XAO Banter Thread

[Trembling Hand]

But the pattern has to be very symmetrical, neat and uncluttered.

There may be one of your problems with the endless search for profitably. Looking for the holy grail in the perfect pattern.

The edges are in the ugly patterns. Not some nice gentle "look everyone can see it" setups.

 

[skyQuake]

I'd say in general the longer the time frame, the more reliable. But the pattern has to be very symmetrical, neat and uncluttered. In other words it has to be obvious to all those who make the prices move - the institutions.

The more perfect the pattern, the more I rub my hands with glee and go in with size for the fls brk.
Moreso futures than equities.

 

[coolcup]

Err you sure about that? Other than folklore you have any evidence that the statement that H&S patterns are reliable?

I am not trying to say that a H&S is reliable or unreliable. I don't really care either way. The point I am making is that the time frame it is measured over does not of itself affect the reliability of the pattern. So if you believe a H&S is reliable (am not saying it is) or any other pattern for that matter, it should be reliable over a short time period or a longer one.

It was in response to nulla nulla who was saying my observation was false because it wasn't over the exact timeframe as the chart leading up to the GFC. I wasn't trying to show an exact match it was just a point of interest to me that the chart patterns looked similar.

 

[Gringotts Bank]

The more perfect the pattern, the more I rub my hands with glee and go in with size for the fls brk.
Moreso futures than equities.

That makes sense for a shorter time frame because it's likely that the price will revisit the breakout point. When you're covering your short, I might be buying for a longer time frame.

 

[tinhat]

I'd say in general the longer the time frame, the more reliable. But the pattern has to be very symmetrical, neat and uncluttered. In other words it has to be obvious to all those who make the prices move - the institutions.

My observation is that P&F charts display more symmetry than time based charts.

 

[Gringotts Bank]

5000 tomorrow. :bananasmi

 

[Gringotts Bank]

Distribution happening last 5 days, imo. Any lower today and a trendline is broken. Might see the banana dance yet!

 

[sinner]

The point I am making is that the time frame it is measured over does not of itself affect the reliability of the pattern. So if you believe a H&S is reliable (am not saying it is) or any other pattern for that matter, it should be reliable over a short time period or a longer one.

My experience is the opposite of this. Timeframe plays a huge role into the reliability of a "pattern", at least those which you can quantitatively define and therefore test.

Do you have any evidence to back up your claim that the time frame does not affect the reliability of the pattern? I can easily show backtests which invalidate this claim.

 

[tech/a]

My experience is the opposite of this. Timeframe plays a huge role into the reliability of a "pattern", at least those which you can quantitatively define and therefore test.

Do you have any evidence to back up your claim that the time frame does not affect the reliability of the pattern? I can easily show backtests which invalidate this claim.

Agree with this.
Patterns have a life span.
You could hardly credit a 200 Tick move in the SPI
to a rectangle breakout of a 10 tick range.

Patterns form in all time frames and some contradict others.
They are simply graphical representation of those involved in the trading of the instrument.

 

[coolcup]

My experience is the opposite of this. Timeframe plays a huge role into the reliability of a "pattern", at least those which you can quantitatively define and therefore test.

Do you have any evidence to back up your claim that the time frame does not affect the reliability of the pattern? I can easily show backtests which invalidate this claim.

Far out dude. I am not interested in back validating anything. I just saw a similar pattern now to pre GFC which was amusing / interesting. If you don't believe in it, then that's fine. I am not advocating it either way, just thought it was interesting.

 

[Gringotts Bank]

I don't like the way the market is now. Needs some fear in it. Without fear there's no market, because fear is greed's flipside.

Maybe that's what Bernake is trying to do - remove fear from the market.

- - - Updated - - -

My experience is the opposite of this. Timeframe plays a huge role into the reliability of a "pattern", at least those which you can quantitatively define and therefore test.

Do you have any evidence to back up your claim that the time frame does not affect the reliability of the pattern? I can easily show backtests which invalidate this claim.

Sinner can I see your backtests on pattern reliability and time frame please?

 

[Trembling Hand]

I don't like the way the market is now.

You never do. And seem to spew up at all the wrong times. Maybe you would be better just trading a system?

 

[sinner]

Sinner can I see your backtests on pattern reliability and time frame please?

Sure, pick a quant pattern, and two timeframes, and I can show you 1000 bar results for both.

If you're picking an intraday timeframe for comparison then I'll be testing FX, because that is the high quality intraday data I have.

I can do daily/weekly/monthly for everything else.

 

[Gringotts Bank]

Sure, pick a quant pattern, and two timeframes, and I can show you 1000 bar results for both.

say descending wedge, upwards breakout, daily vs weekly and weekly versus monthly.

[edited]

 

[sinner]

say descending triangle, daily vs weekly vs monthly.

lol, please quantify what is a descending triangle, that is really not what I meant by quant pattern.

Maybe I should've given some examples: 20 day closing high, last bar in lowest 20% of N day range, Close above 20 day closing average price...hold for N days.

 

[CanOz]

say descending wedge, upwards breakout, daily vs weekly and weekly versus monthly.

[edited]

Not sure you'll find a ton of weekly monthly wedges...what about daily, 60m and 5m?

 

[Gringotts Bank]

Maybe I should've given some examples: 20 day closing high, last bar in lowest 20% of N day range, Close above 20 day closing average price...hold for N days.

That example is fine thanks.

 

[Zedd]

I don't like the way the market is now. Needs some fear in it

Agree. I felt that in the last 12 months we'd finally moved out of the skittishness of the GFC, and massive swings between overly optimistic and overly pessimistic. It felt like global macro-economic news was being incorporated into movements - ie. good news = market up, bad news = market down. Since the "tapering" comment, where we had a brief pull back, optimism is clearly in charge again, and nothing seems to be tempering the move up. FOMO but quite concerned that there seems to be a lot of collective back-patting going on.

 

[CanOz]

Agree. I felt that in the last 12 months we'd finally moved out of the skittishness of the GFC, and massive swings between overly optimistic and overly pessimistic. It felt like global macro-economic news was being incorporated into movements - ie. good news = market up, bad news = market down. Since the "tapering" comment, where we had a brief pull back, optimism is clearly in charge again, and nothing seems to be tempering the move up. FOMO but quite concerned that there seems to be a lot of collective back-patting going on.

lol, now its like good news = market goes up because its good news! Bad news = market goes up because a possibility of a later taper!

 

[sinner]

That example is fine thanks.

Hehe, those were a list of examples? Sorry, I can show whichever you like, but please specify exactly which one ...also with the caveat I prob don't have 1000 bars of monthly data for anything but the SP500 and US credit?