Hope Is Not A Valid Setup – October 19, 2011

[Billy]

I want to add a public confession here.

I have been faithful to the robot because (this, above other reasons) it provides very precise entries and exits. Often I find the proffered opening price so extreme; I cannot believe my orders will get filled. Usually they do.

Only one time did I find the robot out of sync, and this sent Pascal and Billy into a deep think (as chess players say) in search of improvements. That was at the beginning of Augst. the robot went strong long and Dr. K went short. Of course, the short position was hugely profitable. Perhaps as large as the silver run up in March and April. But I missed it.

Today, I tried a small position in IWM related ETFs in compliance with Dr. K's model. Again, the weakness is lack of precision. I took the worst loss in the last month today. But it is under control. No problem. Part of trading. Still, I'm faithful to the robot because it offers precision where the MDM of Dr. K does not.

Nickola,
Dr. K’s MDM has a fantastic and longer proven track record than the robots. But, as it is much emphasized by Dr. K himself, there is a discretionary component in his MDM that is totally absent from the robots. One can of course wonder how does one backtest discretionary components? Sometimes, the discretionary component can change overnight like it did Wednesday morning switching from a confirmed cash (despite a qualifying FTD) after the close to a new buy signal at the open. Dr. K does manage the risk with a fail-safe stop that is usually 3-5% away from his entries, so failing signals seldom lead to significant drawdowns. And the gains on winning signals are usually fairly large once or twice a year to more than compensate for the small drawdowns.
Because the robots do use volatility-based stops, current high volatility would force them to enter new positions with initial stops 8-10% away from entry. In case of a failing signal, the loss/drawdown could be twice as high as Dr. K’s MDM.
In fact, timeframes used are also much different. The robots look at a combination of 20-day money flow with 3-days and 10-days strategies statistics to make up their directional choices, while Dr. K’s MDM is based on multi-week and multi-months price, volume and leadership analysis.
Dr. K enters and exits at market price when a new signal is issued, while the IWM robot only enters at optimal reward-risk price levels derived from the multi-pivot methodology. If it is missed, the robot stays in cash. Hence the reason you find the entry prices so “extreme” but they are indeed often filled, and at the best reward-risk setup, which compensates somehow for the wider initial stop. The robots also exit often at a trailing stop price, while there is no official trailing stop with Dr.K.

Actually, we have backtested the multi-pivot methodology with Dr. K’s MDM signals and it was very disappointing. That’s probably why Dr. K who was the first one (before Pascal) to kindly backtest my multi-pivot methodology with very positive conclusions is not using it with his model.
So, Dr K’s model and the robots are completely different animals. We are not in competition because they fill very different needs, investing/trading philosophies and timeframes. It is sound and normal that they are sometimes in clear contradiction one with another. Both Dr. K and Pascal are trading geniuses IMHO and their services both produce some of the best long term risk-adjusted returns on the planet.
Billy