Methodology
How the screening and stock-selection process works, step by step.
Trading framework
Three indicators: SMI + MACD + ATR. Risk per trade capped at 1% of capital. Stop-loss set at swing low − 1.5×ATR, trailing stop at highest high − 2.0×ATR.
Stock selection criteria
Eight filters a stock must pass before it makes the ranking.
The screener scans stocks listed on US and European exchanges. These are two large, liquid markets with reliable, accessible financial data, which allows thousands of stocks to be evaluated automatically and reliably every day.
Only stocks that grew both revenue and net income quarter-over-quarter move on to further analysis, with year-over-year EPS growth also checked. The goal is to catch companies that are actually growing, not just ones that look good on a chart.
The SMI (Stochastic Momentum Index) shows whether a stock is currently oversold or overbought and whether it’s starting to reverse. The screener looks for the moment the indicator crosses its signal line on the weekly chart — often an early sign of a new price move.
A signal on the weekly chart alone isn’t enough — the screener also checks whether a matching signal appears on the daily chart. When both timeframes agree, the odds that the signal is reliable go up.
A classic technical analysis method from the 1930s that looks at how large investors — funds, institutions — accumulate or distribute shares, based on price and volume behavior. The screener looks for signs of an accumulation phase, i.e. the point where „smart money” may just be starting to build a position.
A growth-stock selection methodology popularized by William O’Neil, based partly on current and annual financial results, relative strength versus the market, and demand for the shares. The screener uses selected elements of this method as an additional quality filter.
Instead of rigidly splitting stocks into „meets criteria” / „doesn’t meet criteria”, the Fuzzy filter assigns each stock a score reflecting how strong and reliable a given signal looks. This means „almost perfect” stocks aren’t automatically dropped — they rank lower instead of disappearing entirely.
Regardless of the entry signal, the screener also checks the monthly chart. If the long-term trend looks concerning, the stock gets an extra warning flag — even if the short-term signal is positive.
The role of AI
Automated market scanning and indicator calculations are performed by the screener’s code. Descriptive commentary and analysis for each stock is prepared with the assistance of AI, based on the data generated by the screener.
Strategy validation
Walk-forward testing (rolling window, non-overlapping test periods) and filtering out weak stocks before they’re allowed into the ranking.