Frequently Asked Questions

Dr K Market Direction Model
You often speak about the MDM as if it's some robotic, non emotional thing that makes it's decision based on x,y,z factors.

Q: You often speak about the MDM as if it's some robotic, non emotional "thing" that makes it's decision based on x,y,z factors. But are you, yourself, making the buy, cash, sell calls? In this way, it has been my assumption that part of the model is discretionary, but please clarify. Also, if you are in fact the person who decides what action the MDM takes on a discretionary basis, how can you back test the signal since the 1970's since the MDM didn't exist then? How would you know what actions you would have taken? As an example the MDM switched to a buy signal on 10/5/11 based on the "wash out". Did you go back and determine what days the market "washed out" and then assumed you would have had a buy signal?


A: I use a set of statistically based rules to determine the model's signal (buy, cash, or sell). I performed backtesting using the price/volume action of leading stocks and major indices going back consecutively to 1974 as well as spot checked various years in the 1920s and 1930s to make sure the model continued to well outperform the market (the Dow Jones Industrial Average back then) in entirely different eras. I also accounted for any material change that stayed with the market for more than a few months. One could say this is discretionary on my part, though I need to see a material change that endures for at least a few months and looks to stay for fundamental reasons if I am going to introduce a new systematic rule into the model. In recent times, it has been quantitative easing in early 2009 that was a material change for the markets.

Such new rules occur very rarely. The model is not a black box but remains flexible in terms of the rare material change that may occur. In the late 1990s, it was the advent of high performing internet stocks with no earnings, which meant a leading stock could have no earnings but had to have substantially accelerating sales.


As for the buy signal on 10/5/11, that was based on prior instances when the market had a steep correction then behaved in the manner it did, thus pushing it into a buy signal.

Here are other FAQs which should round out the answer:

Published: Nov 29 2011