Statistical research lab
Confidence intervals, market-regime attribution, chronological holdout checks and execution-cost stress tests based on recorded paper trades. Diagnostics describe the observed sample; they are not forecasts.
Do the results survive more than one favourable sample?
We test whether returns persist across clustered days, different market conditions, later observations and harsher execution assumptions.
Statistical confidence
Moving-block bootstrap resamples active trade-day P&L in five-day blocks. It reports a 95% interval for mean daily P&L, total P&L, positive-sample frequency and the 95th percentile of simulated drawdown.
Market regimes
Closed trades are grouped using indicators recorded at entry: high volatility, directional up/down trend, range-bound, transition/mixed or insufficient context.
Robustness
We add progressively harsher slippage and charge assumptions, test a chronological 70/30 observation split and remove one exit month at a time to check concentration.
VWAP + EMA Pullback
0 closed trades · 0 active trade-days · 0 exit months
Regime analysis will populate after trades with entry-time assessment records are closed.
Execution-cost robustness
Additional slippage is applied per side on top of the recorded paper-fill assumptions. Charge shocks increase recorded charges.
| Scenario | Net P&L | Expectancy | Profit factor | Trades |
|---|---|---|---|---|
| Recorded paper assumptions | ₹0.00 | — | 0.00 | 0 |
| Additional 0.5× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 1× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 2× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 3× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 25% | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 50% | ₹0.00 | — | 0.00 | 0 |
Configured base slippage: 25.0 bps per side. These are deterministic sensitivities, not historical order-book replay.
Chronological 70/30 observation split
Earlier closed trades versus the most recent 30% of closed trades.
At least 30 closed trades are required for a 70/30 chronological paper-trade split.
Leave-one-month-out concentration check
Recalculate the remaining record after excluding each calendar exit month in turn.
At least three observed exit months are required.
Methodology and limitations
- Bootstrap uses active exit days only; days with no closed trade are not assumed to be zero-return days.
- Regime labels are reconstructed from entry-time stored feature snapshots where available. Legacy trades without those snapshots are classified as insufficient_entry_context.
- Slippage and charge shocks are sensitivities applied to recorded paper fills, not a replay of historical order-book execution.
- Parameter perturbation requires re-running the strategy on historical candles and options data; it is not inferred from closed-trade records.
- Parameter perturbation is not calculated in v1.7 because it requires re-running the full strategy against historical index candles and contemporaneous option quotes. Trade outcomes alone cannot validly reveal the result of different indicator parameters or entry rules.
VWAP + EMA Pullback – Option Selling
0 closed trades · 0 active trade-days · 0 exit months
Regime analysis will populate after trades with entry-time assessment records are closed.
Execution-cost robustness
Additional slippage is applied per side on top of the recorded paper-fill assumptions. Charge shocks increase recorded charges.
| Scenario | Net P&L | Expectancy | Profit factor | Trades |
|---|---|---|---|---|
| Recorded paper assumptions | ₹0.00 | — | 0.00 | 0 |
| Additional 0.5× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 1× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 2× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 3× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 25% | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 50% | ₹0.00 | — | 0.00 | 0 |
Configured base slippage: 25.0 bps per side. These are deterministic sensitivities, not historical order-book replay.
Chronological 70/30 observation split
Earlier closed trades versus the most recent 30% of closed trades.
At least 30 closed trades are required for a 70/30 chronological paper-trade split.
Leave-one-month-out concentration check
Recalculate the remaining record after excluding each calendar exit month in turn.
At least three observed exit months are required.
Methodology and limitations
- Bootstrap uses active exit days only; days with no closed trade are not assumed to be zero-return days.
- Regime labels are reconstructed from entry-time stored feature snapshots where available. Legacy trades without those snapshots are classified as insufficient_entry_context.
- Slippage and charge shocks are sensitivities applied to recorded paper fills, not a replay of historical order-book execution.
- Parameter perturbation requires re-running the strategy on historical candles and options data; it is not inferred from closed-trade records.
- Parameter perturbation is not calculated in v1.7 because it requires re-running the full strategy against historical index candles and contemporaneous option quotes. Trade outcomes alone cannot validly reveal the result of different indicator parameters or entry rules.
Opening Range Breakout + Retest
0 closed trades · 0 active trade-days · 0 exit months
Regime analysis will populate after trades with entry-time assessment records are closed.
Execution-cost robustness
Additional slippage is applied per side on top of the recorded paper-fill assumptions. Charge shocks increase recorded charges.
| Scenario | Net P&L | Expectancy | Profit factor | Trades |
|---|---|---|---|---|
| Recorded paper assumptions | ₹0.00 | — | 0.00 | 0 |
| Additional 0.5× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 1× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 2× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 3× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 25% | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 50% | ₹0.00 | — | 0.00 | 0 |
Configured base slippage: 25.0 bps per side. These are deterministic sensitivities, not historical order-book replay.
Chronological 70/30 observation split
Earlier closed trades versus the most recent 30% of closed trades.
At least 30 closed trades are required for a 70/30 chronological paper-trade split.
Leave-one-month-out concentration check
Recalculate the remaining record after excluding each calendar exit month in turn.
At least three observed exit months are required.
Methodology and limitations
- Bootstrap uses active exit days only; days with no closed trade are not assumed to be zero-return days.
- Regime labels are reconstructed from entry-time stored feature snapshots where available. Legacy trades without those snapshots are classified as insufficient_entry_context.
- Slippage and charge shocks are sensitivities applied to recorded paper fills, not a replay of historical order-book execution.
- Parameter perturbation requires re-running the strategy on historical candles and options data; it is not inferred from closed-trade records.
- Parameter perturbation is not calculated in v1.7 because it requires re-running the full strategy against historical index candles and contemporaneous option quotes. Trade outcomes alone cannot validly reveal the result of different indicator parameters or entry rules.
Opening Range Breakout + Retest – Option Selling
0 closed trades · 0 active trade-days · 0 exit months
Regime analysis will populate after trades with entry-time assessment records are closed.
Execution-cost robustness
Additional slippage is applied per side on top of the recorded paper-fill assumptions. Charge shocks increase recorded charges.
| Scenario | Net P&L | Expectancy | Profit factor | Trades |
|---|---|---|---|---|
| Recorded paper assumptions | ₹0.00 | — | 0.00 | 0 |
| Additional 0.5× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 1× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 2× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 3× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 25% | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 50% | ₹0.00 | — | 0.00 | 0 |
Configured base slippage: 25.0 bps per side. These are deterministic sensitivities, not historical order-book replay.
Chronological 70/30 observation split
Earlier closed trades versus the most recent 30% of closed trades.
At least 30 closed trades are required for a 70/30 chronological paper-trade split.
Leave-one-month-out concentration check
Recalculate the remaining record after excluding each calendar exit month in turn.
At least three observed exit months are required.
Methodology and limitations
- Bootstrap uses active exit days only; days with no closed trade are not assumed to be zero-return days.
- Regime labels are reconstructed from entry-time stored feature snapshots where available. Legacy trades without those snapshots are classified as insufficient_entry_context.
- Slippage and charge shocks are sensitivities applied to recorded paper fills, not a replay of historical order-book execution.
- Parameter perturbation requires re-running the strategy on historical candles and options data; it is not inferred from closed-trade records.
- Parameter perturbation is not calculated in v1.7 because it requires re-running the full strategy against historical index candles and contemporaneous option quotes. Trade outcomes alone cannot validly reveal the result of different indicator parameters or entry rules.
EMA + RSI Pullback
0 closed trades · 0 active trade-days · 0 exit months
Regime analysis will populate after trades with entry-time assessment records are closed.
Execution-cost robustness
Additional slippage is applied per side on top of the recorded paper-fill assumptions. Charge shocks increase recorded charges.
| Scenario | Net P&L | Expectancy | Profit factor | Trades |
|---|---|---|---|---|
| Recorded paper assumptions | ₹0.00 | — | 0.00 | 0 |
| Additional 0.5× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 1× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 2× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 3× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 25% | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 50% | ₹0.00 | — | 0.00 | 0 |
Configured base slippage: 25.0 bps per side. These are deterministic sensitivities, not historical order-book replay.
Chronological 70/30 observation split
Earlier closed trades versus the most recent 30% of closed trades.
At least 30 closed trades are required for a 70/30 chronological paper-trade split.
Leave-one-month-out concentration check
Recalculate the remaining record after excluding each calendar exit month in turn.
At least three observed exit months are required.
Methodology and limitations
- Bootstrap uses active exit days only; days with no closed trade are not assumed to be zero-return days.
- Regime labels are reconstructed from entry-time stored feature snapshots where available. Legacy trades without those snapshots are classified as insufficient_entry_context.
- Slippage and charge shocks are sensitivities applied to recorded paper fills, not a replay of historical order-book execution.
- Parameter perturbation requires re-running the strategy on historical candles and options data; it is not inferred from closed-trade records.
- Parameter perturbation is not calculated in v1.7 because it requires re-running the full strategy against historical index candles and contemporaneous option quotes. Trade outcomes alone cannot validly reveal the result of different indicator parameters or entry rules.
EMA + RSI Pullback – Option Selling
0 closed trades · 0 active trade-days · 0 exit months
Regime analysis will populate after trades with entry-time assessment records are closed.
Execution-cost robustness
Additional slippage is applied per side on top of the recorded paper-fill assumptions. Charge shocks increase recorded charges.
| Scenario | Net P&L | Expectancy | Profit factor | Trades |
|---|---|---|---|---|
| Recorded paper assumptions | ₹0.00 | — | 0.00 | 0 |
| Additional 0.5× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 1× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 2× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 3× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 25% | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 50% | ₹0.00 | — | 0.00 | 0 |
Configured base slippage: 25.0 bps per side. These are deterministic sensitivities, not historical order-book replay.
Chronological 70/30 observation split
Earlier closed trades versus the most recent 30% of closed trades.
At least 30 closed trades are required for a 70/30 chronological paper-trade split.
Leave-one-month-out concentration check
Recalculate the remaining record after excluding each calendar exit month in turn.
At least three observed exit months are required.
Methodology and limitations
- Bootstrap uses active exit days only; days with no closed trade are not assumed to be zero-return days.
- Regime labels are reconstructed from entry-time stored feature snapshots where available. Legacy trades without those snapshots are classified as insufficient_entry_context.
- Slippage and charge shocks are sensitivities applied to recorded paper fills, not a replay of historical order-book execution.
- Parameter perturbation requires re-running the strategy on historical candles and options data; it is not inferred from closed-trade records.
- Parameter perturbation is not calculated in v1.7 because it requires re-running the full strategy against historical index candles and contemporaneous option quotes. Trade outcomes alone cannot validly reveal the result of different indicator parameters or entry rules.
Supertrend + VWAP Confirmation
0 closed trades · 0 active trade-days · 0 exit months
Regime analysis will populate after trades with entry-time assessment records are closed.
Execution-cost robustness
Additional slippage is applied per side on top of the recorded paper-fill assumptions. Charge shocks increase recorded charges.
| Scenario | Net P&L | Expectancy | Profit factor | Trades |
|---|---|---|---|---|
| Recorded paper assumptions | ₹0.00 | — | 0.00 | 0 |
| Additional 0.5× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 1× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 2× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 3× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 25% | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 50% | ₹0.00 | — | 0.00 | 0 |
Configured base slippage: 25.0 bps per side. These are deterministic sensitivities, not historical order-book replay.
Chronological 70/30 observation split
Earlier closed trades versus the most recent 30% of closed trades.
At least 30 closed trades are required for a 70/30 chronological paper-trade split.
Leave-one-month-out concentration check
Recalculate the remaining record after excluding each calendar exit month in turn.
At least three observed exit months are required.
Methodology and limitations
- Bootstrap uses active exit days only; days with no closed trade are not assumed to be zero-return days.
- Regime labels are reconstructed from entry-time stored feature snapshots where available. Legacy trades without those snapshots are classified as insufficient_entry_context.
- Slippage and charge shocks are sensitivities applied to recorded paper fills, not a replay of historical order-book execution.
- Parameter perturbation requires re-running the strategy on historical candles and options data; it is not inferred from closed-trade records.
- Parameter perturbation is not calculated in v1.7 because it requires re-running the full strategy against historical index candles and contemporaneous option quotes. Trade outcomes alone cannot validly reveal the result of different indicator parameters or entry rules.
Supertrend + VWAP Confirmation – Option Selling
0 closed trades · 0 active trade-days · 0 exit months
Regime analysis will populate after trades with entry-time assessment records are closed.
Execution-cost robustness
Additional slippage is applied per side on top of the recorded paper-fill assumptions. Charge shocks increase recorded charges.
| Scenario | Net P&L | Expectancy | Profit factor | Trades |
|---|---|---|---|---|
| Recorded paper assumptions | ₹0.00 | — | 0.00 | 0 |
| Additional 0.5× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 1× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 2× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 3× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 25% | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 50% | ₹0.00 | — | 0.00 | 0 |
Configured base slippage: 25.0 bps per side. These are deterministic sensitivities, not historical order-book replay.
Chronological 70/30 observation split
Earlier closed trades versus the most recent 30% of closed trades.
At least 30 closed trades are required for a 70/30 chronological paper-trade split.
Leave-one-month-out concentration check
Recalculate the remaining record after excluding each calendar exit month in turn.
At least three observed exit months are required.
Methodology and limitations
- Bootstrap uses active exit days only; days with no closed trade are not assumed to be zero-return days.
- Regime labels are reconstructed from entry-time stored feature snapshots where available. Legacy trades without those snapshots are classified as insufficient_entry_context.
- Slippage and charge shocks are sensitivities applied to recorded paper fills, not a replay of historical order-book execution.
- Parameter perturbation requires re-running the strategy on historical candles and options data; it is not inferred from closed-trade records.
- Parameter perturbation is not calculated in v1.7 because it requires re-running the full strategy against historical index candles and contemporaneous option quotes. Trade outcomes alone cannot validly reveal the result of different indicator parameters or entry rules.
9/21 EMA Crossover
0 closed trades · 0 active trade-days · 0 exit months
Regime analysis will populate after trades with entry-time assessment records are closed.
Execution-cost robustness
Additional slippage is applied per side on top of the recorded paper-fill assumptions. Charge shocks increase recorded charges.
| Scenario | Net P&L | Expectancy | Profit factor | Trades |
|---|---|---|---|---|
| Recorded paper assumptions | ₹0.00 | — | 0.00 | 0 |
| Additional 0.5× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 1× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 2× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 3× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 25% | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 50% | ₹0.00 | — | 0.00 | 0 |
Configured base slippage: 25.0 bps per side. These are deterministic sensitivities, not historical order-book replay.
Chronological 70/30 observation split
Earlier closed trades versus the most recent 30% of closed trades.
At least 30 closed trades are required for a 70/30 chronological paper-trade split.
Leave-one-month-out concentration check
Recalculate the remaining record after excluding each calendar exit month in turn.
At least three observed exit months are required.
Methodology and limitations
- Bootstrap uses active exit days only; days with no closed trade are not assumed to be zero-return days.
- Regime labels are reconstructed from entry-time stored feature snapshots where available. Legacy trades without those snapshots are classified as insufficient_entry_context.
- Slippage and charge shocks are sensitivities applied to recorded paper fills, not a replay of historical order-book execution.
- Parameter perturbation requires re-running the strategy on historical candles and options data; it is not inferred from closed-trade records.
- Parameter perturbation is not calculated in v1.7 because it requires re-running the full strategy against historical index candles and contemporaneous option quotes. Trade outcomes alone cannot validly reveal the result of different indicator parameters or entry rules.
9/21 EMA Crossover – Option Selling
0 closed trades · 0 active trade-days · 0 exit months
Regime analysis will populate after trades with entry-time assessment records are closed.
Execution-cost robustness
Additional slippage is applied per side on top of the recorded paper-fill assumptions. Charge shocks increase recorded charges.
| Scenario | Net P&L | Expectancy | Profit factor | Trades |
|---|---|---|---|---|
| Recorded paper assumptions | ₹0.00 | — | 0.00 | 0 |
| Additional 0.5× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 1× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 2× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Additional 3× configured slippage per side | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 25% | ₹0.00 | — | 0.00 | 0 |
| Charges increased by 50% | ₹0.00 | — | 0.00 | 0 |
Configured base slippage: 25.0 bps per side. These are deterministic sensitivities, not historical order-book replay.
Chronological 70/30 observation split
Earlier closed trades versus the most recent 30% of closed trades.
At least 30 closed trades are required for a 70/30 chronological paper-trade split.
Leave-one-month-out concentration check
Recalculate the remaining record after excluding each calendar exit month in turn.
At least three observed exit months are required.
Methodology and limitations
- Bootstrap uses active exit days only; days with no closed trade are not assumed to be zero-return days.
- Regime labels are reconstructed from entry-time stored feature snapshots where available. Legacy trades without those snapshots are classified as insufficient_entry_context.
- Slippage and charge shocks are sensitivities applied to recorded paper fills, not a replay of historical order-book execution.
- Parameter perturbation requires re-running the strategy on historical candles and options data; it is not inferred from closed-trade records.
- Parameter perturbation is not calculated in v1.7 because it requires re-running the full strategy against historical index candles and contemporaneous option quotes. Trade outcomes alone cannot validly reveal the result of different indicator parameters or entry rules.
Bootstrap intervals and resampled drawdowns are conditional on the observed paper-trade sample and the selected block length. A positive-sample frequency is not the probability that future trades or future months will be profitable. Regime buckets can be noisy and can reflect classification thresholds. Small samples must not be used to select a real-money strategy.