Predicting with certainty which specific stock “will go up today” is impossible. What we can do is filter for stocks that, based on historical patterns and current signals, have a higher probability of rising in the next session. The selected filters are designed exactly for that.
Screening Filters
Price: 2–200 USD
- Purpose: Focus on reasonably priced, tradeable stocks and avoid extreme penny stocks or ultra-expensive names.
- Rationale:
- Stocks under $2 often have very low liquidity and can be highly manipulated or subject to extreme volatility that doesn’t reflect real demand.
- Very high-priced stocks (above $200) might move in smaller percentage terms or be less accessible for smaller accounts.
- This range targets a broad universe of “normal” stocks where price action and signals are more meaningful for short-term trading.
Monthly Average Dollar Volume: ≥ 100,000 USD
- Purpose: Ensure the stocks are sufficiently liquid.
- Rationale:
- Dollar volume = price × volume, so this filter requires that at least ~$100k of stock changes hands per day on average over the last month.
- Better liquidity typically leads to tighter bid–ask spreads and more reliable execution at or near the current price.
- Illiquid stocks can be hard to enter/exit and their intraday moves may be “noisy” and unreliable for any prediction.
Moving Average Relationship: PriceAboveMA5
- Purpose: Capture short-term upward momentum.
- Rationale:
- The 5-day moving average (MA5) is a very short-term trend indicator. Price trading above its 5-day average suggests recent strength.
- When the current price is above MA5, the stock is in a short-term uptrend, which is statistically more favorable if you’re looking for a one-day gain than stocks in a downtrend.
- This filter aims to pick names that already have positive near-term momentum rather than attempting to catch falling knives.
One-Day Rise Probability: ≥ 52%
- Purpose: Select stocks with a model-estimated probability of going up tomorrow that is better than a coin flip.
- Rationale:
- A 52%+ probability threshold means we only keep stocks where a predictive model (likely based on historical price behavior, volume, volatility, and technical/other features) believes there is at least a slightly favorable edge for an up day.
- It doesn’t guarantee a gain, but combined over many trades, such an edge is what traders look for.
One-Day Predicted Return: ≥ 0.2
- Purpose: Ensure that the expected size of the move is meaningful, not just barely positive.
- Rationale:
- This value is likely in percentage terms (e.g., ≥ 0.2% expected return for the next day) or model units translating to a small positive move.
- A stock with a tiny expected gain, even if likely, might not be attractive after considering trading costs, spreads, and risk.
- By requiring a minimum predicted return, we focus on candidates with both a higher chance of going up and a non-trivial expected upside.
Why Results Match Your Question
- You asked for stocks that “will go up today.” While we can’t guarantee outcomes, these filters:
- Emphasize short-term upward trends (PriceAboveMA5).
- Use a probabilistic model to favor stocks with a >50% chance of rising (one_day_rise_prob ≥ 52).
- Require a positive and meaningful expected one-day gain (one_day_predict_return ≥ 0.2).
- At the same time, they:
- Restrict to tradable, liquid stocks (price 2–200, dollar volume ≥ 100k) so signals and execution are more reliable.
Overall, the filters are aimed at narrowing the universe down to liquid stocks that are already showing short-term strength and that a model suggests have a slightly better-than-average chance of rising today, with some expected upside in the move.
This list is generated based on data from one or more third party data providers. It is provided for informational purposes only by Intellectia.AI, and is not investment advice or a recommendation. Intellectia does not make any warranty or guarantee relating to the accuracy, timeliness or completeness of any third-party information, and the provision of this information does not constitute a recommendation.