On trading algorithms.

No new trades. Sitting on FIG instead of overtrading my gains. New semester, new classes, and I would like the have the morning free before I commit the time to researching some risers for my new strategy, commented on by MUDDY in my last post.

I lately had two interesting conversations regarding the only studies I use while trading. My indicators for entering a trade have almost always been a combinations of experience enriched price movement and the trusty 10/60 day moving average. I will note to look into bollinger bands, per Muddy's note, as an attempt to diversify my technical indicators. Most people also find despair, yes despair, in the fact that I hardly allow myself to find out what exactly a company does before entering a trade. When a known riser takes off premarket, all you need are those 3-5 ticker letters and quick fingers to get the trade in.

Anyway, I always describe the 10/60 day moving average study set to people as a set of ploted lines, with the 10 day moving average being sinusoidal WRT the 60 day moving average. Now recently I explained this to two people Amit and Steve, who both happen to be engineers. Despite different specialties they both immediately responded wanting to call the 10 day line a Fourier transform of price/time. Interesting. Amit went further to ask what I thought the number of cycles was. Great question. AS I pointed out after I talked with Steve, I think that answer maybe 3. That right there, is easily exploitable. Forget scans, complex algorithims, what if funds were distributed across a number of tickers and bought or sold(to go long or short) depending on real time 10/60 crosses and then bought/sold on the anticipated nodes for the ticker. According to my theory this would be about 5 solid trades per day/ticker. WOW. Worth looking into for its mere simplicity. Who knows, maybe this will all turn out to be a commonplace fact about how 10/60 really works, but I haven't ever heard anyone talk about it.

Where I take this realization about 10/60 to is the efficient market hypothesis. I now feel I have the necessary proof to battle this idea I always thought was a bit immature. It was developed before the internet, so for it to still be vigorously debated today shows just how applicable it still maybe. For me, early one stocks were not about news, and price of a stock was not as good of an indicator as a 5 minute plotted chart. News is said to be something unknowable in present but to affect fture price. In behavioral and technical trading indicators are known to those who choose to note them, and definitely affect future price. My belief is that read if you dare "news" has been created due to development in communications that allowed real time quotes and the online day trader to be birthed. Information spreads quick nowadays. Plus with the influx of traders trading securities today behavioral analysis is alive and well as fundamental. It is in my opinoin a separate field from techincal analysis too. There is no luck involved with patterned observations. I have yet to find the long run to prove my committment to how dead EMH may be, but I'm working on it. 

I find great problem in the assumptions EMH takes too. It is because a great number of unskilled traders will over react and under react that profits will be made. The decisions of these many people cannot be admitted to the "noise" category. I take the opposite view. 

Who knows. As I pointed in my last post. Maybe market makers take those observant folks and allow them the same profit potential as they themselves in an attempt to fool the foolhardy into losses. If this is the case, we are talking about a different type of stock market then many ever expected. But hey if the collapse of our economy can be traced back to the buying and selling of a single security, mortgages, who is to say I am wrong.

I am curious to hear the EMH view.