November 13, 2011

Over the past few weeks, I have presented charts and portfolio tests going from the almost totally random entries to using administrative investment policies. Some of the tests using my trading methods showed extraordinary performance levels; way beyond what is estimated as statistically possible in academic literature or elsewhere.

I showed that you could use Schachermayer's payoff matrix to scale your expected total profits to your desired level simply by adding more capital in order to enable you to increase your bet size. You could, for instance, transform Σ(Q.*ΔP) into Σ(2Q.*ΔP) which would double your total generated portfolio profits just by doubling your initial capital. There are other methods that lead to increased profits as well, increasing your bet size is not the only way available.

Some go on the predicate that: technical analysis fails to work, period. And that a lot of statistical evidence is there to prove it. That technical analysis is not better than betting randomly on price movements and that overall there is no expected gains on what every technical trading procedures you may develop.

Well, I disagree. Anyone and I mean anyone, can develop superior trading techniques using technical analysis or, even as demonstrated in the past few weeks, using almost totally random entries. Whatever trading method you want to design is included in the Schachermayer pay-off matrix: it has only two variables and one of which you can not alter. You are left with designing a holding matrix which is just at most an inventory problem, where your fluctuating inventory level (your position sizing algorithm) is the only thing that really matters.

To demonstrate my point of view, I decided to use a published script that uses technical analysis extensively. My choice went to the Hershey Equities Chartscript v 5.0 which was last modified in 2008 after having some 4 years of continued tweaking and improvements by their authors. It is an elaborate trading system entirely based on technical analysis. My intention is to modify this script, have it stop shooting itself in the foot, change its trading philosophy, alter its look at price movements, change its inventory management system, add more efficient stop losses and finally modify most of its functions to better reflect my mental statistical views of the market.

There is always a need to compare. Therefore, I have done the preliminary test which is to have the original Hershey V 5.0 script go over a portfolio. I chose dataset 2 and I will do my modifications to the same script to be used on the same data set. My objective is to report back when I have reached the 100M profit level using only technical analysis as decision surrogates.

And then, to push even further, I will do another test, change the dataset and show that profitability was maintained even if the script has never seen any of the stocks in the new data set. This is a tall order. If I do not succeed in achieving the 100M in profits there will be a need to address the degree of success depending on the results. And if I do succeed, then there is some re-thinking that needs to be done. Maybe technical analysis has some value after all. So, please, allow me a few days before I report back.

The Hershey original script is no longer available. It was free on the old Wealth-Lab 4 website.

The summary results using the original Hershey V 5.0 follows: 

Original Hershey Equities Chartscript v 5.0 Script

Original Hershey V5.0

(click to enlarge) 
Based on the above results, the original version of the Hershey script does not provide results that are significantly different from a random entry system as was presented previously. Therefore, these results corroborate again the hypothesis that technical analysis has little value. So, I do have a lot of work to do. I would say wish me luck, but I already know the outcome.

(click to enlarge) 

AAU AKAM ARUN ASYS
ATML BIDU CAM CAT
COOL ETN FFIV FIRE
GMCR HK HNL IDCC
IGTE LTXC LULU MELI
MENT MFL MGH MSN
NDSN PFCB PNRA PTI
QCOR QLTY REDF RVBD
SCSS SF SFLY SHS
SPRD SVVS TLEO TPX
UA UTEK VSEA  

( click to enlarge)


Created on ... November 13, 2011,    © Guy R. Fleury. All rights reserved.