Myths of the myth busters?

[daxdaytrader]

I read some material on another site that I found quite questionable...

"The idea that you can use Fibonacci sequences to predict retracements is a favorite ruse of market gurus"

I found this a little amusing as I know a couple of wealthy pro traders who uses Fibs for market exits. Now I realise that doesn't prove they give any statistical edge. On the other hand, since one of them manages millions of dollars of funds, and if market makers decide that is how they are going to exit, then it might well "work", irrespective of the lack of science - price is also reflective of beliefs.

The same site had another guy mention that candlesticks don't work!

"as you shift the time boundaries of the Candlestick formations they could change from a Dragonfly Doji to a Gravestone Doji or shift from a Bearish Harami to a Bullish Harami"

His case is that if you have a 15 minute candlestick and move the candlesticks start times along say 5 or 10 minutes you will get different formations. Of course you will, but that does not mean that the outcome of candlestick formations is random, and one is moving the sample boundary, not the sample data. If a dragonfly forms on any chart, the price undoubtedly did drop down and lift off again up to the same level, no amount of timeframe shuffle changes the sample data. Furthermore, the first candlestick on the chart represents the first 15 minutes of the session - if you want to start it 5 minutes into the session, that's your business (and how would you plot that first 5 minutes?), but I wouldn't think that many people would do that, so you are likely to misread the market. I've found candlesticks plus filtering is very effective, Gregory Morris proves it so in his books.

It seems these trading boards often fill up with messages from folk who really need to learn basics before they comment.

 

[TheBramble]

Good post DDT. And I think you've highlighted the answer. If indeed there was a question.

Fibs work because they work - much of the time. They fall into the same category as any other self-fulfilling prophecy. I'm not saying there isn't any supra-normal aspect to them, just that expecting them to work will cause them to do so if most people believe they will.

As for candlesticks - it's largely irrelevant which bar type you use - your argument is identical - and correct - for all of them.

As you change your timeframe the bar/candle whatever formations can change. So who is the 5 min bar trader trading against? And within those higher timeframes - with whom does that trader trade?

It is the existence of folk trading different timeframes feeding up into the higher timeframe. That's why it makes sense to keep an eye on those higher timeframes. And the lower ones too for timing. You get a different picture from each which tends to add to the overall harmonics.

Liked your post. Good thought.

 

[Splitlink]

Good post DDT. And I think you've highlighted the answer. If indeed there was a question.

Fibs work because they work - much of the time. They fall into the same category as any other self-fulfilling prophecy. I'm not saying there isn't any supra-normal aspect to them, just that expecting them to work will cause them to do so if most people believe they will.

As for candlesticks - it's largely irrelevant which bar type you use - your argument is identical - and correct - for all of them.

As you change your timeframe the bar/candle whatever formations can change. So who is the 5 min bar trader trading against? And within those higher timeframes - with whom does that trader trade?

It is the existence of folk trading different timeframes feeding up into the higher timeframe. That's why it makes sense to keep an eye on those higher timeframes. And the lower ones too for timing. You get a different picture from each which tends to add to the overall harmonics.

Liked your post. Good thought.

I use EOD charts to come to a decision on entry and stop levels and then, the next day, watch the lower time frames. The patterns are quite different and, therefore, the candlesticks must be, too.

Split