Predicting The Trend for SP100 and SP500 ETFs : SP100&500 ETFs - A 30 Day Look Ahead Using 2 Moving Average Cross-overs.

Friday, October 12, 2018

SP100&500 ETFs - A 30 Day Look Ahead Using 2 Moving Average Cross-overs.

New Algorithm using the WEKA RandomForest classifier
Best moving average method: TA-Lib's T3
Fast: 14
Slow: 27
Weighted F Measure: .5733
Average Correctness: .7413

	2018-08-30	2018-08-31	2018-09-04	2018-09-05	2018-09-06	2018-09-07	2018-09-10
Symbol	2018-10-12	2018-10-13	2018-10-17	2018-10-18	2018-10-19	2018-10-20	2018-10-23
OEF	Strong sell	Strong buy	Strong buy	buy	Strong buy	Strong buy	Strong buy
QQQ	sell	Strong buy	Strong buy	sell	Strong buy	Strong buy	Strong buy
TQQQ	buy	buy	Strong buy	buy	Strong buy	buy	Strong buy
SQQQ	Strong buy	Strong sell	Strong sell	sell	Strong sell	Strong buy	sell
QLD	Strong buy	Strong buy	Strong buy	Strong buy	buy	buy	buy
QID	Strong sell	Strong sell	buy	sell	sell	Strong buy	sell
SPY	Strong sell	Strong buy	Strong buy	buy	Strong buy	Strong buy	Strong buy
SPXL	buy	Strong buy	Strong buy	Strong buy	buy	Strong buy	Strong buy
SPXS	sell	sell	Strong sell	sell	sell	buy	Strong sell
UPRO	buy	Strong buy	Strong buy	Strong buy	buy	Strong buy	Strong buy
SPXU	Strong sell	Strong sell	Strong sell	sell	Strong sell	sell	Strong buy

No comments:

Post a Comment

Notes:

Follow my video blog Using WEKA, TA-Lib, and A.I. to estimate stock price trends

Combining the results from 15 different classification models for SP100 and 500 related ETFs - the results say we should expect continued "uptudiness" .Why? inverse (bearish) ETFs are in a cash position. While most bullish related ETFs are in a stock position.

The bullish/bearishness indicator is for the general market. So while an inverse type ETF may show a "go to stock" high percentage I look at this as a sell the market call. And vice versa if an inverse ETF has a high percentage in "go to cash" then it indicates a general market move to the upside.

Models: Best Moving Average, Double Moving Average, Triple Moving Average, MACD, RSI, MACD & RSI, Bollinger Bands, CCI, Derivative Oscillator, Stochastic, Double Stochastic, Ultimate Oscillator and Percentage Change Moving Average.

Attributes are mostly based on cross-overs.
Expected results used either a 3 or 4 day look-ahead.
All intervals for the models are optimized based on the best return for the past 8 months.

Was cash/stocks were models position at previous day.

To cash/stocks models expected next day position.