BAC · Bank of America Corp.
Stock Universe · 全球股票市场High-Precision Structural Swing Capture: Analyzing the BAC 15-Minute Algorithmic Strategy
This quantitative research report evaluates an active algorithmic strategy for Bank of America stock (BAC) operating on a 15-minute execution interval while executing selective, extended position holding. Across a 7.02-year BACktested history from September 2019 to September 2026, the strategy generated a total cumulative return of +2430.20% and an annualized return of 135.23% over 36 completed trades. Featuring a high win rate of 83.33% and a profit factor of 3.75, the decision logic demonstrates rare entry characteristics, averaging just 5.13 trades per year. The strategy achieves equity expansion through multi-week swing capture, holding positions for an average of 1273.71 hours (approximately 53 days) while maintaining a strict maximum drawdown of 15.97%. With a DevioLab score of 70.02 and a rank of 3 for BAC, this quantitative model highlights how lower-frequency automated trading agents can utilize intraday candle resolution to optimize position timing without succumbing to over-trading or elevated portfolio churn.
分析日期： 22.09.2026. 策略数据单独更新。阅读完整分析
1. Strategy profile
The algorithmic strategy designated for BAC stock operates on a 15-minute execution interval, utilizing high-resolution market data to identify entry and exit triggers while executing a selective position-trading profile. Over a historical evaluation period spanning 7.02 years from September 16, 2019, to September 22, 2026, the model logged 36 completed trades. Despite the micro-structure interval of 15 minutes, the algorithm is far from high-frequency; it registers a modest trade frequency of 5.13 trades per year and 1.24 trades per active month. The primary objective of this quantitative model is to capture substantial structural trends in BAC equity rather than short-term intraday noise. Across the full sample, the strategy achieved a cumulative return of +2430.20%, translating to an annualized performance of 135.23%. The strategy holds a DevioLab score of 70.02 and ranks 3rd among evaluated models for the BAC asset. This performance profile indicates that automated trading bots deploying this logic act as patient trend-following agents, filtering out secondary market fluctuations to focus exclusively on macro-level price expansions.
2. Trading rhythm and position duration
The strategy exhibits a unique dual-timeframe dynamic, pairing intraday entry execution on 15-minute candles with extended holding horizons. The average position holding duration is 1273.71 hours, which translates to approximately 53.07 days, while the median holding duration stands at 612.00 hours, or roughly 25.50 days. This discrepancy between the mean and median holding times reveals that while half of the trades resolve within roughly one month, a segment of highly profitable positions is allowed to ride broader trends for several months. Exits are similarly sparse, with an average interval of 68.41 days between trade completions and a median spacing of 50.96 days. By restricting market engagement to roughly 5 completed positions per year, the model drastically reduces exposure to short-term market noise. For trading robots and automated execution algorithms, this low transaction cadence eliminates active portfolio churn and mitigates the execution risks associated with frequent order submission, allowing the trading agent to remain patient during chop and engage only when structural conditions align.
3. Quality of historical results
The quality of historical trades is characterized by high win consistency and balanced profit distribution. Out of 36 completed trades, 30 ended in profit and 6 resulted in losses, yielding an 83.33% win rate. The strategy achieved a profit factor of 3.75, signifying that gross gains were nearly four times the magnitude of gross losses. The average trade yield is +10.29%, while the median trade return closely mirrors this figure at +9.75%. The tight alignment between average and median returns indicates that performance is driven by consistent, repeatable trade outcomes rather than skewed by extreme statistical outliers. The best individual trade generated a return of +34.98%, whereas the worst trade logged a loss of -15.97%. Furthermore, the top three winning trades accounted for 25.28% of the strategy's total gross profit. Because less than a third of total profits originate from its three best events, the model demonstrates broad structural durability across its trade population rather than over-reliance on a few fortunate market moves.
4. Risk, drawdown and losing behavior
Risk management metrics reveal a tightly controlled equity curve with remarkably brief drawdown phases. The maximum historical drawdown experienced by the strategy was 15.97%, which precisely mirrors the single worst trade loss of -15.97%. This structural alignment indicates that peak equity declines were primarily driven by individual trade exits rather than compounding strings of consecutive losses. Indeed, the strategy's longest historical losing streak was just 1 trade, meaning that every single loss was immediately followed by a winning trade throughout the entire 7-year sample. Conversely, the longest winning streak reached 14 consecutive trades. The inability of losses to cluster reflects effective market regime identification by the algorithm. However, holding positions for an average of 1273.71 hours entails carrying overnight risk and structural equity drawdowns during multi-week retracements before positions reach their target exits. For automated trading agents, this requires capital allocation models capable of enduring temporary unrealized drawdowns without prematurely invalidating the strategy's broader position rules.
5. Behavior through time and yearly stability
A multi-year examination of annual breakdowns demonstrates robust temporal stability across varying macroeconomic conditions in stock markets. In 2020, the strategy recorded 5 trades, achieving a 100% win rate and a cumulative return of +80.09%. In 2021, it again logged 5 trades with a 100% win rate, yielding +92.50%. The highest activity occurred in 2022 with 9 completed trades, resulting in 7 wins and 2 losses for a 77.78% win rate and a +66.09% return sum. The strategy experienced its sole negative calendar year in 2023, generating a modest loss of -14.13% across 3 trades with 2 wins and 1 loss. Recovery was swift in 2024, producing a +84.80% return over 4 trades, all of which were profitable. In 2025, performance remained solid with 5 trades, an 80.00% win rate, and a sum of +23.59%. In 2026, 5 completed trades produced 3 wins and 2 losses, returning +37.35%. With 6 out of 7 calendar years posting positive net returns, the model shows consistent year-over-year profitability and structural adaptability.
6. Recent period since 2024-06-01 versus full history
Evaluating recent behavior provides crucial insights into whether the algorithm maintains its edge in contemporary market environments. Since June 1, 2024, the strategy completed 13 trades, accumulating a combined profit sum of +110.76%. Remarkably, these 13 recent trades represent over 36% of the strategy's total historical trade count of 36, signaling an acceleration in signal generation compared to its multi-year average. Despite this increased trade cadence, performance quality remained exceptionally strong. The recent period delivered a substantial portion of the overall historical gains, reinforcing that the quantitative model has not suffered from regime decay or curve-fitting obsolescence. Comparing the recent sum of +110.76% across 13 trades against the historical average trade yield of +10.29% confirms that modern trades continue to align with long-term performance expectations. For trading bots deployed in current conditions, this recent sample provides statistical confirmation that the core mathematical mechanics remain effective.
7. Strengths and limitations
The strategy exhibits notable statistical strengths alongside distinct operational trade-offs. Its primary strength lies in its exceptional win rate of 83.33% combined with a robust profit factor of 3.75 and a modest maximum drawdown of 15.97%. The distribution of gains is remarkably uniform, as evidenced by the top three winning trades representing only 25.28% of gross profits and the median trade return (+9.75%) remaining close to the average trade return (+10.29%). Furthermore, losing streaks never exceeded 1 trade. Conversely, the principal limitation is the strategy's low trade count of 36 over 7.02 years. While filtering out bad trades improves profitability, small sample sizes carry higher statistical uncertainty than high-frequency models. Additionally, long holding durations (averaging 1273.71 hours) require substantial patience from traders and expose open positions to macro gaps and earnings announcements. The strategy also showed vulnerability during 2023, where a low trade volume of 3 trades resulted in a net loss of -14.13%.
8. DevioLab analytical conclusion
The DevioLab rank of 3 and score of 70.02 for BAC underscore a highly successful quantitative architecture optimized for patient trend capture. By utilizing a 15-minute timeframe to pinpoint execution while maintaining position horizons that stretch over weeks or months, the algorithm achieves an optimal balance between micro-precision and macro-trend harvesting. The strategy successfully avoids the friction of over-trading, generating an annualized return of 135.23% on just 5.13 trades per year. For quantitative developers constructing automated trading robots or algorithmic agents, this model provides a compelling template for stock trading. It demonstrates that superior risk-adjusted returns can be achieved not through hyperactive execution, but through strict signal selectivity, resilient drawdown control, and the disciplined trailing of major market expansions.
9. Data scope and methodology
The statistics analyzed in this report are derived from historical BACktested execution data for the BAC stock strategy between September 16, 2019, and September 22, 2026. All metrics, including trade counts, holding periods, win rates, drawdowns, and annualized returns, are based on completed, closed trade events within this 7.02-year window. The recent sample window isolates trades with exit timestamps on or after June 1, 2024. These BACktested figures represent simulated quantitative performance under standardized historical data modeling and do not account for real-time order slippage, institutional transaction fees, liquidity shifts, or unexpected corporate actions. Past performance derived from historical strategy simulations does not guarantee future results or live trading returns.