StockFetcher Forums · Filter Exchange · setting up the most accurate graph to predict stock movement<< 1 2 3 >>Post Follow-up
porti
20 posts
msg #82607
Ignore porti
11/6/2009 6:20:57 PM

Question: Does this mean the SPY is half way, or is it hitting the resistance line now?





porti
20 posts
msg #82608
Ignore porti
11/6/2009 6:27:31 PM

To clarfify myself, the divergence MACD EMA 9 / MACD 12,26 hasn't crossed 0 till now. Thats makes me thinking that we are only at half.

Is this correct? Please dont ban me ... :)

tmanbone
124 posts
msg #82609
Ignore tmanbone
11/6/2009 7:07:16 PM

Nice, Thanks all.

chetron
2,817 posts
msg #82613
Ignore chetron
modified
11/6/2009 9:23:15 PM

spy looks like a head and shoulders.


edit:nov. 19

Kevin_in_GA
4,553 posts
msg #82620
Ignore Kevin_in_GA
modified
11/7/2009 8:38:52 AM

OK, at some point everyone should try to write a filter that picks market tops and bottoms, if for no other reason than to see how hard it is.

Here's my thinking on how I would do it. By the way, I like simple filters over more complex ones whenever possible - too many lines or divergences to keep track of leads me to decision paralysis.

Fetcher[
/*KEVIN'S ATTEMPT TO PICK MARKET TOPS AND BOTTOM'S*/

set{trigger, slope of Linear regression(10) indicator}

draw trigger line at 0
draw LRI(10)
apply to symlist(spy)

draw rsi(2) on plot weekly rsi(2)

and chart-time is 6 months
]



I have been intrigued by the linear regression indicator for some time, seeing it as more responsive than an EMA (which it is), but not really sure how I would use it.

First thought was to buy when the price crossed above, and sell when the price crossed back below - only modest predictive accuracy, and a lot of buying and selling as the price fluctuated around the LRI line.

Second was to use two timeframes, a 10 day and a 50 day, and buy and sell at the crosses - again profitable, but it acts more like "responding" than "predicting". You are several days late on the market turns both in and out.

Looking at the LRI(10), what struck me was that the slope change - from positive to negative or negative to positive - coincided in many cases with a short term top or bottom. The above filter uses that slope change as a trigger for buying or selling.

The choice of 10 days as the time frame was arbitrary, and other time frames might be a little better, but 10 days is a simple round number that is fast enough to catch turns but not so fast that you get a lot of whipsaw action.

I included M4M's RSI trick as well so that the two can be compared as to accuracy and whipsaws.

Thoughts?

the_trade123
107 posts
msg #82625
Ignore the_trade123
11/7/2009 1:23:26 PM

Look at INTC. Looks like a very good candidate. How different is this from the "BOLLINGER BAND TRADING STRATEGY"? Isn't It simple to see that farther away from the mean BB, more the force to come back towards it. You may call it, rubber band, spring etc...essentially what you want is a confirmed uptrending stock with improving MACD and CCI.

jnafach
74 posts
msg #82630
Ignore jnafach
11/7/2009 7:07:22 PM

I am not sure we can use rb, in march rb was 0.85 range and same for november lows

trendscanner
265 posts
msg #82639
Ignore trendscanner
modified
11/8/2009 8:56:51 AM

Porti, just for clarification, the divergence I was referring to is not the MACD divergence line crossing above 0 in the chart below. The divergence I was referring to is the divergence between slope of the two sequential lows made by price (a negative slope) and the positive slope of the corresponding sequential RSI and MACD lows. Easier to see on the chart below.

That is the divergence that I believe M4M is referring to, as he has mentioned that type of diveregence on his previous posts. When price is making new lows but key indicators (MACD, RSI) are not confirming those lows, it's a strong TA signal that a reversal in price may be coming and is a fairly reliable signal to trade.

Kevin, you're bottom filter is pretty good too. I think bottoms are easier to sense than tops, since bottoms tend to be sharper, while tops can roll over or eke out new highs for longer periods.






mystiq
650 posts
msg #82640
Ignore mystiq
modified
11/8/2009 9:37:17 AM

Kevin_in_GA
- Ignore Kevin_in_GA
modified 11/7/2009 8:38:52 AM

OK, at some point everyone should try to write a filter that picks market tops and bottoms, if for no other reason than to see how hard it is.

Here's my thinking on how I would do it. By the way, I like simple filters over more complex ones whenever possible - too many lines or divergences to keep track of leads me to decision paralysis.


Fetcher[
/*KEVIN'S ATTEMPT TO PICK MARKET TOPS AND BOTTOM'S*/

set{trigger, slope of Linear regression(10) indicator}

draw trigger line at 0
draw LRI(10)
apply to symlist(spy)

draw rsi(2) on plot weekly rsi(2)

and chart-time is 6 months
]




I have been intrigued by the linear regression indicator for some time, seeing it as more responsive than an EMA (which it is), but not really sure how I would use it.

First thought was to buy when the price crossed above, and sell when the price crossed back below - only modest predictive accuracy, and a lot of buying and selling as the price fluctuated around the LRI line.

Second was to use two timeframes, a 10 day and a 50 day, and buy and sell at the crosses - again profitable, but it acts more like "responding" than "predicting". You are several days late on the market turns both in and out.

Looking at the LRI(10), what struck me was that the slope change - from positive to negative or negative to positive - coincided in many cases with a short term top or bottom. The above filter uses that slope change as a trigger for buying or selling.

The choice of 10 days as the time frame was arbitrary, and other time frames might be a little better, but 10 days is a simple round number that is fast enough to catch turns but not so fast that you get a lot of whipsaw action.

I included M4M's RSI trick as well so that the two can be compared as to accuracy and whipsaws.

Thoughts?

**great indicator**
/*KEVIN'S ATTEMPT TO PICK MARKET TOPS AND BOTTOM'S with MACD*/
Fetcher[
set{trigger, slope of Linear regression(10) indicator}
draw macd(3,16,10) histogram on plot trigger line at 0
draw LRI(10)
apply to symlist(spy)

draw rsi(2) on plot weekly rsi(2)

and chart-time is 6 months
]




porti
20 posts
msg #82643
Ignore porti
11/8/2009 12:25:48 PM

Thank you for the explanation!

StockFetcher Forums · Filter Exchange · setting up the most accurate graph to predict stock movement<< 1 2 3 >>Post Follow-up

*** Disclaimer *** StockFetcher.com does not endorse or suggest any of the securities which are returned in any of the searches or filters. They are provided purely for informational and research purposes. StockFetcher.com does not recommend particular securities. StockFetcher.com, Vestyl Software, L.L.C. and involved content providers shall not be liable for any errors or delays in the content, or for any actions taken based on the content.


Copyright 2016 - Vestyl Software L.L.C.Terms of Service | License | Questions or comments? Contact Us
EOD Data sources: DDFPlus & CSI Data Quotes delayed during active market hours. Delay times are at least 15 mins for NASDAQ, 20 mins for NYSE and Amex. Delayed intraday data provided by DDFPlus