AMT · American Tower Corporation
Stock Universe · marché actions mondialAMT Quantitative Swing Trading Analysis: Evaluating a 15-Minute Algorithmic Model with 83.59% Win Rate and 9.81% Maximum Drawdown
This quantitative research study examines an algorithmic trading strategy evaluated on American Tower Corporation (AMT) equities across a 15-minute candle interval spanning 6.97 years from September 2019 to September 2026. Generating a cumulative closed-trade gain of 1146.76% across 195 completed transactions, the strategy achieves a profit factor of 2.50 and an 83.59% historical win rate. Notably, maximum equity drawdown was constrained to 9.81%, while the top three winning trades accounted for only 11.15% of total gross profits, indicating a broadly distributed return profile rather than reliance on extreme market outliers. Designed as decision logic suitable for deployment in automated trading bots, trading robots, or quantitative trading agents, the strategy exhibits low concentration risk and a consistent multi-day holding structure averaging 78.25 hours per position. This analysis delves into the underlying statistical mechanics, payoff balance, drawdown containment, and temporal stability across full-history and recent evidence windows.
Analyse au 16.09.2026. Les données de la stratégie sont mises à jour séparément.Lire l’analyse complète
Strategy Profile
The evaluated algorithmic trading strategy operates on American Tower Corporation (AMT) stock using a 15-minute price bar resolution. Over a total historical sample of 6.97 years spanning from September 16, 2019, through September 3, 2026, the strategy executed 195 completed trades. It attained a total cumulative sum of closed-trade returns equal to 1146.76%, yielding an annualized return calculation of 179.46%. Out of 195 closed positions, 163 concluded with positive returns while 32 resulted in net losses, establishing an overall hit rate of 83.59%. The strategy's performance profile is capped by a maximum recorded peak-to-trough drawdown of 9.81%, reflecting disciplined downside mitigation within backtested execution parameters. Achieving a DevioLab Score of 82.03, this model ranks number 1 among strategies evaluated for AMT and is designated as a Core strategy due to its independent quantitative top-rank standing. The underlying logic offers systematic entry and exit signals designed for integration into execution infrastructure such as automated trading bots, algorithmic trading robots, or programmatic trading agents operating on medium-resolution equity market data.
Trading Rhythm and Position Duration
Although the strategy evaluates market conditions on a granular 15-minute bar timeframe, its operational rhythm is characterized by multi-day swing positions rather than high-frequency execution. Over the nearly seven-year backtested period, the model closed an average of 27.99 trades per year, which translates to approximately 2.32 completed trades per active month. Position duration data reveals an average holding time of 78.25 hours (approximately 3.26 calendar days) and a median holding duration of 48.50 hours (roughly 2.02 calendar days). The divergence between average and median holding durations indicates that while typical trades settle within two trading days, select positions extend across several additional sessions to accommodate prevailing trends. In terms of trade frequency, the average interval between consecutive position exits stands at 12.95 days, with a median exit spacing of 10.90 days. This indicates that the algorithm remains out of the market or holds active positions across multi-bar windows without triggering over-trading. Automated trading agents executing this logic are therefore subjected to moderate turnover, avoiding the transaction churn associated with sub-hourly scalping while maintaining active exposure during localized price swings in AMT.
Quality of Historical Results
A detailed inspection of the strategy's return distribution reveals a balanced payoff structure. The average return across all 195 completed trades is 1.54%, closely aligned with the median trade return of 1.44%. The tight proximity between mean and median values suggests that trade returns are symmetrically distributed, with minimal skewness or distortion from extreme individual trades. The best single transaction yielded a return of 15.50%, whereas the worst individual trade resulted in a loss of -7.39%. This establishes a maximum trade payout asymmetry ratio of roughly 2.1 to 1 in favor of the positive outlier relative to the negative extreme. Crucially, the top three winning trades collectively account for just 11.15% of total gross profits. In quantitative systems, a low profit concentration metric demonstrates that cumulative performance is sustained by consistent, repeatable wins across the trade population rather than a handful of anomalous windfalls. Combined with a profit factor of 2.50, these metrics demonstrate that historical profitability stems from high hit-rate consistency supported by controlled average trade sizing.
Risk, Drawdown and Losing Behavior
Risk containment represents a primary structural feature of this AMT strategy. The historical maximum peak-to-trough drawdown reached 9.81%, maintaining single-digit downside exposure despite operating across dynamic market environments over nearly seven years. This peak drawdown is notably constrained when evaluated alongside the model's overall profit factor of 2.50. Analysis of streak dynamics shows that the strategy experienced a maximum consecutive losing streak of only 3 trades, compared to a maximum winning streak of 18 consecutive positive trades. The brevity of losing streaks prevents rapid capital erosion during adverse regimes, enabling equity curves to recover efficiently. Because the maximum individual loss was bounded at -7.39%, the strategy did not suffer catastrophic single-trade drawdowns. The combination of an 83.59% win rate and short losing sequences suggests that automated trading robots utilizing this decision logic would historically experience relatively low duration equity stagnation, though past drawdown bounds do not guarantee absolute protection against future structural shifts.
Behavior Through Time and Yearly Stability
Evaluating yearly breakdown statistics demonstrates persistent performance across distinct operational periods. In 2019 (partial year starting mid-September), the strategy logged 5 trades with a 100% win rate and a cumulative return of 5.68%. In 2020, activity expanded to 31 trades, delivering a 77.42% win rate and a 74.67% return sum across 24 wins and 7 losses. Subsequent full calendar years maintained steady execution: 2021 produced 26 trades (84.62% win rate, 23.23% return sum), 2022 generated 32 trades (75.00% win rate, 32.36% return sum), and 2023 saw 25 trades (80.00% win rate, 28.90% return sum). Trade frequency remained relatively stable, ranging between 24 and 33 completed transactions per full year. The win rate never dropped below 75.00% in any individual calendar year. Profit generation accelerated in 2024 (24 trades, 91.67% win rate, 35.25% sum) and 2025 (33 trades, 84.85% win rate, 43.03% sum), continuing into partial-year 2026 (19 trades, 94.74% win rate, 58.07% sum). This uniform yearly distribution illustrates that positive performance was broad-based rather than localized to a single trending year.
Recent Period Since 2024-06-01 Versus Full History
A critical test for any quantitative model is its recent operational integrity. In the evaluation window beginning June 1, 2024, through September 3, 2026, the strategy recorded 64 completed trades, representing nearly 32.8% of all transactions in the 6.97-year dataset. Cumulative closed-trade returns during this recent period reached 117.47%, accounting for approximately 10.2% of total historical net profits. Comparing recent trade frequency to the full history indicates a slight acceleration in execution pace: the strategy averaged roughly 2.8 trades per month during the recent window compared to the full-history average of 2.32 trades per active month. Win rates in recent calendar years also exhibited elevated levels, reaching 91.67% in 2024 and 94.74% in 2026. The substantial sample size of 64 recent trades provides robust evidence that the algorithm's statistical edge remained intact and highly active in recent market cycles, avoiding the performance degradation that frequently affects legacy quantitative models over time.
Strengths and Limitations
The primary quantitative strength of this strategy lies in its combination of high win rate (83.59%), strong profit factor (2.50%), and limited peak drawdown (9.81%). The low concentration of gross profit in top winning trades (11.15%) confirms that returns are structurally sound and broad-based. Furthermore, the moderate trade frequency of ~28 trades per year reduces execution overhead for automated trading agents. However, specific structural limitations must be recognized. First, the strategy relies on a higher win rate to drive returns, meaning that its average win magnitude is smaller relative to potential losses if a trade breaches normal parameters. Second, because total historical trades count 195 over nearly 7 years, individual year sample sizes average around 25 to 33 trades; while statistically meaningful, larger trade samples provide even stronger confidence bounds. Finally, as with all backtested quantitative models, historical holding dynamics and return distributions reflect past price action and cannot guarantee identical behavior under altered market regimes or unexpected structural volatility.
DevioLab Analytical Conclusion
With a DevioLab Score of 82.03 and a rank of #1 for AMT equity strategies, this algorithmic model represents a highly effective swing trading architecture for American Tower Corporation. Its design strikes a favorable balance between high win consistency, modest holding periods averaging 78.25 hours, and strict capital protection reflected in a 9.81% maximum drawdown. The strategy avoids over-reliance on outlier trades, delivering steady progress supported by a maximum losing streak of just 3 trades. Recent performance since mid-2024 reinforces its temporal durability, showing robust trade generation and strong total return capture. For quantitative developers and traders seeking logic suitable for deployment in automated trading bots or algorithmic execution agents, this AMT strategy provides a compelling reference model combining high hit-rate efficiency with robust risk metrics.
Data Scope and Methodology
The analysis presented in this report is based strictly on historical backtested transaction data generated between September 16, 2019, and September 3, 2026, covering 6.97 calendar years. All calculations reflect closed trades executed on 15-minute price candle intervals for American Tower Corporation (AMT) stock. Net trade return statistics, holding periods, profit factor calculations, and drawdown metrics are derived directly from closed-trade log sequences. Percentage metrics represent cumulative sums of individual trade performance and do not reflect direct compounded portfolio equity curves, slippage modeling, or broker fee adjustments unless explicitly stated. Historical performance outcomes are evaluated purely for analytical and quantitative research purposes and do not constitute investment advice or guarantees of future live trading results.