AIG · American International Group, Inc.
Stock Universe · global stock marketAIG · AIG 15-Minute Algorithmic Strategy Analysis: High Precision Position Holding and Tail Risk Management in Equity Automation
This quantitative study evaluates the top-ranked algorithmic trading strategy for AIG on a 15-minute timeframe. Designed as executable logic for automated trading bots, trading robots, and trading agents, the strategy exhibits an exceptional historical win rate of 88.89% and a cumulative net profit of +1,475.20% across 72 completed trades over nearly seven years of historical simulation. Earning a DevioLab score of 73.98 and ranking first for this equity symbol, the algorithm operates with low frequency and extended position durations, averaging almost 589 hours per trade. While achieving an outstanding profit factor of 8.33 and maintaining profitability across every single calendar year analyzed, the strategy carries structural downside fragility, highlighted by a maximum drawdown of 47.30% driven primarily by a single severe worst trade of -46.06%. This paper provides an exhaustive statistical examination of the model's pay-off distribution, temporal stability, recent post-2024 performance, and inherent risk-reward trade-offs.
Analysis as of 13.09.2026. Strategy data is updated separately.Read full analysis
Strategy profile
The quantitative model designated for AIG operates on a 15-minute candle resolution and represents the core primary strategy for this symbol within the DevioLab framework, ranking first overall with a score of 73.98. Across a simulated sample history spanning from September 16, 2019, through September 13, 2026—a period covering roughly 6.99 years—the strategy completed exactly 72 trades. Despite using an intraday 15-minute chart for entry and exit evaluation, the system is fundamentally non-scalping and highly selective, generating an average of only 10.29 trades per year. Over the full backtested history, the strategy produced a cumulative profit of +1,475.20%, translating to an annualized performance metric of 98.77%. The decision logic is structured to serve as an automated trading agent or trading bot algorithm, filtering noise on short timeframes while capturing sustained multi-week equity trends. The dataset demonstrates that high-frequency chart monitoring can be effectively paired with low-frequency execution to capture substantial macro moves in individual stock equities.
Trading rhythm and position duration
A critical structural feature of this algorithmic approach is the vast discrepancy between its underlying chart interval and its actual holding duration. While market data is ingested on a 15-minute basis, position holding times reflect a patient swing-to-position trading style. The average holding duration per trade is 588.91 hours, which equates to roughly 24.5 days, while the median holding duration stands at 402.75 hours, or approximately 16.8 days. Exits occur infrequently, with an average interval between closed trades of 35.93 days and a median of 27.05 days. On an active operational scale, the strategy generates approximately 1.41 trades per active month. These spatial properties indicate that automated trading robots running this code will spend long periods holding active market exposure without altering position state. The algorithm filters out short-term market noise, waiting for prolonged trend continuation before triggering exit conditions. This extended holding structure minimizes trade churn, though it requires deployed execution engines to withstand prolonged overnight and weekend market exposure.
Quality of historical results
The statistical distribution of trade outcomes reveals remarkable consistency and payoff health. Out of 72 closed trades, 64 were profitable, yielding an exceptional historical win rate of 88.89%, while only 8 trades resulted in losses. This hit rate translated into a profit factor of 8.33, meaning gross profits exceeded gross losses by more than eight times. The average trade gain across all closed signals reached +4.55%, closely aligning with the median trade return of +4.95%. This tight alignment between mean and median expectations indicates that positive outcomes are systematically distributed across trades rather than skewed by a few extreme events. Furthermore, the top three largest winning trades collectively accounted for only 16.12% of total gross profit, confirming that historical outperformance was driven by a broad cohort of winning trades rather than reliance on rare, unpredictable windfalls. The strategy's maximum single winning trade generated a gain of +29.26%, demonstrating solid upside participation during favorable market trends.
Risk, drawdown and losing behavior
Despite its dominant hit rate and high profit factor, the strategy displays a pronounced risk asymmetry that demands rigorous risk management when deployed in automated trading systems. The maximum historical drawdown reached 47.30%. This severe equity contraction was almost entirely caused by a single extreme outlier event: the strategy's worst trade recorded a loss of -46.06%. While the algorithm achieved an extraordinary winning streak of 20 consecutive profitable trades and never suffered more than 2 consecutive losing trades, the severe magnitude of its worst trade illustrates that losing positions can occasionally run deeply against the system before an exit is executed. This structural vulnerability implies that while loss frequency is extremely low, loss magnitude can be disproportionately high. Systems operating this logic without secondary safety stops or allocation limits face non-trivial tail risk during abrupt structural market shifts or gap events.
Behavior through time and yearly stability
An examination of annual performance demonstrates impressive temporal consistency across varying market regimes. The strategy recorded positive net returns in every full calendar year evaluated. In 2019, during the initial partial period, the strategy executed 1 trade with a 100% win rate, yielding +3.98%. In 2020, trade activity expanded to 10 trades with an 80% win rate, generating +56.99%. Performance peaked over the next two years: 2021 delivered 13 trades with a 92.31% win rate (+70.60% cumulative gain), and 2022 produced 13 trades with an identical 92.31% win rate (+76.32% cumulative gain). The subsequent years maintained robust hit rates: 2023 saw 9 trades with an 88.89% win rate (+22.66%), 2024 generated 8 trades with an 87.50% win rate (+26.90%), 2025 delivered 8 trades with a 100% win rate (+39.12%), and 2026 produced 10 trades with an 80% win rate (+30.82%). Trade volume remained steadily bound between 8 and 13 trades annually, confirming stable signal generation across market cycles.
Recent period since 2024-06-01 versus full history
Evaluating performance in the recent testing window from June 1, 2024, onward reveals strong, sustained operational effectiveness. During this period, the strategy completed 22 trades, accumulating a combined gain of +75.36%. Notably, these 22 trades account for nearly 30.5% of the total 72 trades generated over the entire seven-year dataset, reflecting a higher concentration of completed cycles in recent market conditions. The performance achieved in this window demonstrates that the core edge has remained functional and highly profitable without experiencing decay. The strategy's ability to maintain strong absolute returns in recent market environments validates its ongoing relevance for automated trading agents seeking persistent statistical edges in stock equities.
Strengths and limitations
The primary strength of the AIG algorithm lies in its exceptional historical hit rate of 88.89%, high profit factor of 8.33, and well-distributed profit structure where the top three trades represent under 17% of gross gains. Its annual stability is noteworthy, generating positive returns across every calendar year tested. The primary limitation stems from trade frequency and tail risk. With only 72 completed trades over nearly seven years, the absolute sample size is relatively small, which introduces statistical uncertainty. Furthermore, the worst trade loss of -46.06% and maximum drawdown of 47.30% demonstrate that when losses do occur, they can severely impact portfolio equity. Deployed execution logic must balance the high win rate against the potential for deep individual trade drawdowns.
DevioLab analytical conclusion
The AIG 15-minute algorithmic strategy presents a compelling quantitative profile, earning its top rank on DevioLab with a score of 73.98. By translating intraday 15-minute price action into multi-week trend holds, the algorithm achieves an extraordinary balance of high win frequency and strong overall return efficiency. It provides robust candidate logic for automated trading bots and trading robots optimized for high-probability position trading. However, practitioners must account for the structural tail risk highlighted by its 47.30% maximum drawdown. Implementing prudent capital allocation, position sizing rules, or protective overlay stops is essential to harness the strategy's 88.89% win rate while mitigating the impact of rare but severe downside events.
Data scope and methodology
This analysis is based strictly on backtested, simulated historical trade statistics for AIG spanning from September 16, 2019, to September 13, 2026. All reported performance metrics, including holding times, returns, drawdowns, and win rates, are calculated from completed closed trades within the simulated environment. Metrics do not incorporate real-world execution friction such as broker commissions, exchange fees, slippage, market impact, or borrow costs. Recent period metrics post-2024 reflect trades closed after June 1, 2024. Historical backtested performance does not guarantee future results. This article is published exclusively for educational and analytical purposes and does not constitute financial or investment advice.