The consensus on this thread so far seems to be that trading is 'hard' but not impossible to derive a consistent reliable secondary/primary income from ??
Following on from my post on this thread #34 let us examine how you might go about dveloping your edge and what you need to know about it.
Let's say you are drawn towards tech analysis rather than fundie or other types of analysis and this is how you wish to derive an edge.
Firstly you accept what an edge is:
A Trading Edge is a repeating circumstance or set of circumstances (ie a set-up or combination of set-ups) that suggest a greater probability of one thing happening over another, based on historical precedent. Additionally and/or alternatively; if the edge does not suggest a greater probability of one thing happening over another, then based on historical precedent, it suggests that more gain is available when it succeeds than when it fails, sufficient to realise a net gain over any given sample.
So then you need to develop such a tech edge. Let's say for the sake of argument that on a trawl back through the charts you have noticed that when a Pinbar or other classic (pre-determined) reversal candle forms outside or straddling the 20bol supported by regular divergence on the stochastic oscillator [measured between the last swing on that oscillator outside the 20 or 80 extreme level] - that price seems mostly to snap back inside the bollinger and there is pip gain to be had. (By the way I am not saying this is an edge or indeed is not an edge - I have no idea- it is just a simplistic example to illustrate the point.)
An example of this is below at point 1,2 and 3
So next you need to gather a large back sample of this set-up and how it performed
on the instruments and t/f's you intend to trade from And from this data you need to know the following
Firstly you need to observe the typical to maximum risk:reward ratio achieved, ie would it have been better aiming for 1:1, 1:2 etc, and when would it have been optimum to move stop in to minimise the risk without it affecting the chance of the trade running on to achieve optimum R:R over the back sample.
Then,
1. Sample size (no of set-ups observed)
2. Total winning trades at the R;R, Total losing trade sand total B/e trades given the trade managemnet rules.
3. Know it's overall strike rate (winning trades as a % of total trades, at your minimum r:r ratios.)
4. Know it's typical/maximum distributions of wins/losses
5. Know it's typical / maximum consecutive run of losing/winning set-ups
6. What was the largest drawdown experienced across the sample
7. Over the sample size know the probability of a consecutive losing run of x,y and z , x being the typical, y being the maximum experienced across the back sample, z being say double or treble the maximum experienced.
Knowing all this information will help you to optimise your money, risk and trade management sensibly to this
potential edge.
Let's assume the back test stacks up and an edge may be there. Next thing to do is to forward test it in the live market on the same instrument and time frames used in the back sample and to see whether you can follow the rules of entry, trade management and exit you have determined from the stats gathered over the back sample. Ie Can you obtain similar results over the forward test of the same size or larger than that of the back sample ? Ie can you, as the excellent Mark Douglas says -close the 'Profit Gap' that exists between having even a potyential or proven edge and actually being able to tyrade it profitably yourself. If you cannot or more of things may have occurred
a. The potential edge just didn't satand up in the forward test - ie the back sample was not sufficiently representative.
b. You did not stick to the entry, trade management and exit rules of the potential edge and this has skewed the forward test data.
If the forward test does resemble the back test sample then repeat the forward test again with same sample size, and if this again stands up, repeat it again...and if this stands up - you may just have an edge and feel confident enough to begin trading it with real money.
G/L
). There I was, patting myself on the back when the entry "worked" but finding that many bags of gold had turned to purses of copper by the time I closed the trades or, even worse, to bills from my broker.
