Bulletin 53 What Do We Know For Sure

The Traderclub Forum: Traders Club Bulletins: Bulletin 53 What Do We Know For Sure
By
webmaster (Admin) on Sunday, February 23, 2003 - 11:45 am:

BULLETIN #53

What Do We Know For Sure?


By Chuck LeBeau

While chatting with one of our members recently I casually remarked that I had often observed that markets tend to go down much faster than they go up. He suggested that it would be a good idea to test that popular assumption. He then proceeded to run a few simple tests on a broad portfolio of futures markets using various length moving averages and reported back that his results did not support the conclusion that markets go down faster than they go up. His conclusion was that most futures markets were surprisingly symmetrical and his study indicated that they actually went up and down at approximately the same rate. (In fact I think his study showed that the grain markets tended to go up faster than they went down.)

Although I am a bit wary of the moving averages testing method used in this particular test our very interesting discussion got me to thinking about how little we know for sure about the nature of the markets we trade. We tend to take a lot of information at face value and build systems around critical assumptions that may have very little basis in fact. What do we really know about the markets we trade? Probably not nearly as much as we think we know.

I think that this Forum would be the ideal place to start a discussion about what we know "for sure" about the nature of the markets. What simple statements about the markets can we actually prove? For example, I recall having seen an academic study many years ago that showed that futures markets do have big trends (particularly on the up side). The professors who did the study offered proof in the form of a study of the distribution of prices which produced an abnormally thick tail on the positive side of the curve. I was impressed that someone had offered some statistical proof of an assumption that futures traders have been relying on ever since I can remember. Trend following works because it is based on a valid and provable assumption. (If any members can recall or retrieve this study and post it here I would greatly appreciate the help.)

Now I would propose that we use the knowledge and resources of our members to make a list of what we know "for sure". There are members of this group who are well disciplined in statistical testing methods that might be able to prove or disprove any assumptions we propose.

I think we need three groups to do this research effectively. The first group can offer simple statements of what they believe to be true about the nature of the markets. This group does not have to be very experienced or have any testing skills or software. They will simply ask important questions or make a statement and throw it on the table for discussion and research. The second group will need to have programming skills and access to testing software so they can try to prove or disprove each statement. The third group would evaluate the testing procedures to see if the research was done in a manner that offered conclusive proof. We need to be careful and take a hard look at the "proof" that might be offered to make sure it really does prove what we are testing. The researchers will have to be willing to submit tests that may be rejected as flawed but we will all be indebted to them for their hard work. Eventually we will begin to accumulate a valuable list of truths about the markets.

I will offer a couple of statements that might serve to get things started. I'm hoping that some of our more knowledgeable and talented members will try to figure out how to test these assumptions and post their findings here. We should prepare to have the results challenged and discussed.

1. Prices go down faster than they go up. (Suggestion on testing - look at price excursions in both directions and measure slope or rate of change.)
2. Prices are serially correlated over the short run. For example: Up moves are followed by more up moves. Down moves are followed by further down moves.
3. Prices tend to return to the mean. (Are statements 2 and 3 mutually exclusive?)

Right now our list of what we know "for sure" is empty. Let's see how hard it is to expand the list. Please offer additional suggestions on what we think we know about the markets.

I'm hoping that this topic "What Do We Know For Sure" will generate some lively and productive discussions on our FORUM. Please join us and contribute whatever knowledge you can at http://www.traderclub.com/discus/board.html


Good luck and good trading.


* * * * * * * *
Special Discount to System Trader's Club members ONLY

Giles Korzenecki is a good friend of mine and a good system developer. His bond system, "US one" was rated number one in the country last year. (It was up 167%!!!) As a special to the System Trader's Club, he is offering his system to members for 1/2 off the regular price.

For details please contact him at 888-281-9566.

Giles Korzenecki
MarketDirection.net
312-528-3378
888-281-9566


* * * * * * * *

Note from Chuck LeBeau: We still have a limited supply of our Video Workshop packages which contain more than 5 hours of video instruction, a workbook and a simple ADX system for stock trading. Total price is $250. ($285 for non-U.S. orders.) For further details Go to www.traderclub.com/video.htm We have received many wonderful testimonials about these videos. I can assure you that our members have enjoyed these videos and so will you.

* * * * * * * * *