Demo

And there it goes as i post

Just watch the weak area's..
"Strong" moves Leaves a weak market structure..
When market structure is weak there is opportunity.
Next time you see a quick pop.... think how weak the market structure is.

How to take advantage of those opportunities?

Entering is more of an art then a science...

Might be looking for a lower high... might be a key level like NFP high..
Might be time of day where bears seem to always go home and stop pushing..
Might be a hammer or shooting star candle formation.

Thanks

Could you expand a little more when you have some time? maybe with some charts.

What is the best time to look for fake moves to fade and when expect a real continuation move?