Short and Long Trade Filter for Finding Price Extreme PivotsOnly short when shorter moving average (80 SMA) is over the upper thick bollinger band (400 SMA square root of 2 standard deviation band). Only long when it's below the lower.
The actual entry trigger is one of six things (in order of most importance):
1.) test of major horizontal resistance line
2.) break of a pattern that supports the bias (e.g. M pattern for short; W pattern for long)
3.) test of an hourly, 4 hour, daily, monthly, or yearly pivot (classical pivots) level: S1,S2,S3,R1,R2,R3, etc.
4.) test of a major fibs extension level (1.414, 1.5, 1.618, 2.0, 2.618, 3.0, etc.)
5.) RSI divergence AND a favorable MACD cross
6.) a test of one of the bands under (for long) or over (for short) the thick band - this is a really aggressive entry when used by itself.
The more of #'s 1-6 that happen simultaneously, the higher the probability that price will pivot favorably from that level and the easier it becomes to decide on a logical stop loss (price patterns and horizontal S/R levels give structure to base stop losses from).
This set-up is especially strong when the 400 SMA bands are squeezing. The theory is that the price should, more often than not, explode out of the squeeze on the side that the 80 SMA is in relationship to the 400, or in the direction that the 400 SMA is moving at the time of the squeeze, UNLESS the 80 is near the thick band (within 10% distance between it, the thick band and the 400 SMA). If the 400 bands are squeezing and the 80 SMA is near to touching, or moving outside of, one of the thick bands, then price is more likely to shoot out of the squeeze in the opposite direction.
So this set-up is best used for two things:
1.) finding price extremes, from which price is likely to pivot into decent gains as a counter-trend play, if not getting us in at the start of a new trend altogether.
2.) increasing the odds that we get in on the right side during a long-term price range (400 SMA bollinger band squeeze), on fast moves with high percent gains.
I've only tested this set-up on shorter time frames for SPY, with decent results in back-testing the idea. Whether or not it is useful at longer time-frames and other stocks, forex, etc., is yet to be seen.
