Analysis: retrospective narratives

[trendie]

What do we mean by analysis?
I am always fascinated by the people who construct retrospective narratives to "explain" why something happened?

Why do we need an explanation?
For example, The price of [commodity/currency] rose/fell BECAUSE of reason A/B/C.

Its the "because" bit that fascinates. Why must something have a "because".
NFP comes out. figures better than expected. SPX goes down.
reason is because "news was factored in".

NFP comes out. figures better than expected. SPX goes up.
reason given because "market was reacting to good news".

Whether market goes up on good news, or goes down, there very rarely is a consistent pattern of prediction.
But the reasons, the retro-analysis, seeks to make sense of different reactions (going up/down) to ostensibly similar initial parameters (good news).

My trading is simply reacting to situations, and trading a pattern.
I have no expert knowledge other than playing the odds.
I KNOW the trade wont always work out.

But I DONT try to "make sense" of it, other than to monitor the odds of wins/losses to determine change in market characteristics over time. (x number of trades)

Why is there so much analysis?
Why do we need to impose a narrative?
Is it that we seek to impose a sense of "meaning", since we are uncomfortable with the idea that similar initial parameters give rise to different outcomes?

It is what it is.
Why the huge industry to "make sense" of things?

 

[brewski1984]

Too true. I'm going to be one of those suckers that quotes Livermore and say it doesn't matter why it happened all that matters is it happened

The ones that really annoy me are the random people/companies that post occasionally and say something like "Gold dipped below 1700, if it carries on going down we will likely see 1675 as support but if it breaks back through 1700 we will likely see 1740 as resistance." Yeah thanks mate, awesome insight

 

[Shakone]

I don't believe everything works on a cause-effect basis, and I don't think it is necessary to know the cause to make a profit. But I do think it can give confidence, and I do think it is one of the best ways to learn. Why did it work this time and not that time? Why is this support stronger than that one, even though they seem the same? etc. Whether the answer you get is the real cause is less important than whether this process can improve your trade selection. And I believe it can.

Also cause and effect is great for manipulation.

 

[wackypete2]

Cause and effect is more reliable on a long term fundamental basis. If the Fed, Eurozone, or other sovereign fundamentally change their tactics or outlook then you can see a long term effect. For short term news trades it's pot luck.

Analysts play the retro-analysis game because it makes them look good. Simple as that.

I agree with Shakone though, newer traders need some indication of what went wrong. It is how they learn. But experienced traders who can make a profits with a 50% win rate don't need to know why each wrong trade went bad. They know it's within the tolerances of the method they are using. That's all that matters.

Peter