Interesting Malcom,Originally Posted by charttv
This tend to support a few market cliches' like "Santa Claus Rally", "Sell in May and Go Away" etc.
Another interesting concept along these lines is described in Stuart Johnstons book "Trading options To Win". This guy thinks out of the square and is often hilarious to-boot.
Basically as one who like to write option premium on commodity futures, he introduces the concept of "non-seasonality"... basically where does the price tend not to go.
These tendencies show up more often than seasonality, because in addition to identifying quiet times of the year, there is a corresponding non-seasonal trade for every seasonal tendency.
For instance based on your work, writing XJO index puts in December appears to be an absolute no-brainer.
Just another way of looking at statistics that I found interesting.
Completely unconvincing. I have already posted a statistical analysis of this yet you blather on without a single p-value in sight. Would I be right in guessing that you are trying to sell some sort of trading system, tips, coaching or whatever to the gullible? You'll probably do all right.Originally Posted by charttv
Thanks for the feedback Chemist. I can assure you that I have no hidden agenda and post on here because I wish to share what I think are worthwhile analytical approaches.
If it angers you so to see such a ham fisted approach to stat research please feel free to knock out tutorials on how it should be done so we can learn and apply your expert know how.
Please inform us of what tools and knowledge we require to do this correctly.
MalcolmOriginally Posted by charttv
Ham fisted works quite well in commodity markets. Seasonality has been around a long time, for good reason.
ChemistOriginally Posted by chemist
Im interested---what is your approach to trading?
Do you trade?
Kris my son is finishing his PHD in Physics specialising in Photonics.
The hardest thing I'm finding with Kris in teaching him trading is that he believes there is a mathamatical answer to everything.
How do you approach this?
What have you adopted?
I enjoy Malcolm's work.Originally Posted by chemist
If he has something to sell then I say good on him.
What we don't need is another whinging Australian with a tall poppy syndrome attitude to whatever is presented in front of them.
Oh, and the mathematics is a con for those who study it too much.
Discussion only! Posts may be factually incorrect due to ignorance, taken out of context, misinterpreted, or just opinionated discussion.
I just had a look at the above presentation.Originally Posted by charttv
The presentation itself is well done but to be honest I'm leaning towards chemist's view in that the results are not very convincing, but I wouldn't say it as bluntly as he/she did
The technique you use in your analysis appears to be sound but since you only go back to 1987, imo there isn't enough data there to provide convincing results......eg...I don't have your chart up anymore but from memory one month made highs 21% of the time but also made lows 16% of the time and I think you called that month a generally bullish month.
From my point of view, if I see a month making highs 21% of the time and lows 16% of the time there isn't a big enough difference in the 2 values for me personally to say whether the month is genarally a bullish or bearish month.
Perhaps if the dataset you use went back say at least 50 years then the results might be more conclusive than these appear to be for me at least.....but as I said earlier, your analysis technique looks sound to me, although I'm sure others like chemist could probably suggest techniques to further enhance the work you did in the above presentation.
Thanks for the food for thought...
Bullmarket - I agree! less than 20 years of data is a pretty small sample size. The only reason I started compiling stats like these is because I needed stats to back up my observations over the years.
Interestingly, there is data on the S&P 500 going back to 1950 and for the DJIA the data goes as far back as 1928. This data is available for free from finance.yahoo.com.
I should add that I am not a statistician.
Another way of showing seasonality and market high points,see figure 1 of Faber's report on the S&P, plot only particular months as different series: http://www.gloomboomdoom.com/marketc...ads/060605.pdf
My posts are not recommendations (even when I rave about something). Always rely on your own research & judgement.
I'm a huge Faber fan!
Sooooooo..... You're a bear Malcolm?Originally Posted by charttv
Depends what you're bearish on. Faber is bullish gold, oil and Asia as well as being bearish the US in general.
Sooooo..... you're bearish on western economies? I nead to know so I know whether to issue you with a "Bear Club" Badge.Originally Posted by charttv
It's OK for a bear to be bullish on gold and oil.
But to earn an "Uberbear badge", you have to be bearish on asia as well
Asian equities in the short term perhaps.
Wayne, I consider myself a trend follower but yes, it is hard to ignore the fundamental macroeconomic influences at work nowadays.
Quite! The wise bear is also pragmatic. Is it possible to be a bear and be long on everything? Absolutely! We bears ain't dumb! Only Salmon... and gamblers... swim against the current. (and commercial in the futures, but they eventually create the current)Originally Posted by charttv
But I must say, it has been fun being a bear for a few weeks.