What analysis do you use for trading?

[nunthewiser]

risk:reward is of little use to me. My risk on any position is around 20%. But in reality my avg loss is around 5%. The initial risk on any trade is still 20%.

I .

WTF?

 

[tech/a]

Why not move the stop above break even and get some negative R:R action happening?

Yes you could it was only one example.

Surely the R:R of a trade must use the initial risk, not the decided risk when the trade is closed out?

Yes of course and that will normally be a % of capital.

The Risk on a trade for me is the maximum amount I'm prepared to lose when i take the trade. The reward is the result of the trade. But I do not trade base on R:R, and like Mr J believe that the horse comes before the cart, and changes to the way we trade effect results, in turn effecting R:R. The R:R only changes because you're making a fundamental change to the trade or the way you trade (eg tightening stop).

Yes true of most traders.

risk:reward is of little use to me. My risk on any position is around 20%. But in reality my avg loss is around 5%. The initial risk on any trade is still 20%.

I would re-phrase "Initial R/R is of little importance to me".

I find a profit:loss ratio that incorperates the win rate to be more useful. Risk constantly changes throughout a trade, but the end results are fixed and provide a much more stable benchmark.

Agreed.

 

[Mr J]

Surely the R:R of a trade must use the initial risk, not the decided risk when the trade is closed out?

I don't think so. We're using our average R:R to determine performance, and being able to minimise our risk to be less than our original risk is part of that. If my average loss is less than my original risk, then using the original risk doesn't reflect the performance of the trades. The original risk is just a limit for maximim loss. We could use the same idea for profit and implement a 1000% target profit. Should I start using that for R:R because it's my maximum potential profit? If not, why should I use my maximum potential loss?

My risk on any position is around 20%. But in reality my avg loss is around 5%. The initial risk on any trade is still 20%

Initial risk yes, but using the 20% figure overlooks your ability to keep losses to far smaller amounts. If we're judging performance, that's information we'd want to include.

 

[MRC & Co]

............

LOL!!!!

 

[tech/a]

I know what your doing Synergy and agree.

 

[motorway]

http://www.fullermoney.com/content/...onsFromInconsistentForecastingSkillsNov07.pdf


Nice overview of many various aspects of Stops.

SURELY IS an IMPORTANT TOPIC
esp since MARKETS
are not random
But hard to PREDICT

They TREND but EVOLVE...

Motorway

 

[skyQuake]

http://www.fullermoney.com/content/...onsFromInconsistentForecastingSkillsNov07.pdf


Nice overview of many various aspects of Stops.

SURELY IS an IMPORTANT TOPIC
esp since MARKETS
are not random
But hard to PREDICT

They TREND but EVOLVE...

Motorway

Cheers for the great link motorway. Now to look up 'Consensus Economics of the UK' and fade their euro plays!

 

[Mr J]

Talking about stops, one analysis that seems to be underused (maybe not, but I haven't seen anyone talk about it) is finding the optimal stop size. A lot of people may use the previous swing high/low, a moving average, previous support or resistance etc, but I think there's a better way to go about it by analysing how much of our stop our previous trades have used.

If we find our profitable trades rarely go 30 ticks against us initially, why would we use a 50 tick stop? If we dropped it to 30 we'd boost our return by 66% (minus the odd profitable trade that did go past -30). Big boost to be had there.

 

[motorway]

Talking about stops, one analysis that seems to be underused (maybe not, but I haven't seen anyone talk about it) is finding the optimal stop size. A lot of people may use the previous swing high/low, a moving average, previous support or resistance etc, but I think there's a better way to go about it by analysing how much of our stop our previous trades have used.

If we find our profitable trades rarely go 30 ticks against us initially, why would we use a 50 tick stop? If we dropped it to 30 we'd boost our return by 66% (minus the odd profitable trade that did go past -30). Big boost to be had there.

What sort of method do we have when we we let market events guide us ?
===> We use a contingent Buy order
an initial Stop
We move to Break even , we adjust to the unfolding market events..
We trail the stop... We use contingent Sell orders etc

We then enter on other market signals.. AS THEY OCCUR .

What sort of method is this ?

It is using a digitalized method ( think in terms of MOVES of % )
an Event driven method
of
Non periodic cycles ( The signals have their INTRINSIC TIME )
THEY are Market Time driven.

A NON LINEAR method .. A POWER LAW DRIVEN METHOD

what TECH/A would call a METHOD that is in harmony of WHY we make $$$$

It's old name is POINT AND FIGURE METHOD

In fact that paper on STOP LOSSES
is just another way of describing P&F

IF YOU CAN HOLD IT ALL IN YOUR HEAD you don't need the charts

BUT if you want to filter amongst the best opportunities
and to have revealed before your eyes ( with mathematical PRECISION ) the
FACTORS you are considering.. ( range etc )

You need to use a method of representation that is
what TALEB would call EMPIRICAL & SKEPTICAL , non linear and POWER LAW DRIVEN... (In his first book he gives a good description of P&F method )

FOOD FOR THOUGHT



TO THE extent you use a CONTINGENT ORDER MODEL ( stop loss )
you are overlaying A P&F METHODOLOGY on

Your Linear Time framed ( illusory ? ) etc REPRESENTATIONS.


Rule number 1) NEVER DOUBT MARKET EVENTS


They make the $$ and take therm AWAY...



Motorway

 

[xinyu09]

When we finally solves all chess puzzles then the next step is analysis a totally random market and make it rational! No more emotional buys/sells, we are here to make a profit and we have workout strategies that will work forever. Then who's going to invest to the prefect world knowning that someone else will always win?! It's like the prefect sword and prefect shield dilemma

Anyway, I believe all strategies that make a profit will make a profit and the odd to make a profit is a lot higher than 50% (if not 100%) for a good trade!

Hisitory wise, a good company always make a profit and I guess the odd for investing in such company is 100% profitable in the long run.

Just my opinion, DYOR

 

[GumbyLearner]

When we finally solves all chess puzzles then the next step is analysis a totally random market and make it rational! No more emotional buys/sells, we are here to make a profit and we have workout strategies that will work forever.

Just my opinion, DYOR

Like you xinyu I'm here to make a profit also but I'm also not willing to give away my rights to question the principles of a Westminster Democratic System that currently exist in Australia. So come get me! :

 

[Mr J]

Motorway, sorry but I have no idea what you just said, and it doesn't seem to be anything about analysing the size of the initial stop?

Anyway, I believe all strategies that make a profit will make a profit and the odd to make a profit is a lot higher than 50% (if not 100%) for a good trade!

All strategies make a profit for someone, even if it's only the broker.

Hisitory wise, a good company always make a profit and I guess the odd for investing in such company is 100% profitable in the long run.

Even if it were true and that easy, it doesn't seem to have made a lot people much money.

 

[motorway]

Mr J

Just food for thought

Instead of over laying a time based chart
with a contingent event driven method

Be that STOPS
contingent BUY sells ( SIGNALS etc )

A SYSTEM

Just chart the EVENTS themselves and what do you end up with ?
what sort of chart ?

It won't have a TIME FRAME for a start
because you don't know in advance the periodicity of those EVENTS
( and why create lag ? )

YOU will end up with a chart like this

A pure EVENT chart

Motorway, sorry but I have no idea what you just said, and it doesn't seem to be anything about analysing the size of the initial stop?

All there are, are market events of a certain size

A child could work out were to place the stops on this chart

Someone a little older could apply some mathematics
THERE ARE ONLY TWO DIMENSIONS after all
each BOX is a digital unit of 4.43%
they go up or they go down
producing columns of a certain size and speed
have a characteristic relationship ==> "RIO"

The digital "FINGERPRINT" of "RIO"

"TLS" "BHP" etc will all have their own "FINGERPRINTS"

You could just call this chart
a contingent order chart
or a stop chart ( in fact an old name for it )

The events that would trigger an action
are all that are being represented


It is only such events that EXIST

Say you were trading between RIO and another stock
you see how the events themselves would time your trades
stop your trades, SWITCH YOUR TRADES ?

What people call systems are an overlay of such event triggers
I am pointing out that you could just chart the events..
with a pure event chart

All the things in that stop pdf
are before your eyes
All the stages


It is analysed for you
because it is a dynamic history

JUST PLACE THE STOPS
another stocks chart will reveal
that particular "HOW" in turn..

A whole lot of rubbish is not there because it really does not exist..
Only the events exist...

Just food for thought

Your question is answered
On this chart
By the size of the columns and their relation to each other..

Because it is only those that matter
that make the $$$
If the chart stops moving .. No more $$$

But I think there's a better way to go about it by analysing how much of our stop our previous trades have used.

motorway

 

[motorway]

Here is a TLS chart

9.05% with TLS is equiv to 4.43% with RIO

The events
trigger the trades
SWITCH THE TRADES
STOP the trades

so why chart anything extra ? ( anything extra is noise and lag )

could you use 4.43% with TLS too
SURE==> But there is no more relevant signal on it just longer columns...


( mentioned back in the thread )

AGAIN
you can see what stops to set and where


Motorway

 

[motorway]

( mentioned back in the thread )

And on that POST..

TRENDS ARE NOT SUBJECTIVE
when defined objectively

On that TLS chart
the move down from the top
ANY one doubt what they see ?

So who was buying ?

Those who think trends are SUBJECTIVE !
==which is ==>Who think trends are whatever they think or want them==> HOPE them to be !

ONLY the EVENTS matter ( and they do , LOOK AT THE CHART )

The chart has no indicators
It is as naked as you can get.

Pure events


So some idea
of

What analysis I use for trading

Motorway

 

[Mr J]

I did see P&F chart mentioned, I see what you're getting at. However, I've gone through phases with different charts - range, volume, tick, time etc - and came to the conclusion that they're all the same, it's just a matter of getting used to reading it. I'm tempted to try P&F out of curiousity, but Amibroker doesn't support it. There looks like there's some code out there for it though.

As for knowing where to place stops, I think that is best decided by analysis of previous situations, not by the chart that is used.

 

[brty]

Back to R:R.

When you are learning to trade you learn pretty quickly that a paper loss is a real loss. Holding on for the stock to turn is a mugs game.

However with the opposite occurring, a paper gain is often not treated as a real gain.

Personally I think money is money.

In Tech/a's example of a 3R trade , risking 10c to make 30c, the risk reward ratio changes during the trade, even when the stop is brought to 'break-even'.

When the trade is 20c in your favor, with the stop at 'break-even', you are still effectively risking 20c to make 10c. Unless of course 'won money' is different to 'lost money'.

Making any changes to a system immediately means changing the results of that system, so that employing a trailing stop to keep the loss at no more than 10c from the peak equity of a trade, means that there will be many times that the trailing stop takes you out of a trade before the 30c profit is hit.

My question is why do people treat an initial stop differently to a stop within a trade, ie only give it 10c to begin with, but happy to let it have up to 29c when trying for the 30c??

To me the notion of a trade being 'no-risk' just because the stock went in your favor, is not correct. You are still risking the gain.

brty

 

[Mr J]

In Tech/a's example of a 3R trade , risking 10c to make 30c, the risk reward ratio changes during the trade, even when the stop is brought to 'break-even'.

When the trade is 20c in your favor, with the stop at 'break-even', you are still effectively risking 20c to make 10c. Unless of course 'won money' is different to 'lost money'.

Yes, which is why I disagree when people talk of "free trades" when moving the stop to break even. It's not free, because the unrealised profits are still at risk.

However, I wouldn't agree that in your example the reward-risk has changed to 1:2 at all. We're not risking that 20c to lock in a 10c gain, we're risking it for further gains. Once a profit is realised, it can be ignored for the rest of the trade. The 10c is irrelevant as soon as we lock it in, it's the 20c currently at risk and the potential gains that are important.

 

[It's Snake Pliskin]

Yes, which is why I disagree when people talk of "free trades" when moving the stop to break even. It's not free, because the unrealised profits are still at risk.

But at breakeven there is no profit realised, nor unrealised.

Are you referring to opportunity cost?

 

[Mr J]

But at breakeven there is no profit realised, nor unrealised.

Sorry, you're right in that there's no realised profit, but there is certainly unrealised profit because when people move the stop to breakeven, the trade is in profit. I said realised profit when talking of the 10c locked in, because I wanted to distinguish it from the 20c of unrealised profit that isn't locked in.