Monday, October 22, 2012

Back Testing 52 Week High Theory

MetaStock SPRS Series - Week 91 - TechniTrader® Stock Discussion for MetaStock Users - Back Testing 52 Week High Theory - October 22, 2012
By: Martha Stokes C.M.T.


I was a speaker at the Las Vegas MetaStock Users Conference on October 13th. During my training session a couple of important questions were asked by the attendees that I did not have time to answer to my satisfaction.

I am going to answer those questions here in this forum so that everyone can read the answers.

Question posed: There is a study out that used “Back Testing” to derive a theory about where the institutions are buying the heaviest. The theory was that since 52 week highs had higher volume, that this must be where institutions buy into stocks.

My answer to that question was no, not in the current market conditions with the current market structure. The person asking was bewildered as he believes that back testing is irrefutable evidence. But there is an obvious flaw in the theory most traders are unaware of, due to a lack of understanding that has occurred regarding the market structure in recent years.

What is immediately apparent to me as a cycle theorist, is a critical error in the original back testing data set.

An absolute rule in science and cycles is that in order for an accurate test of any theory, the data and conditions must be relatively consistent and the same throughout the entire set of data. If conditions alter during the data set used, then the testing method and theories will be skewed and inaccurate.

Let’s examine the Back Testing Method first to determine if indeed the back testing supplied an accurate set of data on which to make this assumptive theory. The question I am posing then is this, “Did the conditions of the stock market, the market participant groups, the number and type of groups, and how they bought and sold stocks remain constant for 40-50 years back in time from today?” A major change during the period of time, will cause the data to be skewed and the back test results inaccurate.

If we go back in time 50 years from the year 2012, we are starting the test data around 1962. In that decade, the Dow Theory of 3 market participants was still intact. The informed investors which are the investment banks, wealthy individuals, and corporations controlled about 50% of all the market activity. Mutual Funds were just beginning to be popular again, after the huge Mutual Fund debunking in the 1920’s when Mutual Funds were first invented and sold heavily.

In the 1970’s the markets were stalled in a wide range bound pattern going nowhere. Mutual Fund investing slowed, but the market participant groups remained intact as Charles Dow’s theory had stated.

There were no pension funds allowed in the market in the 1960’s and 70’s.

That means that 50% of all the activity in the market was the small lot investor and odd lot investor. Remember there were no day traders, no retail traders, no small funds, no HFTs, no institutional traders as there are today. Investing was mostly long term. The Dow Theory clearly defined that the informed investors which are the banks, corporations and wealthy individuals started the bull market, and the buying by the average and odd lot investors was toward the end of the bull market.

By the early 1980’s Pension Funds had finally won their battle with Congress and pension funds entered the stock market. This created the bull market of the 80’s and it changed the matrix of the market participant cycle. However these were all still long term investments with very limited and constrained short term trading for all fiduciary funds. That means all during the first 2 decades of the back testing for the theory was primarily long term investing either by a fund or by a small investor.

The ratio of 50% informed and 50% uninformed held into the 80’s, so for the first 2 ½ -3 decades the data set is reasonably accurate for back testing up to that time, but then in the late 80’s changes commenced.

In the late 80’s and early 90’s floor traders took PC computers and started trading against their market maker employers. They became the institutional and professional traders market participant group we know today. Their short term trading activity did not alter the balance much, perhaps 2-5% at most.

In 1990 there were still no online brokers, or short term retail trading. It was all professional floor traders who had access to the exchanges and understood how the intraday action worked. At this time the speculative activity of the markets increased to some extent. Small lot investors began to lose market share. Their activity dropped from 55% of the daily market activity to 45%. The shift of balance between the “informed” and “uninformed” had started. What once had been an evenly balanced market was now more professional activity than average investor activity. Volumes started to increase in 1994 as PC computers and online brokers became popular.

But it wasn’t until 1998, a pivotal year for the stock market structural changes, that the balance between the average investor and the professional aka “informed” side of the market shifted dramatically. In the Roth IRA Bill was a little known rider that eliminated the “Rule of 3” that had been in place since the pension funds had been allowed into the market. With the elimination of the “Rule of 3” and the deregulation of the banks merging commercial and investment banking, the entire market structure changed almost overnight.

Now the balance between the professional side and the average investor/retail trader shifted even more, with 70% of the market activity now professional and 30% average investor/retail trader. The high volumes of shares traded on stock charts that can be seen during the 1998-2000 are dramatic. In addition there are now 8 levels of Market Participants, far more than the three Dow identified.

By 2005 High Frequency Trading had begun adding the final 9th Market Participant. This encouraged more Dark Pool activity which is an off the exchange transaction by giant funds, the largest funds in the market. Dark Pool activity does not go through the exchanges but is done “Over the Counter.” These are the largest orders used by any market participant and this is another major change to the market conditions and structure. In doing so, the professional side of the market dominance increased to 80% while the average investor and retail trader activity dropped to 20%.

The initial analysis of the data set, therefore provides clear evidence that the data set included in the back testing theory of 52 week highs as a buy entry point for institutions is flawed. Conditions, market structure, and market participants change significantly during the data set period, which has skewed the statistics for this theory. The data set used is not reliable due to the major changes in the market structure from 1998-2012. These changes were not small or insignificant, these changes altered the entire market structure including, how orders were processed, and the volume activity on which the “Back Testing 52 Week High Theory” was based.

Trade wisely,

Martha Stokes, C.M.T.
Member of Market Technicians Association
Master Rated Technical Analyst: Decisions Unlimited, Inc.
Instructor and Developer of TechniTrader® Stock Market Courses
http://technitrader.com
MetaStock Partner

©2012 Decisions Unlimited, Inc.

Disclaimer: All statements, whether expressed verbally or in writing are the opinions of TechniTrader, its instructors and or employees, and are not to be construed as anything more than an opinion. Student/subscribers are responsible for making their own choices and decisions regarding all purchases or sales of stocks or issues. At no time is any stock or issue on any list written or sent to a student/subscriber by TechniTrader and its employees to be construed as a recommendation to buy or sell any stock or issue. TechniTrader is not a broker or an investment advisor it is strictly an educational service.

2 comments:

daisey said...

Thanks for share MyCCPay

dgmeeee said...

I'll be sure to keep an eye on this thread. Looking for the same issue. Bumped into your thread. Thanks for creating it. Looking forward for solution.

dollar general pay stub