how can something like this be expressed?

Originally posted by inandlong
I recognize the benefit of quantifying your observation, although relative to each stock, using a % rule would be accurate. But I know what you want. Too bad you can't say, "when the price of the stock is 1 inch from its 5 sma. :)
More...
i laughed when i first read that, but it just now occurred to me that you're exactly right.

i literally just took a ruler to my extremes.jpg image and it seems an extreme for that stock is about 3/4 of an inch (or greater) away from it's 5sma. lol
 
So the question remains "how do we quantify your observation?

Geez, and it's a full moon tonite too. I might as well just head back to my office and put on some coffee. I won't be sleeping anytime soon.

Which is why I still think about esignal international and IB for the DAX. If I'm awake, I might as well trade. :D

Because you know as well as I do...
 
Originally posted by inandlong
So the question remains "how do we quantify your observation?
More...
before that, i'd ask if my observation is worth quantifying? if it is, i'm willing to work on it.
 
Gordon,

Anything can be expressed mathematically in my opinion.

What you are after is probably variance study as applied to statistics and standard deviations.

If you are adjusting for every stock, then you could also have a volatility measurement and then base the price move on that volatility measurement to see how many sigmas it moves away from it.

Picture this scenerio (assume 1 minute ticks)

1
2
3
4
5

Obviously, the trend here is up, but each tick, the trend slows down from the previous tick -- even though it increases by the same amount each time (this is why LOG charts are better than LINEAR charts for percentage approximations).

Now, you could right a small sub-routine that determines the averaging moving volatility and then compare the most recent price change with that moving average and red-flag it if it exceeds your threshold parameters.

Unfortunately, I would never trade off any academic system because they all fail in the long run.

aphie
 
I think it is because it leads to at least two tradeable scenarios: one that is immediate and one that sets up.

Say your scan provides 5 candidates. Many times on those parobolic moves there are gaps from bar to bar, especially the shorter time frames. When the price stalls and takes out the previous bar's low - in the case of an upswing - you have a very real reason to jump in short.

If there are no gaps and the price stalls and starts to move sideways, now you have a great breakout setup forming for the next leg, whether up or down. The nice move from that breakout comes when it is very near the MA and either breaks out looking like MA support, or breaks down, crossing the MA. Either way, all the MA watchers will be getting in too.

All those MA traders augment our knowledge that...
 
Originally posted by aphexcoil
Gordon,

...you could also have a volatility measurement and then base the price move on that volatility measurement to see how many sigmas it moves away from it.
Now, you could right a small sub-routine that determines the averaging moving volatility and then compare the most recent price change with that moving average and red-flag it if it exceeds your threshold parameters.

More...

My point exactly :)



Originally posted by aphexcoil
Unfortunately, I would never trade off any academic system because they all fail in the long run.

aphie [/B]
More...

Most - maybe. All - I wouldn't say so....
 
You must log in or register to reply here.
Back
Top