Short AI Stock Strategy Backtest Against SPY
Document Lumibot
Summary
This report presents a short backtest of a large-cap stock strategy attributed to a multi-agent AI trading bot and compares it with SPY. The stated test ran from January 4 to January 15, 2026, using Yahoo data and a universe of large technology and other prominent U.S. companies. The report lists performance, risk, and benchmark statistics, including returns, drawdown, Sharpe and Sortino ratios, volatility, time in the market, and winning-day frequency.
Over this brief sample, the strategy’s reported total return was approximately flat but slightly negative, while SPY gained about 1%. The strategy also showed a larger maximum drawdown and volatility, and negative risk-adjusted performance measures. The report provides no trade rationale, model decision record, transaction cost assumptions, or detailed methodology for the signals. Its roughly two-week window is far too short to establish durable performance, and the metrics should be treated as a snapshot rather than evidence of a reliable edge.
Key ideas
- The report compares a large-cap stock strategy with SPY over a short January 2026 test period.
- The strategy’s reported return was slightly negative while the benchmark gained about 1%.
- The strategy had higher reported volatility and a deeper maximum drawdown than SPY.
- The document gives performance statistics but little detail about signal construction or trading costs.
- The short sample does not establish whether the approach can perform consistently.
Tags
Full text
# bull vs bear ai stock trading bot
Tearsheet (generated by QuantStats)
large-cap-plain Compared to SPY 4 Jan, 2026 - 15 Jan, 2026
Benchmark is SPY | LumiBot 4.6.3 | DataSource yahoo | Backtest time 24:01 | Generated by QuantStats (Lumiwealth Version) (v.1.1.5)
Annual Return ⓘ
-9.76%
Total Return ⓘ
-0%
Max Drawdown ⓘ
-1.81%
RoMaD ⓘ
-5.38
Longest DD Days ⓘ
4
Sharpe ⓘ
-0.72
Sortino ⓘ
-1.14
Key Performance Metrics
MetricSPYStrategy
Risk-Free Rate3.56%3.56%
Time in Market67.0%75.0%
Total Return1%-0%
CAGR% (Annual Return)24.3%-9.76%
Sharpe2.59-0.72
RoMaD35.19-5.38
Corr to Benchmark1.00.49
Prob. Sharpe Ratio67.18%41.91%
Smart Sharpe2.38-0.66
Sortino4.73-1.14
Smart Sortino4.34-1.04
Sortino/√23.35-0.8
Smart Sortino/√23.07-0.74
Omega1.50.9
Max Drawdown-0.69%-1.81%
Longest DD Days34
Volatility (ann.)6.41%16.17%
R^20.240.24
Information Ratio-0.1-0.1
Calmar35.19-5.38
Skew0.450.72
Kurtosis0.140.28
Expected Daily%0.05%-0.03%
Expected Monthly%0.66%-0.31%
Expected Yearly%0.66%-0.31%
Daily Value-at-Risk-0.5%-1.41%
Expected Shortfall (cVaR)-0.5%-1.41%
MTD0.66%-0.31%
3M0.66%-0.31%
6M0.66%-0.31%
YTD0.66%-0.31%
1Y0.66%-0.31%
3Y (ann.)24.3%-9.76%
5Y (ann.)24.3%-9.76%
10Y (ann.)24.3%-9.76%
All-time (ann.)24.3%-9.76%
Best Day0.66%1.73%
Worst Day-0.49%-1.06%
Best Month0.66%-0.31%
Worst 1-Month Return0.66%-0.31%
Best Year0.66%-0.31%
Worst Year0.66%-0.31%
Avg. Drawdown-0.51%-1.64%
Avg. Drawdown Days24
Recovery Factor0.960.15
Ulcer Index0.00.01
Serenity Index-4.9-1.53
Annualized Return on Risk Capital2,895.24%-517.88%
Worst 3-Month Return--
Time to Recovery (Days)12
5th Percentile Tail Loss-0.4%-1.01%
Time Underwater (Days)58
Percent Positive Months100.00.0
Avg. Up Month--
Avg. Down Month--
Win Days6.05.33
Loss Days6.06.67
Win Days%50.0%44.44%
Win Month%100.0%0.0%
Win Quarter%100.0%0.0%
Win Year%100.0%0.0%
Beta-1.22
Alpha--0.33
Correlation-48.53%
Treynor Ratio--3.16%
EOY Returns vs Benchmark
YearSPYStrategyMultiplierWon
20260.66-0.31-0.47-
Worst 10 Drawdowns
StartedRecoveredDrawdownDays
2026-01-122026-01-15-1.814
2026-01-052026-01-08-1.464
2026-09-29T23:19:58.443939
image/svg+xml
Matplotlib v3.10.9, https://matplotlib.org/
2026-09-29T23:19:58.490898
image/svg+xml
Matplotlib v3.10.9, https://matplotlib.org/
2026-09-29T23:19:58.668691
image/svg+xml
Matplotlib v3.10.9, https://matplotlib.org/
2026-09-29T23:19:58.729776
image/svg+xml
Matplotlib v3.10.9, https://matplotlib.org/
2026-09-29T23:19:58.776783
image/svg+xml
Matplotlib v3.10.9, https://matplotlib.org/
2026-09-29T23:19:58.825779
image/svg+xml
Matplotlib v3.10.9, https://matplotlib.org/
2026-09-29T23:19:58.883507
image/svg+xml
Matplotlib v3.10.9, https://matplotlib.org/
2026-09-29T23:19:58.944332
image/svg+xml
Matplotlib v3.10.9, https://matplotlib.org/
2026-09-29T23:19:59.000486
image/svg+xml
Matplotlib v3.10.9, https://matplotlib.org/
2026-09-29T23:19:59.051519
image/svg+xml
Matplotlib v3.10.9, https://matplotlib.org/
2026-09-29T23:19:59.096250
image/svg+xml
Matplotlib v3.10.9, https://matplotlib.org/
2026-09-29T23:19:59.150864
image/svg+xml
Matplotlib v3.10.9, https://matplotlib.org/
2026-09-29T23:19:59.207438
image/svg+xml
Matplotlib v3.10.9, https://matplotlib.org/
Disclaimer: This report is for informational purposes only and
should not be considered as investment advice. Past performance
is not indicative of future results.
Parameters Used
ParameterValue
universe['AAPL', 'MSFT', 'NVDA', 'AMZN', 'META', 'GOOGL', 'TSLA', 'AVGO', 'COST', 'JPM', 'V', 'LLY', 'XOM']
agent_max_model_calls200
agent_researcher_modelopenai/gpt-6-luna
agent_bull_modelopenai/gpt-6-luna
agent_bear_modelopenai/gpt-6-luna
agent_trader_modelopenai/gpt-6-luna
agent_model_calls36
agent_researcher_calls9
agent_researcher_cache_hits0
agent_researcher_tool_calls55
agent_researcher_input_tokens826874
agent_researcher_output_tokens24824
agent_researcher_total_tokens851698
agent_researcher_thinking_tokens12069
agent_researcher_cached_input_tokens714612
agent_researcher_cache_write_input_tokens0
agent_researcher_uncached_input_tokens112262
agent_researcher_tool_use_input_tokens0
agent_researcher_latency_ms_total244378
agent_researcher_latency_ms_avg27153.11
agent_researcher_first_event_latency_ms_avg3801.22
agent_researcher_detail_parquetlogs/large-cap-plain_2026-09-29_22-55_DXopM0_agent_detail.parquet
agent_bull_calls9
agent_bull_cache_hits0
agent_bull_tool_calls67
agent_bull_input_tokens824569
agent_bull_output_tokens20035
agent_bull_total_tokens844604
agent_bull_thinking_tokens11108
agent_bull_cached_input_tokens714319
agent_bull_cache_write_input_tokens0
agent_bull_uncached_input_tokens110250
agent_bull_tool_use_input_tokens0
agent_bull_latency_ms_total214296
agent_bull_latency_ms_avg23810.67
agent_bull_first_event_latency_ms_avg3593.89
agent_bull_detail_parquetlogs/large-cap-plain_2026-09-29_22-55_DXopM0_agent_detail.parquet
agent_bear_calls9
agent_bear_cache_hits0
agent_bear_tool_calls87
agent_bear_input_tokens911170
agent_bear_output_tokens33330
agent_bear_total_tokens944500
agent_bear_thinking_tokens17175
agent_bear_cached_input_tokens772479
agent_bear_cache_write_input_tokens0
agent_bear_uncached_input_tokens138691
agent_bear_tool_use_input_tokens0
agent_bear_latency_ms_total343691
agent_bear_latency_ms_avg38187.89
agent_bear_first_event_latency_ms_avg4197.78
agent_bear_detail_parquetlogs/large-cap-plain_2026-09-29_22-55_DXopM0_agent_detail.parquet
agent_trader_calls9
agent_trader_cache_hits0
agent_trader_tool_calls242
agent_trader_input_tokens3612756
agent_trader_output_tokens69886
agent_trader_total_tokens3682642
agent_trader_thinking_tokens53319
agent_trader_cached_input_tokens3428867
agent_trader_cache_write_input_tokens0
agent_trader_uncached_input_tokens183889
agent_trader_tool_use_input_tokens0
agent_trader_latency_ms_total850553
agent_trader_latency_ms_avg94505.89
agent_trader_first_event_latency_ms_avg3672.56
agent_trader_detail_parquetlogs/large-cap-plain_2026-09-29_22-55_DXopM0_agent_detail.parquet
BACKTESTING_DATA_SOURCEyahooShown in full with attribution under the source's licence. Licence: GPL-3.0
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.