Evidence based TA, anyone ?

[combotrader]

No he says nothing about hit rates in the book. That was just me:clap:

He goes for mean returns as the main parameter to judge success. Obviously this could take other forms : sharpe ratio etc.

For those who like theory the book delves deeper into cognitive and psychological biases for thinking " my hardworked backtest finally works". In a nutshell the author is criticising backtesting results one takes for granted.

Furthermore - conventional bootstrapping is not enough. A white reality check has to be attached to it.
BSD - the osler study you mention covers only 2 years. Merely bootstrapping is not the same as statistical inferencing.

 

[zupcon]

I really cant make my mind up about this book. On the one hand, its treatment of statistical techniques, and their application to technical analysis is extremely sound, and for that reason, it remains one of my favorite books.

The book also frustrates me immensely. I believe the author has some institutional trading experience, but most of the time he just comes across as someone who's completely bewildered by whats going on.

Maybe its just my interpretation but the book almost has a feeling of absolute desperation about it. Its like he's spent years getting his head around the techniques, back testing 1001 lame unprofitable idea's, given up and then thought sod it, I'll write a stats book for traders. Its the book equivalent of listening to Leonard Cohen

 

[DaxDestroyer]

I really cant make my mind up about this book. On the one hand, its treatment of statistical techniques, and their application to technical analysis is extremely sound, and for that reason, it remains one of my favorite books.

The book also frustrates me immensely. I believe the author has some institutional trading experience, but most of the time he just comes across as someone who's completely bewildered by whats going on.

Maybe its just my interpretation but the book almost has a feeling of absolute desperation about it. Its like he's spent years getting his head around the techniques, back testing 1001 lame unprofitable idea's, given up and then thought sod it, I'll write a stats book for traders. Its the book equivalent of listening to Leonard Cohen

You could be right

 

[Mathemagician]

I really cant make my mind up about this book. On the one hand, its treatment of statistical techniques, and their application to technical analysis is extremely sound, and for that reason, it remains one of my favorite books.

The book also frustrates me immensely. I believe the author has some institutional trading experience, but most of the time he just comes across as someone who's completely bewildered by whats going on.

Maybe its just my interpretation but the book almost has a feeling of absolute desperation about it. Its like he's spent years getting his head around the techniques, back testing 1001 lame unprofitable idea's, given up and then thought sod it, I'll write a stats book for traders. Its the book equivalent of listening to Leonard Cohen

This is entirely possible. I guess there's a reason pharmaceutical companies hire both statisticians and biochemists...

jj