How Our Engines Work
Full transparency into the formulas, factors, and assumptions behind every stock recommendation. No black boxes.
Short-term (intraday/swing). Technical indicators. 3 sessions/day. Top 20 ranked picks.
6–24 month holds. Fundamental analysis. Quarterly review cadence. Compounder stocks.
Engine 1: Stock Intelligence Scoring
Every stock is scored 0–100 using 14 weighted technical factors. Scores ≥65 trigger a BUY, 36–64 = HOLD, ≤35 = SELL/AVOID. The engine runs up to 3 times daily (morning, midday, close sessions).
Scoring Factors
Price Target Formulas
Entry just below the current market price for a clean fill. Avoids chasing spikes at the exact ask.
ATR measures a stock's typical daily price swing. A high ATR (e.g. NVDA ~$7) means wider stops/targets. A low ATR (e.g. KO ~$0.60) means tighter stops/targets. This normalizes risk across all volatility types.
3× ATR above entry = 3:1 reward-to-risk. Example: Buy NVDA at $900, ATR=$7 → Target = $921. Always maintains 2:1 minimum R:R regardless of the stock's volatility.
1.5× ATR below entry. Example: Buy NVDA at $900, ATR=$7 → Stop = $889.50 (~1.2% stop). Buy KO at $60, ATR=$0.60 → Stop = $59.10 (~1.5% stop). Adapts to each stock's volatility profile.
Reward:Risk Ratio: Our ATR model always targets a minimum 2:1 R:R. This means even if only 40% of trades are winners, the portfolio can still be net profitable — because wins are at least 2× the size of losses.
Engine 2: Long-Term Fundamental Scoring
Scores 0–100 based on fundamental quality indicators. Designed for 6–24 month holds. Quarterly review is recommended. Strong Buy ≥75, Buy 60–74, Hold 36–59, Avoid ≤35.
Fundamental Factors
12-Month Price Target Formula
Uses the stock's own 1-year historical growth rate as the baseline, adjusted by sector tailwind multiplier. Example: Stock grew 25% last year, sector tailwind = 1.1× → Target = Price × 1.25 × 1.1 = +37.5% implied upside. This is an estimate, not a guarantee.
Simple percentage gain from current price to the 12-month target. Used to rank and compare picks. A positive upside does not guarantee the target will be reached.
Portfolio Projection Model
The projection banners shown in the dashboard use specific assumptions. These are mathematical models, not forecasts.
Stock Intelligence Projection Assumptions
At 55% win rate and 2:1 R:R: EV = (0.55 × R) − (0.45 × 0.5R) = 0.325R. This positive expected value is the mathematical basis for the compounding projection. EV is capped at 0.8%/session to prevent unrealistic projections.
Long-Term Projection Assumptions
Long-term projections use a straight-line path to the 12-month target (not compound growth), which is more conservative and realistic for buy-and-hold positions.
Honest Limitations
The engine has not been backtested across multiple full market cycles (bull, bear, sideways, crash). The RecommendationOutcome entity is accumulating real data — win rates shown are based on actual outcomes as they build up, not historical simulation.
Real trades rarely fill at exactly the buy target price. Slippage (filling above target) reduces actual returns vs. projected. In fast-moving markets, slippage can be 0.1–0.5%+ per trade.
Projection models assume zero commissions. Frequent trading (3 sessions/day × 20 picks) with per-trade fees would meaningfully reduce net returns. Factor in your broker's fee structure.
Earnings surprises, macro shocks, geopolitical events, and Fed announcements can invalidate all technical signals instantly. No model fully captures these tail risks.
A strategy with strong signals in a trending bull market may underperform in choppy or bear markets. Indicator effectiveness changes with market regime.
Knowing the right trade and executing it consistently are different challenges. Emotional discipline in following stops and targets is entirely the user's responsibility.
Forward-Looking vs Historical Performance Lists
A common source of confusion: financial media publishes "Best Performing Stocks" lists that rank stocks by past returns (e.g., SNDK returned 3,185% in 2026). Our engine works differently.
- • Shows stocks that already delivered 20–3000% returns
- • Useful for understanding past winners
- • But: buying at the peak after the move is done
- • Example: SNDK at 3,185% gain in 2026 → probably already overextended
- • Scans for current signal alignment
- • Identifies entry points BEFORE major moves
- • Daily picks: technical momentum & sector strength today
- • Long-term picks: reasonable valuations with solid fundamentals
Why the Difference Matters
Media lists are rear-view mirrors. When SNDK appears on a "best performers" list in May 2026 (having already gained 3,000%), buying it at that price means:
— You've already missed most of the move
— Risk/reward is inverted (limited upside, large downside if it corrects)
— You're buying into exhaustion, not momentum
Our system aims to buy the SNDK of tomorrow at today's prices. We look for:
— Stocks showing early technical strength (RSI improving, volume surge, sector tailwind) before the masses notice
— Solid fundamentals at reasonable valuations before the Wall Street upgrade tsunami
— Entry points where risk is controlled and reward potential is high
Different time horizon = different list. A stock might be:
— A strong technical entry today (Stock Intelligence pick)
— But fundamentally overextended on a 12-month horizon (Hold or Avoid on Long-Term)
— Or vice versa: cheap fundamentals with poor near-term technicals
The goal: Be an early participant in the next SNDK move, not a late buyer after it's already delivered 3,185%. This is why we focus on current signals and relative entry value, not historical rankings.
Methodology FAQ
This methodology document is for educational and transparency purposes only. Nothing on this platform constitutes financial advice, investment advice, or a solicitation to buy or sell securities. Past performance and projected returns are not indicative of future results. All investing involves risk, including the potential loss of principal. Read our full Terms & Conditions →