AXP · American Express Company
Stock Universe · marché actions mondialQuantitative Analysis of AXP · AXP 15m Algorithmic Strategy: High-Selectivity Multi-Week Position Holding with Exceptional Profit Factor
An analytical evaluation of the top-ranked algorithmic trading strategy for AXP on DevioLab.com (DevioLab Score: 88.61). Operating on a 15-minute execution timeframe, this core quantitative strategy demonstrates high selectivity, yielding an 84.09% win rate across 44 completed trades over a seven-year historical dataset spanning 2019 to 2026. Characterized by long average position holding times of 772.94 hours (~32.2 days) and a low trade frequency of 6.27 trades per year, the strategy achieved a backtested profit factor of 6.88 alongside a maximum drawdown strictly bounded at 8.11%. The strategy's total cumulative simulated return reaches +4386.42%, providing a compelling historical baseline for automated trading bots, trading robots, and quantitative trading agents designed for swing-to-position execution in equity markets.
Analyse au 21.09.2026. Les données de la stratégie sont mises à jour séparément.Lire l’analyse complète
Strategy profile
The evaluated algorithmic model, designated as AXP 0 +4386.42% 1TRAD-RHH8, occupies the rank 1 position for the asset symbol AXP on DevioLab.com with a quantitative score of 88.61. Classified as a core strategy operating on 15-minute price bars, the algorithm spans a historical evaluation window of 7.02 years from September 16, 2019, through September 21, 2026. Over this multi-year evaluation, the strategy completed 44 closed trades, achieving 37 winning trades against only 7 losing trades, which translates to a historical win rate of 84.09%. The simulated cumulative trade gain across the complete sample reaches +4386.42%, which corresponds to a theoretical annualized return metric of 222.07%. Although the underlying price series is sampled on a 15-minute intraday interval, the system's operational logic functions as a high-conviction position-selection engine. It filters out short-term market noise in favor of extended trend captures, making it an ideal decision logic for automated trading bots and algorithmic trading agents looking to execute strategic swing allocations on AXP.
Trading rhythm and position duration
A critical structural feature of this strategy is its exceptionally low trade velocity combined with long holding durations. Over 7.02 years of backtested history, the strategy registered 44 total completed trades, yielding an average trade frequency of 6.27 trades per year, or approximately 1.19 trades per active month. This low turnover clearly establishes that the strategy does not operate as intraday scalping or high-frequency momentum trading. Position holding times reflect a macro swing framework: the average holding duration stands at 772.94 hours (approximately 32.2 days), while the median holding duration is 542.13 hours (approximately 22.6 days). Spacing between completed trade exits is similarly wide, averaging 56.56 days with a median interval of 49.00 days. These figures indicate that an automated trading robot executing this algorithm remains inactive for weeks at a time between signals, holding active positions through multi-week price cycles to capture broad moves rather than trading short-term noise.
Quality of historical results
The statistical distribution of trade returns indicates superior profit generation balanced across the sample. The average trade gain across all 44 trades is +9.66%, while the median trade yield sits at +7.06%. The strategy's single best trade produced a gain of +32.07%, whereas its worst individual trade lost -8.11%. The overall profit factor reaches an outstanding 6.88, indicating that total gross winning trade percentages exceed total gross losing trade percentages by nearly seven to one. To evaluate whether historical gains were driven by isolated extreme events, we examine profit concentration: the top three winning trades combined account for only 20.98% of total gross profit. This low concentration ratio confirms that historical performance is broad-based and structurally resilient, relying on steady repeated wins rather than an unrepeatable single windfall trade.
Risk, drawdown and losing behavior
Risk parameters across the 7.02-year dataset reflect remarkable equity preservation. The maximum historical drawdown recorded for the strategy was -8.11%, a figure that precisely mirrors the loss incurred in its single worst completed trade (-8.11%). This equivalence demonstrates that equity retracements were tightly bound, with peak-to-trough drawdowns never cascading into compounding multi-trade losses. This risk containment is further reinforced by the strategy's streak metrics: the longest historical losing streak was limited to just 2 consecutive trades, compared to a peak winning streak of 13 consecutive profitable trades. For trading agents and quantitative risk models, this statistical profile demonstrates minimal equity curve fragility, low downside volatility, and an ability to recover rapidly from temporary adverse price swings.
Behavior through time and yearly stability
Analyzing annual performance metrics from 2020 through 2026 reveals exceptional temporal consistency across varying equity market regimes. In 2020, the strategy recorded 7 trades with a 100% win rate and a cumulative return sum of +141.75%. In 2021, 4 completed trades achieved a 100% win rate, delivering +59.98%. The strategy faced its most challenging conditions in 2022, completing 7 trades with 4 wins and 3 losses (57.14% win rate) while remaining net positive at +48.25%. In 2023, 6 trades produced 4 wins (66.67% win rate) for a combined +23.56%. Performance surged again in 2024, logging 8 trades with an 800% win rate (8 wins, 0 losses) and +69.06%. In 2025, 5 trades resulted in 4 wins (80.00% win rate) for +36.98%, followed by 2026 with 7 trades, 6 wins (85.71% win rate), and +45.48%. Crucially, every single calendar year in the seven-year sample generated positive summed returns, demonstrating multi-year stability.
Recent period since 2024-06-01 versus full history
Filtering performance to the recent evaluation window beginning June 1, 2024, provides vital insight into modern strategy efficacy. Since 2024-06-01, the strategy completed 18 trades, generating a cumulative return sum of +134.67%. Representing roughly 40.9% of all 44 lifetime trades, this recent period reflects a higher density of signal triggers relative to the strategy's multi-year average. The average yield per trade during this window remains strong at approximately +7.48% (+134.67% sum over 18 trades). This recent clustering of successful trades demonstrates that the algorithm's decision logic has remained highly effective in recent market structure, continuing to generate reliable long-side swing entries without experiencing performance decay or structural drift.
Strengths and limitations
The primary statistical strengths of this AXP trading model include its high win rate (84.09%), robust profit factor (6.88), tight maximum drawdown (-8.11%), and healthy profit distribution, where the top three winners represent only 20.98% of gross gains. Furthermore, its perfect annual stability record across seven calendar years provides strong backtested evidence of cross-regime functionality. Conversely, material limitations exist. With only 44 completed trades over 7.02 years, the total statistical sample size is relatively small, which naturally expands the statistical confidence intervals. Additionally, the average holding time of 772.94 hours (~32.2 days) means deployed trading robots must hold positions through significant overnight gaps, earnings announcements, and macro news events, exposing open equity to unhedged market risks between 15-minute signal checks.
DevioLab analytical conclusion
With a top ranking on AXP and a DevioLab Score of 88.61, this algorithmic strategy stands out as a high-precision, low-turnover quantitative model. Its historical profile balances an 84.09% hit rate with controlled downside exposure, as evidenced by its matching -8.11% worst trade and maximum drawdown figures. The strategy's multi-week holding structure effectively extracts substantial swing moves while avoiding over-trading and premature exits. For quantitative developers, portfolio managers, and operators of automated trading bots, the strategy offers a robust, stress-tested logic suited for deployment environments where signal quality, low drawdown, and capital protection take priority over high trade frequency.
Data scope and methodology
All analytical conclusions presented in this report are grounded strictly in the backtested performance dataset generated for strategy identifier AXP 0 +4386.42% 1TRAD-RHH8 on 15-minute price history from September 16, 2019, to September 21, 2026. The recent evaluation window analyzes completed trades with exit timestamps on or after June 1, 2024. Statistics reflect simulated gross trade performance and do not incorporate live market execution parameters such as broker commissions, exchange fees, bid-ask spreads, order slippage, or borrowing costs. Historical backtested results are provided exclusively for quantitative research and educational context and do not constitute financial advice or guarantees of future live trading returns.