GIS · General Mills Inc.
Stock Universe · global stock marketGIS Algorithmic Trading Strategy Analysis: 85% Win Rate and 3.75 Profit Factor Across Multi-Year Stock Execution
This quantitative evaluation analyzes an algorithmic trading strategy for GIS (General Mills) spanning 6.91 years of historical data from September 2019 through August 2026. Designed for implementation in automated trading bots and algorithmic trading agents, the strategy operates on a 15-minute timeframe while capturing multi-week structural swings. Across 80 completed trades, the model generated a cumulative backtested return of +587.39% with an annualized return profile of 47.96%. The performance is characterized by an exceptional win rate of 85.00%, a profit factor of 3.75, and a peak historical maximum drawdown of 13.07%. Notably, profit distribution is remarkably uniform, with the top three winning trades contributing just 12.63% of total gross gains. Recent testing from June 1, 2024 onward demonstrates robust activity with 31 completed trades generating +64.92% in aggregate returns, affirming sustained structural efficacy across varying equity market regimes.
Analysis as of 05.11.2026. Strategy data is updated separately.Read full analysis
Strategy profile
The evaluated algorithm for GIS operates within equity market dynamics, utilizing a 15-minute candle resolution to determine trade timing while maintaining position horizons that extend across weeks. Over a backtested timeframe of 6.91 years—spanning from September 16, 2019 to August 13, 2026—the strategy achieved a DevioLab Score of 83.92, ranking first overall among strategies tested for the GIS asset ticker. Total historical closed-trade performance reached +587.39%, translating to an annualized return figure of 47.96% across 80 completed trade cycles. The strategy's decision logic provides a deterministic framework suitable for deployment via automated trading bots, trading robots, or quantitative trading agents seeking structured exposure to GIS price movements.
Trading rhythm and position duration
Despite evaluating market data on a 15-minute intraday chart interval, the strategy displays a low-frequency, medium-duration swing trading rhythm. Over the 6.91-year evaluation period, the system completed 80 trades, averaging 11.58 trades per year or approximately 1.29 trades per active month. Position holding times reflect a patient approach: the average holding duration is 344.28 hours (approximately 14.3 days), while the median holding duration stands at 314.00 hours (roughly 13.1 days). This close alignment between mean and median holding metrics indicates consistent position exposure rather than statistical skewing from extreme hold-time outliers. Furthermore, the average interval between consecutive exit events is 31.66 days, with a median of 27.99 days. This temporal spacing confirms that automated trading agents executing this framework spend considerable time out of active market positions or holding established trades through minor intraday noise.
Quality of historical results
The quantitative quality of historical trade executions is defined by exceptional accuracy and balanced return distribution. Out of 80 completed trades, 68 resolved in positive return territory against 12 losing trades, yielding a overall win rate of 85.00%. The system achieved a profit factor of 3.75, meaning gross profits exceeded gross losses by nearly four to one. The average trade performance recorded +2.89%, while the median trade performance stood closely at +2.82%. This minimal variance between average and median outcomes demonstrates that profitability is steady and predictable across individual trade samples. The maximum positive outcome recorded was +12.22%, contrasted against a worst-case trade result of -10.70%. A critical insight into performance stability lies in profit concentration: the strategy's top three winning trades accounted for only 12.63% of total gross profit. This low concentration metric confirms that strategy gains are broad-based across the sample rather than dependent on a few isolated windfall trades.
Risk, drawdown and losing behavior
Risk parameters throughout the 6.91-year history reveal controlled drawdowns relative to cumulative yield. The maximum historical drawdown was restricted to 13.07%, reflecting a high ratio of equity accumulation to peak-to-trough equity decline. The algorithm demonstrated an extraordinary sequence of profitable executions, achieving a peak winning streak of 39 consecutive trades without a loss between late 2019 and early 2023. Conversely, the longest consecutive losing streak was limited to just 2 trades. The strategy's worst single trade loss of -10.70% occurred within an overall regime of constrained risk exposure, preventing localized market shocks from compounding into broader structural equity erosion for automated trading robots executing the strategy logic.
Behavior through time and yearly stability
Annual distribution of trade results illustrates strong consistency across distinct financial market environments. In 2019, across a partial sample, the strategy completed 2 trades (100% win rate) yielding +4.11%. The years 2020, 2021, and 2022 produced flawless historical accuracy, each logging exactly 12 trades with a 100% win rate and annual return sums of +62.07%, +50.93%, and +31.10% respectively. Performance underwent a regime shift in 2023, where 6 completed trades yielded 3 wins and 3 losses (50.00% win rate), accumulating +7.52%. Operational performance rebounded in 2024, recording 12 trades with an 83.33% win rate and +33.65% total return. In 2025, 14 trades were executed with 9 wins and 5 losses (64.29% win rate), generating +16.82%. The final partial period of 2026 logged 10 trades with 8 wins (80.00% win rate), accruing +25.30%. Notably, every full calendar year produced positive net performance.
Recent period since 2024-06-01 versus full history
Analyzing performance from June 1, 2024 to August 13, 2026 provides a dedicated observation window into recent market conditions. During this period, the strategy recorded 31 completed trades, accumulating +64.92% in aggregate returns. This recent activity represents 38.75% of all historical trades logged across the full 6.91-year history, indicating an increase in trade opportunity frequency in recent years relative to the earlier baseline. Despite the increased execution density, the recent window maintained strong profitability, demonstrating that the strategy's logic remained functional amidst recent GIS price dynamics without suffering from performance decay.
Strengths and limitations
The primary analytical strengths of this strategy include its high historical win rate (85.00%), robust profit factor (3.75%), modest peak drawdown (13.07%), and low reliance on outlier trades (top 3 winners contribute 12.63% of gross profit). The alignment between average trade (+2.89%) and median trade (+2.82%) further underscores system stability. Limitations stem primarily from sample size constraints: 80 closed trades over nearly 7 years represent a relatively small sample for statistically rigorous backtesting, averaging just over 11 trades per year. Additionally, holding times averaging ~14 days require patient capital commitment and operational discipline from trading agents, while the shift in win rate from 100% in 2020-2022 to 50%-64% in 2023 and 2025 highlights vulnerability to shifting volatility regimes.
DevioLab analytical conclusion
The quantitative evaluation of this GIS strategy demonstrates a disciplined, long-horizon swing model operating effectively on a 15-minute execution frame. Securing the top rank for the GIS asset with a DevioLab Score of 83.92, the system balances high accuracy with low drawdowns. Its modest trade frequency and multi-week hold times make it suitable for algorithmic trading agents designed to minimize unnecessary over-trading. While historical backtested metrics reflect strong performance, future live deployments depend on maintaining risk control standards during extended periods of market transition.
Data scope and methodology
This analysis is based on historical backtested closed-trade data for GIS on a 15-minute timeframe from September 16, 2019 to August 13, 2026. All reported statistics, including win rates, holding times, profit factors, drawdowns, and annual breakdowns, refer strictly to closed trades within the simulated dataset. Trade metrics assume fixed non-leveraged trade execution logic without incorporating live execution variables such as slippage, exchange fees, or order routing delays. Historical performance results are evaluated for quantitative analysis only and do not constitute investment advice or guarantee future trading performance.