Technical Analysis Software Project
Back in April this year I wrote a Technical Analysis Software package based on spectral analysis, not Hurst Cycles. I used a spreadsheet to project the group of dominant signals. I got bored with the software and set it aside until this week. When I loaded all that market data I got quite the surprise when the projections of the model were extremely accurate to what actually happened. The model caught the major trend, but not the October dip. I wouldn't expect it to.
To the point, I recalculated the model projections. The trajectory shows nearly flat activity / top within the next 30 trading days, and heading lower over the remaining 140 days. There is no more long term upside potential. Since the SPX is above the trajectory, and expected to fall below it, now is an excellent time to short or go double inverse fund.
This model has a 90% chance of calling a direction change resulting in a market move greater than 20%. So this model is very useful for intermediate term buy-and-hold. I would need months of 1 minute data to be able to begin studies to see if the algorithms are useful for day-trading.
