Everybody hurts...sometimes...

[forker]

I have to agree with DT on this one. Time really isn't significant. What is significant are levels

 

[timsk]

And that is a great quote Tim.

Thanks DT and MrG', although the credit belongs to dbphoenix and not me . . .

"The most common error that people make here is the same error they make when "interpreting" candles: focusing on each bar or bar-pair or cluster as a discrete event. Whatever happens is part of what happened before and will affect what happens next. Therefore, focusing on the "waves" of buying pressure and selling pressure is far more productive than worrying about what this or that particular bar or bar-pair means. The market is a movie, not a slide-show."

Quoted from post #4 on P1 of the discussion thread relating to the article: Volume Analysis by Sebastian Manby.
Tim.

 

[LiamH]

@Scotty.. I agree re sample size needing more than most think although trading shouldn't take anywhere near as many instances to get reliable stats as poker.. In poker there are so many more ways to win\lose a hand that can skew your short term results one way or the other. In trading you're basically either right or wrong so assuming a mechanical strat it shouldn't take even 1% as many trades as it would take hands in poker to determine profitability.

 

[tafita]

Maybe wrong but sensing some sarcasm there. I wasnt even talking or thinking about if someone has an edge or not, was about how big a sample you need before you can start applying reliable probabilities to setups

Tharp suggests a minimum of 30 trades but recommends 100

 

[forker]

A 30 trade sample size should me more than enough to Judge how a method is working however, it should span a month at the very least. That way you are testing it through various data points.

 

[Hotch]

seeing... for technical analysis to "work" (whatever that means), lots of people need to see the same thing and arrive at the same decision...

I don't agree with this. Arrive at the same decision, quite possibly, but people don't need to see the same thing to agree.

There are all sorts of people who believe in God, but for different reasons. Similarly there are all sorts of reasons people are bearish bullish, you don't have to be bearish or bullish for the same reasons.

 

[forker]

I don't agree with this. Arrive at the same decision, quite possibly, but people don't need to see the same thing to agree.

There are all sorts of people who believe in God, but for different reasons. Similarly there are all sorts of reasons people are bearish bullish, you don't have to be bearish or bullish for the same reasons.

Well, unless you on the same page as the Market movers you don't make money. So there has to be common elements to analysis both news and technical

 

[scose-no-doubt]

for technical analysis to "work" (whatever that means), lots of people need to see the same thing and arrive at the same decision - which is why there are established timeframes that most people look at when making these types of decisions.

I've always thought of technical analysis as interpreting what the market moving big players and people who need to transact as part of their operations are doing.

Kind of like following clues or something... I've never even thought of approaching TA from the viewpoint that market movers will be using it.

Interesting point you make.

 

[Black Swan]

Thanks DT and MrG', although the credit belongs to dbphoenix and not me . . .

"The most common error that people make here is the same error they make when "interpreting" candles: focusing on each bar or bar-pair or cluster as a discrete event. Whatever happens is part of what happened before and will affect what happens next. Therefore, focusing on the "waves" of buying pressure and selling pressure is far more productive than worrying about what this or that particular bar or bar-pair means. The market is a movie, not a slide-show."

Quoted from post #4 on P1 of the discussion thread relating to the article: Volume Analysis by Sebastian Manby.
Tim.

Good quote

 

[Rooster34]

Can you remember the stage were you finally begin to get an edge/strat together, you follow it to the absolute letter (planning trades/trading the plan). For a couple of days you bag pips, then the next day you lose a couple of percent from your account.."but how, I followed the plan to the letter..?" Then the next day you have a losing day, not as bad as the previous, only 1%, but now the gains you made on the previous two are wiped out and you fear another losing day as that would completely blow apart the faith you had in the edge you thought you had, an *edge* that you've taken a month to get to grips with..You also realise it's going to take at least a week of trading to get back to where you were..you're going to need perhaps 4 days of straight wins to break even...5hit..

Ever done this; you trade off 10/15 min TFs, you have a great day, you take 6 trades, 4 winners 2 losses = 80 pips for the day. "That's it *grail* found! This time next year Rodney".. Then the next day you have 4 losses 2 winners, you give back your gains by losing 80 pips. You then look at what would have happened on day 2 if you'd traded using the same entry exit criteria and MM on a 1 hr TF. You'd have taken 2 trades, one loss of 50 pips one winner of 30 pips, only a twenty pip loss.."OK decision made, I'll move up a few TFs that'll sort it..!"

So you now trade off the 1 hr TFs, and guess what you have a 2 day losing streak, but you've taken 4 trades, 3 loss, 1 winner a loss over the 2 days of 60 pips but there's something different, you felt in control, despite the alerts flashing and all your indicators lining up screaming "trade" you didn't feel rushed, not that you were happy to give pips back to the market, but it somehow felt different to how you had traded off the 10/15 mins. Then it hits you, "what would it have been like trading off the original 10/15 mins?" You look and you see that you'd have taken ten trades off the 10/15, 7 losses 2 winners 1 breakeven,,"Fook, what a mess! Am I glad I moved away from that type of trading.." Then it hits you, a real Eureka moment in relation to trading, when the market is gyrating or stuck in a tight range, (which it does for the 70/80% of the time), you're best sitting on your hands because chasing it, engaging with it is a total waste of time, so if you do engage you're taking a much better probability shot than on the smaller TFs..

We're told to look after the money, but what about looking after your psyche, how emotionally (and at times physically) exhausted do you become if you take 4 losing 15 min TF trades (4 x 25 pip losses) in the morning session, realising you now may need 5 winners to have a decent winning day? Compare that to the *hurt* of taking one losing trade in the morning session for 40-50 pips and still feeling/believing that you can finish up for the day when NY opens.

If you trade FX off small TFs you need lightening reactions, you need the best/fastest hardware, you need close to zero spreads from the 'Autobahns' of this world, you need (genuine) NDD/ECN, you need to have a mindset that is crucially different from the normal retail trader, a mindest that has been honed to 'trader perfection', over years of proven success and mentorship from colleagues sitting next to you that regularly pull down mega salaries because they've proven their worth over the years. If that description doesn't fit you don't fight it, you can pick enough fights with the market where you may fight and live to fight another day, you'll never win trading as a normal retail trader if you trade off small TFs.

I started off on the 5m timeframe. Why? Because you could get a tight stop, crank up the leverage, and make 2-3% of your account several times a day. Predictably I lost 75% of my account within a few months. I only started with a few hundred dollars, but I lost over 1500 pips. fortunately they were low value pips.That hurts. It's real pain.
I moved to the daily timeframe and viola, I've won 3 out of 4 trades for a positive gain of well over 700 pips, and I should have let them run even farther. I'm working on that. I was infatuated with the low timeframes, I'm at home with the larger ones. It's almost stress free.
The thing is, that I think this kind of trading gives you a life too, no need to sit in front of a computer from London open to NY close trying to spot a few small moves.
Having said that, there are guys over in the forex live calls thread that do just fine on smaller TFs. But they aren't noobs......

 

[BSD]

Haven't read the thread, but one thing is clear, the big money is in the big moves, and trading intraday means you're pitting transaction costs against your chances of ending up net profitable while cutting short the profit potential by exiting for the arbitrary reason that the day is over.

No wonder Paul Tudor Jones, who started out as a broker, realized that 95% of daytraders are losers, the odds are simply against you.

Nobody can catch all intraday micro fluctuations in the real world (wasps post facto charts implied that he could, but then we know what happened then haha), so the safest and most lucrative way is to forget chasing non existent perfection, and instead just follow the path of least resistance and catch the big swings.

Forget winning percentages, go for great risk / reward trades you can hold for a couple of days, and you'll have more money than you know what to do with.

 

[the hare]

A 30 trade sample size should me more than enough to Judge how a method is working however, it should span a month at the very least. That way you are testing it through various data points.

The uncertainty in the stats is a function of the square root of the sample size. I would suggest that those of us who have a decent edge, returning for example double digit monthly returns are acutely aware of just how small that edge actually is.

The sampling error based on 30 trades is quite possibly in most cases, orders of magnitude above the edge which your trying to measure.

To add to the confusion, you have to remember that your dealing with a non stationary time series. That is the harsh reality.

Of course in the fantasy world of T2W everyone has very sizable edges that allow them to extract thousands of points per week from any instrument, with no losses or drawdowns. If thats the case, a sample of 30 is just about OK.

 

[Black Swan]

The uncertainty in the stats is a function of the square root of the sample size. I would suggest that those of us who have a decent edge, returning for example double digit monthly returns are acutely aware of just how small that edge actually is.

Is it probability at work that the last 3 posts have (imho) nailed it?

 

[Black Swan]

I started off on the 5m timeframe. Why? Because you could get a tight stop, crank up the leverage, and make 2-3% of your account several times a day. Predictably I lost 75% of my account within a few months. I only started with a few hundred dollars, but I lost over 1500 pips. fortunately they were low value pips.That hurts. It's real pain.
I moved to the daily timeframe and viola, I've won 3 out of 4 trades for a positive gain of well over 700 pips, and I should have let them run even farther. I'm working on that. I was infatuated with the low timeframes, I'm at home with the larger ones. It's almost stress free.
The thing is, that I think this kind of trading gives you a life too, no need to sit in front of a computer from London open to NY close trying to spot a few small moves.
Having said that, there are guys over in the forex live calls thread that do just fine on smaller TFs. But they aren't noobs......

On T2W? Are you sure on that?

 

[LiamH]

On T2W? Are you sure on that?

I think Wackypete, redart, dinos and chowclown all do pretty well on the lower tf's don't they?

 

[DionysusToast]

Haven't read the thread, but one thing is clear, the big money is in the big moves, and trading intraday means you're pitting transaction costs against your chances of ending up net profitable while cutting short the profit potential by exiting for the arbitrary reason that the day is over.

I make on average 1 or 2 trades a day. Transaction costs are hardly killing me.

Nobody can catch all intraday micro fluctuations in the real world (wasps post facto charts implied that he could, but then we know what happened then haha), so the safest and most lucrative way is to forget chasing non existent perfection, and instead just follow the path of least resistance and catch the big swings..

Well - I don't think anyone claims to be able to catch all intraday fluctuations or to be chasing 'micro' ones.

Are you actually saying that no day traders make money ?

I'm not sure how big the pot is that you p1ss in but the most succesful day trader I know personally has a little over $18M in his personal accounts. All from day trading stocks.

He doesn't catch micro moves either.

Should I call him and break the news to him that he actually doesn't exist?

I think you are confusing day trading and scalping.

As was mentioned earlier - it's DAY trading, not 10 second trading.

 

[Hotch]

Well, unless you on the same page as the Market movers you don't make money. So there has to be common elements to analysis both news and technical

Why? As long as you come to the same conclusion it doesn't matter what methods you use.

 

[DionysusToast]

Why? As long as you come to the same conclusion it doesn't matter what methods you use.

Hmmmm - but what are the chances of that ?

 

[bedsit]

Forget winning percentages, go for great risk / reward trades you can hold for a couple of days, and you'll have more money than you know what to do with.

I like the idea - made me question my trading as I do only intra-day trading.

I'll check some longer term set ups and see if I can achieve better results with less effort (don't like getting up early every morning).

 

[diy_]

Oonu see weh mi a seh, see star. Yuh read higha time bars wen di price a lick di level dem an yuh affi brook dem down, hopen dem up soh an look pon di lower time bars fi get a hidea of wat a gwaarn inna de higha bar sinten dem. Puuure info an data inna dem ya knuh.

rass!! a ja deh iyah come fram?