Backtested research example
🗓 Backtest period: 2018-01-01..2023-01-01
Spec ID: spec-multitimeframe-momentum-us-stocks-1783167355 · Generated: 2026-07-04 12:32 UTC
Cluster: Momentum · Sub Cluster: Weekly-Daily Trend Alignment
Long-only US stock trend-following strategy that buys liquid names only when both weekly and daily moving-average trends are bullish. Positions are ATR risk-sized, capped per name, and exited on trend deterioration, ATR trailing stop, or maximum holding period.
The trade is based on the empirical persistence of intermediate-term equity trends. Requiring agreement between a daily trend filter and a slower weekly trend filter attempts to reduce whipsaws by entering only when short- and medium-horizon market participants are aligned.
The structural rationale is that large-cap stocks can trend after information diffusion, institutional rebalancing, earnings revisions, and investor underreaction. ATR-based sizing and stops convert the signal from a pure direction rule into a volatility-normalized implementation, limiting concentration in high-volatility names.
[code omitted from public view]
| Param | Value | Notes |
|---|---|---|
| Rule type | MULTI_TIMEFRAME_TREND_MOMENTUM |
Long-only trend-following rule |
| Daily trend | 20d SMA > 50d SMA, close > 20d SMA, 20d SMA 5d slope > 0 | Daily bullish filter |
| Weekly trend | 20w SMA > 50w SMA, weekly close > 20w SMA, 20w SMA 4w slope > 0 | Uses completed weekly bars only |
| Weekly resampling | Friday close, last available daily close | Daily data resampled to weekly |
| Entry liquidity | 20d average dollar volume >= $10,000,000 | Skip illiquid signals |
| Position sizing | ATR volatility-adjusted, 14d ATR | Target risk per position = 1% of equity |
| Position cap | 10% of equity per symbol | Single-name concentration limit |
| Max leverage | 4.0x | Portfolio leverage ceiling |
| Exits | Daily/weekly trend failure, 3.0x ATR trailing stop, 126d max hold | Multiple independent exit triggers |
| Universe | 20 US stocks sorted by capitalization | Includes AAPL, AMZN, MSFT, NVDA, TSLA, BRK share classes, and others |
| Backtest window | 2018-01-01 to 2023-01-01 | Daily bars |
| Costs | $0.004/share commission, $1 minimum/order, 1% max commission, 0 bps slippage | Slippage assumption is optimistic |
| # | Concern | Status |
|---|---|---|
| 1 | Weekly signal may use incomplete current-week data | Mitigated: weekly indicators use completed weeks only |
| 2 | Same-close signal and execution timing | Needs scrutiny: rebalance at close while signals use close/high/low/volume |
| 3 | Universe selection survivorship bias | Needs scrutiny: fixed listed symbols may not reflect point-in-time membership |
| 4 | Corporate actions and adjusted prices | Requires validation that price history is split/dividend adjusted consistently |
| 5 | Liquidity and execution costs | Partially covered by dollar-volume filter, but slippage is set to 0 bps |
The backtest produced a positive total return of 30.84% with moderate risk-adjusted performance: Sharpe 0.70, Sortino 0.65, and Calmar 0.46. Max drawdown was contained at -11.89% with 8.84% realized volatility, suggesting the ATR sizing and trend exits helped reduce downside exposure. The win rate was low at 39.40%, but profit factor was 1.58, consistent with a trend-following profile where larger winners compensate for frequent small losses across 802 trades.
| Metric | Value |
|---|---|
| Total Return | 30.84% |
| Sharpe | 0.70 |
| Sortino | 0.65 |
| Calmar | 0.46 |
| Max Drawdown | -11.89% |
| Volatility | 8.84% |
| Win Rate | 39.40% |
| Profit Factor | 1.58 |
| Total Trades | 802 |
| Symbols | 20 (AAPL, ACON, AMZN, BETR, BRK-A, BRK-B, GOODO, GOOG, GOOGL, HOLO, +10 more) |
Backtests are historical simulations for research purposes only. They are not investment advice and do not guarantee future performance.