Catalogue of Strats Which DID NOT Work

[Tams 20]

somewhere else in the forum, there is a thread on

Catalogue of Strats Which Worked

http://www.traderslaboratory.com/forums/f106/catalogue-strats-worked-7750.html

 

well... let's attack the grail from both sides,

let's assemble a

Catalogue of Strats Which DID NOT Work

 

the other thread focus on automated systems,

I am starting this thread in the general trading area,

so that we can talk about Concepts, Methods, Tactics, Strategies as well as Systems.

[Tams 20]

Let me kick off the discussion...

 

regarding the old adage... the Trend is Your Friend

 

 

one strategy that does not work is to go against the trend.

 

but, you might ask... how do I know the trend?

 

a kid can eyeball the chart and tell you the trend.

but we are not kids,

we over analyze things,

we can see the trees, but not the forest.

 

ok, a few people other than kids can see the trend too,

I am referring to the price action folks.

 

but not everybody are gifted to be price action traders, what am I to do?

 

 

May I suggest a moving average?

any moving average would do... simple, exponential, weighted, zero lag... etc.

 

what about the length?

try 20 period to start... then adjust as required.

 

try this analysis:

measure the trough to peak range of the moving average...

which range is bigger?

the down leg or the up leg?

 

ask yourself...

which leg has more profit potential?

which leg would you rather be on?

 

 

 

side note:

what not to trade:

instrument that has little volume

instrument that has little direction

they maybe good for investing/gambling, but not trading

[macdfx 10]

If you consider using moving averages as trading with the trend, then what is your opinion about "moving average crossovers"? Do they work for you?

[Tams 20]

If you consider using moving averages as trading with the trend, then what is your opinion about "moving average crossovers"? Do they work for you?

 

you are welcomed to start a new thread on the subject.

[macdfx 10]

:bang head:

you are welcomed to start a new thread on the subject.

I thought this was a discussion, and you brought up moving averages.

 

I expounded on your comment and asked about moving average crossovers and your opinion of them....

 

Or what exactly is it that you are trying to discuss here?

[Tams 20]

:bang head:

I thought this was a discussion, and you brought up moving averages.

 

I expounded on your comment and asked about moving average crossovers and your opinion of them....

 

Or what exactly is it that you are trying to discuss here?

 

I am trying to assemble a

Catalogue of Strats Which DID NOT Work.

 

if you think MA crossover does not work, then post it here.

if you don't know if it works or not, then start a thread to discuss.

[BlowFish 10]

Actually tough to come up with qualified answers.

 

I'll offer this one buying 'oversold' selling 'overbought' markets based on oscillators (such as stochastic s) in trending markets.

[Frank 10]

doing the same things in different market environments.

[uexkuell 10]

NOT working:

 

Trying to reverse a NOT working strategy and thereby producing a working strategy.

 

That's why this thread is pointless.

 

 

 

Some strategies that are not working:

(For the following assuming that "not working" means "producing loss")

 

 

Strategy 1.

Enter.

Wait until the position goes into negative.

(=inverse buy&hold)

 

Strategy 2.

a) Enter.

b) When positions turns increasingly negative increase position size (=try to average down).

c) When positions comes back close to break-even liquidate most of position (=realize loss).

d) Wait until condition b) is met again.

 

Strategy 2 is very frequently used obviously.

 

 

What can be taken from the examples:

Bad money management is a fail-safe road to loss.

(Point of entry is not that important if you want to loose big.)

[Tams 20]

NOT working:

 

...Strategy 2.

 

b) When positions turns increasingly negative increase position size (=try to average down).

 

average down is no doubt the biggest sin of all.

 

I have yet to meet a long term survivor who uses average down as a strategy.

(it might work on a buy and hold long term investment, but even Buffet got burnt a few times)

 

 

a side note:

What is the quickest way to lose the "newbie" moniker? --- try average down.

Edited April 15, 2010 by Tams

[mike21 10]

If you consider using moving averages as trading with the trend, then what is your opinion about "moving average crossovers"? Do they work for you?

 

using 'moving average crossovers' is definitely a strategy that DOES NOT WORK

[macdfx 10]

using 'moving average crossovers' is definitely a strategy that DOES NOT WORK

 

OK, but any insight as to WHY NOT, since moving averages themselves seem to follow the trend?

 

Kinda curious, as that's what I was looking to gain some further understanding about here....

[Tams 20]

OK, but any insight as to WHY NOT, since moving averages themselves seem to follow the trend?

 

Kinda curious, as that's what I was looking to gain some further understanding about here....

 

lots of talks here:

 

http://www.traderslaboratory.com/forums/f34/ma-cross-ranging-day-whipsaws-5995.html

http://www.traderslaboratory.com/forums/f34/fresh-first-crossover-strategies-7338.html

http://www.traderslaboratory.com/forums/f34/moving-average-crossover-volume-surge-2945.html

[macdfx 10]

thxs.................................

[brownsfan019 17]

I have yet to see any evidence that the Jack Hershey / Spydertrader way of drawing lines all over the place work. I would definitely put that in the does not work category.

 

 

We could however add it to the does not work but people cling to it like it does category.

[Tams 20]

risking one dollar for a 50 cents target is definitely not going to work...

[DugDug 10]

This one is debatable - and also a long term trading subject - and also questionable in terms of what is long term, what parameters are used, what markets are traded.

The original turtle trading rules don't appear to work anymore.

They worked very well from the 70s to the 90s then have been pretty poor since. Variations on the model appear to work, however the original seem to have stopped working. (weather it works again in the future is the debate)

[BlowFish 10]

OK, but any insight as to WHY NOT, since moving averages themselves seem to follow the trend?

 

Kinda curious, as that's what I was looking to gain some further understanding about here....

 

Could be because a lot of the time markets don't 'trend' (depending on how one defines 'trend' of course).

[BlowFish 10]

risking one dollar for a 50 cents target is definitely not going to work...

 

Not sure about that. As long as you are winning round 70% why not?

[lote_tree 10]

Ok, so I’m still waiting to see whether anybody can provide any evidence to support their positions on strategies they deem to be ineffective.

 

I've been going through this thread and have seen claims being made that a certain method doesn’t work yet no rigorous data has been provided to support such a hypothesis. I thought evidenced based trading was the fulcrum of good trading but hey, what do I know. May be there is a floating teapot in outer space! :missy:

 

If I came across somebody making a claim that they think XYZ method is ineffective I would immediately ask them the following questions:

 

1. Did you understand the basis of the methodology?

2. If you did understand the basis of the methodology how did you know that you understood it correctly?

3. Did you test the method over large enough data set?

4. Did you conduct out of sample tests?

5. Did you forward test the method and compare it to the statistics produced from your out of sample tests?

6. Did you test whether the method produces trades better than random?

 

These are just a few that come to mind.

 

Testing new trading strategies is like going out on many different dates. You have to interview the woman before you allow her into your bedroom. She will either turn you on or turn you off depending on what answers she gives. It's the same with trading strategies. You have to interview your trading strategy so it can prove to you whether it's worthy to have your hard cash inserted into! Treat new methods with respect and give them a chance through rigorous enquiry. One day you may get lucky with one and it may end up being a cash cow for you

 

Peace..........

[DugDug 10]

 

Testing new trading strategies is like going out on many different dates. You have to interview the woman before you allow her into your bedroom. She will either turn you on or turn you off depending on what answers she gives. It's the same with trading strategies.

.

 

except those women who look so good you just throw your cash at them in the hope they will let you into their bedroom.

 

And then you marry them and they stop working!

 

(i know this is very un PC - i apologise for those who are offended)

[Tams 20]

risking one dollar for a 50 cents target is definitely not going to work...

 

Not sure about that. As long as you are winning round 70% why not?

 

mathematically it sounds good... if life (ie trading) is a linear arithmetic event.

 

consider this:

 

trade#1= 0.50

trade#2= 0.50

trade#3= 0.50

trade#4= 0.50

trade#5= 0.50

trade#6= 0.50

trade#7= 0.50

trade#8= (1.00)

trade#9= (1.00)

trade#10= (1.00)

net = 0.50

 

 

1. if you win in the beginning, then lose later, you will break even.

but life is not linear... if you experienced multiple losses before accumulating enough "risk" capital, your predicament is a fast wipe out.

 

2. in reality... most practitioners would win some, then lose some, then win some again...

in the best case scenario, the trading career would ended up a statistically perpetual borderline break even event.

 

3. people increase their trade sizes (especially during a loss. ie average down)...

what happens is, whatever (meager profit and inflated ego) you accumulated at the beginning is not enough to cover the magnified losses later.

 

4. people move their stops and take quick profits... (I know, that is another story....)

 

 

there are more to it...

but this should be enough to stop any semi-intelligent newbie.

Edited April 16, 2010 by Tams

[BlowFish 10]

mathematically it sounds good... if life (ie trading) is a linear arithmetic event.

 

consider this:

 

trade#1= 0.50

trade#2= 0.50

trade#3= 0.50

trade#4= 0.50

trade#5= 0.50

trade#6= 0.50

trade#7= 0.50

trade#8= (1.00)

trade#9= (1.00)

trade#10= (1.00)

net = 0.50

 

 

1. if you win in the beginning, then lose later, you will break even.

but life is not linear... if you experienced multiple losses before accumulating enough "risk" capital, your predicament is a fast wipe out.

 

2. in reality... most practitioners would win some, then lose some, then win some again...

in the best case scenario, the trading career would ended up a statistically perpetual borderline break even event.

 

3. people increase their trade sizes (especially during a loss. ie average down)...

what happens is, whatever (meager profit and inflated ego) you accumulated at the beginning is not enough to cover the magnified losses later.

 

4. people move their stops and take quick profits... (I know, that is another story....)

 

 

there are more to it...

but this should be enough to stop any semi-intelligent newbie.

 

No. This does not happen if you properly calculated your position size to give an acceptable risk of ruin. And you know what a high win rate actually is much better for risk of ruin than high R:R. It also tends to produce a smoother equity curve. This means you can actually bet larger with the same RoR as a lower %winner system

 

Sure you get streaks of losers but they are much shorter with a 70% win rate than with say a 30% win rate. Trying to find some tables for you but can't right now having a senior moment with google.

 

Edit: I should add I would probably want 80%+ to trade 2:1 risk:reward

[Tams 20]

Picking Top/Bottom does not work.

 

if you have the power to profitably pick top/bottom...

you should also have the power to correctly pick the 6 numbers to win a lottery.

[sneo 10]

i disagree with picking tops/bottoms and trading against the trend as "do not work" strategies.

 

why? because:

 

1. professionally, i trade profitably using both reversals and trend following strategies.

 

2. factually, Vegasoul (a highly successful hedge fund) has three major components to their strategy, of which one is trading reversals. their results and strategy summary proves that reversal trading is not a "do not work" strategy. for the summary of their strategy and results, its available in EurekaHedge (subscription based) and Cogent Hedge (free upon registeration).