Please Note: Blog posts are not selected, edited or screened by Seeking Alpha editors.

Book Review: Millard on Channel Analysis

One benefit I get from being a contributor to Seeking Alpha is that I get chances to read and review books from a large selection, and this time I decided to read a book that is out of my normal reading list so that I can practice jumping out of my existing mindset. 

I picked Millard on Channel Analysis published by Harriman House.

Being a hardcore 'fundamentalist', I do not have a good understanding of technical analysis, but I do recognize that stocks often demonstrate strong tendencies following certain patterns regardless of fundamentals. For example, 'fill the gap' is a typical common phenomenon. And stocks tend to channel up, down or sideways. The S&P 500 has been in an up channel since March, except for the past month during which it has been flat. So I was searching a book that talks in depth about channels, and I decided to read Brian Millard's book. 

Harriman House was very kind in sending me the book from UK, so I feel awkwardly sorry to say that I cannot recommend this book, especially to investors who trust fundamental analysis more than technical analysis. 

The book started with evidence of stocks trend in channels, which prompts investors to predict stock prices based on channel analysis. 

The book then used three chapters to look at the nature of the trend. The stock price movement looks quite like aggregation of sine waves with different amplitude, frequency and phase. So the author went on simulating such systems and comparing them to the stock price movement.

This makes perfect sense to me because it is how scientific research is done. You first observe a phenomenon; second you figure out a model; third you start the model with the simplest form; fourth you add in extra features to get it closer to the real physical world. Test with fresh data. If the model describes the data well, you have a publishable scientific paper. 

This is no coincidence because Brian Millard was a scientist. He was a senior lecturer at the University of London until 1980 and published over seventy scientific papers. The darn thing is that I happen to have gone through all the formal science training, published four papers and decorated myself with a doctorate degree. So I know the weakest link in the whole process -- the model. A good model describes not only the data, but the nature of the phenomenon as well. Many times you can fit a made-up model, but the model just lacks the power to predict. In my view, cycles can describe stock prices, but such a theory is made up for the form, not for the nature. Consequently such a model is weak when used to predict future price movement. 

The book went on using the existing price chart to derive intrinsic price cycles and use that to predict the future price movement, with both simulation and some real stock data. Several examples are given, but they are not generic enough in my view to draw reliable conclusions. 

The book concluded with advocating disciplines in practice. Such as preserving capital, diversify and stop loss. 

I am sorry to say that this book is not my cup of tea. I probably was looking for more of a book talking about channel pattern, volume momentum and market psychology. My short of recommendation could also just be my own failure to jump out of my shell because I could not bring myself to believe the efficacy of applying fourier series decomposition on stock prices. 


Disclosure: No Positions