There is a post here that looks at the entire question and basis of ‘Dow Theory’ from two different comments:

I always get a kick out of reading how history repeats itself and this empirical historical data validates one’s view point today. Here is my viewpoint. Whatever happened in the past doesn’t mean squat in this new paradigm . Look at the 50 year S& P chart, look at when computers started trading.

Now ask yourself what is the stock market … a vehicle for companies to raise cash to support growth and innovation. I don’t think so my naive pal . Its a mechanism to extract as much money from the economy as it can with HFT, CDO’s and all means of smoke and mirrors so excuse me when I laugh in the face of a anachronism such as yourself . I’m an early adopter and I get it…all things flawed rule and your charting was not flawed in previous paradigms therefore cannot rule now.. Nice try though.

Brad, it’s a game of risk and probabilities, that’s all. Your comment shows that you don’t understand the post or don’t trade often. The post doesn’t try to take a 1920′s theory and fit it to today. It uses the theory to allow a trader to access risk and trade based upon that information.

‘Theory’. Implies that the premises are derived, or deduced from an axiom. The resulting premises if logically consistent, then have to be true, each deduction leading to more complex and less obvious truths. Mathematics is the classic example.

A ‘hypothesis’ however is a premise, made through observations, and then holding other variables constant, repeatedly testing that premise. ‘Dow Theory’ is no theory at all, it is a ‘hypothesis’ and as such falls towards Popper’s observation, assumed true until proven false, of which you only need 1 example. If then there is an example of ‘Dow Theory’ non-confirmation within the history of the two averages, the ‘hypothesis’ fails.