Just occasionally while reading a book I hit something that throws me so much I can’t continue reading it until I get it sorted out.
I’m currently part way through Your Money and Your Brain by Jason Zweig. Subtitled Become a smarter, more successful investor the neuroscience way it’s a look at what’s happening in our brains when we react (sometimes bizarrely) to the financial markets – rather appropriate at the current time.
Here’s the stopper. According to Mr Zweig:
People were told that an expert claimed that the market always went up after he predicted it would rise. They were told they could verify the expert’s claim by choosing to observe any or all of the following evidence:
1. What the market did after he predicted it would rise
2. What the market did after he predicted it would fall
3. What he predicted before the market rose
4. What he predicted before the market fell
Then they were asked what was the minimum evidence they would need in order to establish for certain that the expert’s claim was true. Fully 48% of the people responded that No 1 was all they would need. Only 22% gave the correct answer: the minimum evidence needed to see whether the experts claim was true is No 1 and No 4. Even though he says the market always goes up when he predicts it will, you still need to know what he said before the market went down. (After all it does not always go up.) Subjecting him to both these tests is the only way to be positive about the truth.
Now I can’t see this. Assume we apply test 1. To require a second test, there must be a circumstance where the second test will provide information about the statement (‘the market always goes up after I predict it will rise’) that isn’t provided by the first test. I can’t think of a circumstance where this will occur.
Maybe I’m being dumb, along with the ‘professors and graduate students of the University of London’ who provided those 48% ‘wrong’ answers, but I can’t see it.
What I think Zweig means is that the after the expert says ‘the market will rise’ the market might rise, then fall without the expert saying anything else. If so, however, Zweig is incorrect to say the 48% were wrong. They correctly test the expert’s assertion. He doesn’t say ‘and it won’t go down again afterwards.’ Only test 1 is required establish what he says is true – if Mr Zweig wants to read anything into it, that’s his mistake!
If you are going to use an example to prove something it ought to prove it – and for me, this does entirely the opposite. But I am being dumb, so I’m probably missing something. Any volunteers to show me where I’m going wrong?
I’m with stupid too (and sequels are never as good as the original).
It only matters what he said before the market went down if he said it would rise. But that’s covered by 1 already.
Maybe I’m dumb too but how useful is that kind of prediction anyway when markets raise and fall. If it’s raining I can make a prediction it will stop raining, does that make me a prophet?
Sabine –
The theory is that the ‘expert’ predicts the market’s next significant change will be a rise, so as a day trader you can in principle make a profit before it subsequently falls again.
To be fair to Mr Zweig, he’s using the example in the context of saying there’s no point listening to so-called experts. But my issue was not with whether or not the prediction is useful, but rather the dubious logic in the testing of the statement that led to those professors and graduate students being labelled as ‘wrong’.
Hi Brian, Thanks for the clarification. Do I get that correctly that should understand a prediction for ‘will rise’ as ‘will rise without previously falling’? I am still confused though how useful is this prediction without a time for the ‘next significant change’ and ‘subsequently falls again’.