AMD: Bollinger band re-entry, 15m (profile 2) - validation
Selected historical result. 16 preset candidates were evaluated for this market and bar interval; 6 passed the training quality and profitability checks. Training return: 3.36% after assumed costs over 2025-10-11 to 2026-07-11. Selection ranked return divided by maximum drawdown (with a 1% floor), then return and trade count. The choice was frozen before the separate validation period, 2026-07-19 to 2026-10-16. Only profitable validation results with usable robustness evidence are published. This selected subset does not show the success rate of all trials or predict future returns. Robustness comparisons may lose money. TradesViz strategy discovery candidate: Bollinger band re-entry on AMD.US, 15-minute bars. Validation period: 2026-07-19 through 2026-10-16. Historical simulation, not a live account. Tests whether a close back inside the lower Bollinger band is followed by a return toward its middle band. Entry: [CROSSED(input_col=close, input_col2=BOLLBANDS_LOWER(period=30, sd_levels=2), direction=above)] Exit: [close > BOLLBANDS_MIDDLE(period=30, sd_levels=2)] Preset parameter profile 2. Settings: long only; next-bar-open fills; 25,000 starting equity in quote currency; 25% of current equity per position; 14-period ATR, 2.5 ATR initial stop, 2R target, 4 ATR trailing stop, and a 60-bar maximum holding period. Session 09:30 to 16:00 in exchange time. Positions close at the end of each session. Assumed costs: 0.01% of traded value commission per side and 0.01 points slippage per market fill. These are illustrative fixed assumptions, not broker quotes. Funding, financing, borrow fees, varying spreads, and market impact are not separately modeled. Selection design: 16 preset candidates share the training period 2025-10-11 to 2026-07-11. Only the training-selected candidate may be evaluated on the fixed validation period 2026-07-19 to 2026-10-16, after a 7-day gap. Validation failures are retained; the runner does not try another candidate on that holdout. Training profitability is selection-biased. Published reports are a selected subset, not the overall success rate of these strategies. Limitations: Volatility bands do not bound losses; a strong downtrend can continue after a re-entry signal. Validation and robustness checks do not establish future profitability. Data availability, gaps and ambiguous intrabar fills must be checked in the completed report. Requested period: 2026-07-19 to 2026-10-16. Available test data: 2026-07-20 to 2026-10-16, 1,664 bars (America/New_York dates). Simulated historical strategy test for AMD.US using 15-minute bars from 2026-07-19 to 2026-10-16. Asset class: stocks. Indicators and functions used: BOLLBANDS_LOWER, BOLLBANDS_MIDDLE, CROSSED; Box 1 condition: [CROSSED(input_col=close, input_col2=BOLLBANDS_LOWER(period=30, sd_levels=2), direction=above)]; Box 2 condition: [close > BOLLBANDS_MIDDLE(period=30, sd_levels=2)]. Long only. Fills at the next bar's open. Opposite signal reverses the position. Starting equity: 25,000. Position size: 25 % of equity. Stop loss: 2.5 ATR. Take profit: 2 R. Trailing stop: 4 ATR. Commission per side: 0.01 %. Slippage per fill: 0.01 points. Maximum holding period: 60 bars. Close positions at session end. Risk-setting ATR period: 14. Session: 09:30 to 16:00 (exchange time). Saved results. Trades: 33; Net P&L: 1,211.29282; Return: 4.845171%; Win rate: 69.69697%.
Shared by pk78 · Published 2026-10-15 · Snapshot 2026-10-17 · Report revision 4
What was tested
- Market
- AMD.US
- Timeframe
- 15-minute bars
- Test window
- 2026-07-19 – 2026-10-16
- Engine
- tv-engine-3
Simulated historical results, not live trading performance. Historical results do not predict future returns. Shared sections are selected by the author; hidden sections have not been independently verified. The author controls which sections are visible. This is a saved result; market data revisions or engine updates can change a later rerun.
Result and evidence
Promising
Profitable with a profit factor of 1.2 or more. A strategy with no real edge would do this well less than 1% of the time. Check it on other periods before trusting it.
The luck check is a statistical estimate for this sample, not the probability of future success. Costs, sample size and testing many variants affect how much confidence a result deserves.
Equity and drawdown
Saved simulated equity across this test window. Drag the range control to inspect a period.
Robustness check
The same strategy tested across periods, symbols and nearby settings. These checks do not make a strategy immune to overfitting.
Earlier periods
Profitable in 5 of 5 periods with sufficient price coverage. Every period made money.
| Test | Trades | Net P&L | Profit factor | Drawdown | Finding |
|---|---|---|---|---|---|
| Jul 19 to Oct 16, 2026 Available data: 2026-07-20 to 2026-10-16, 1,664 bars. | 33 | 1,211.29 | 3.88 | -0.82% | Promising |
| Apr 20 to Jul 18, 2026 Available data: 2026-04-20 to 2026-07-17, 1,612 bars. | 27 | 663.94 | 1.58 | -2.47% | Too few trades |
| Jan 20 to Apr 19, 2026 Available data: 2026-01-20 to 2026-04-17, 1,612 bars. | 30 | 150.85 | 1.13 | -2.30% | Inconclusive |
| Oct 22, 2025 to Jan 19, 2026 Available data: 2025-10-22 to 2026-01-16, 1,560 bars. | 32 | 225.44 | 1.21 | -1.93% | Inconclusive |
| Jul 24 to Oct 21, 2025 Available data: 2025-07-24 to 2025-10-21, 1,638 bars. | 32 | 50.63 | 1.05 | -1.88% | Inconclusive |
Other symbols
Profitable on 2 of 3 other symbols. The rules hold up on most of them.
| Test | Trades | Net P&L | Profit factor | Drawdown | Finding |
|---|---|---|---|---|---|
| Jul 19 to Oct 16, 2026 Available data: 2026-07-20 to 2026-10-16, 1,664 bars. | 33 | 1,211.29 | 3.88 | -0.82% | Promising |
| AAPL.US Available data: 2026-07-20 to 2026-10-16, 1,664 bars. | 29 | 303.47 | 1.91 | N/A | Too few trades |
| MSFT.US Available data: 2026-07-20 to 2026-10-15, 1,300 bars. | 13 | 259.57 | 3.74 | N/A | Too few trades |
| NVDA.US Available data: 2026-07-20 to 2026-10-15, 1,300 bars. | 29 | -182.51 | 0.72 | N/A | Too few trades |
Stops and targets
42 of 42 stop/target settings were profitable in your window. Yours ranks 7. 39 were profitable in most of the 5 periods; the steadiest is a 1.5 ATR stop with no target (5 of 5 periods, $2,661.65 in total).
| Setting | No target | 1R | 1.5R | 2R | 3R | 4R | 5R |
|---|---|---|---|---|---|---|---|
| 0.5 ATR | 250.73 | 102.82 | 294.28 | 312.88 | 330.76 | 252.19 | 264.04 |
| 1 ATR | 624.73 | 527.74 | 584.67 | 518.24 | 614.37 | 624.73 | 624.73 |
| 1.5 ATR | 1,268.83 | 1,259.82 | 1,192.20 | 1,258.61 | 1,268.83 | 1,268.83 | 1,268.83 |
| 2 ATR | 1,159.83 | 1,064.75 | 1,149.47 | 1,159.83 | 1,159.83 | 1,159.83 | 1,159.83 |
| 2.5 ATR | 1,211.29 | 1,173.09 | 1,211.29 | 1,211.29 | 1,211.29 | 1,211.29 | 1,211.29 |
| 3 ATR | 1,191.79 | 1,181.44 | 1,191.79 | 1,191.79 | 1,191.79 | 1,191.79 | 1,191.79 |
Condition numbers
5 of 5 condition settings were profitable in your window. Yours ranks 1. 5 were profitable in most of the 5 periods; the steadiest is your own setting (5 of 5 periods, $2,302.14 in total).
| Setting | |
|---|---|
| 15 | 248.45 |
| 22 | 547.80 |
| 30 | 1,211.29 |
| 45 | 592.81 |
| 60 | 353.37 |
Simulated trades
33 trades in this snapshot. All times are UTC.