Nepse Decode · Chart & signals · rule-based, no black boxes
Reversal Signal
Built for falling and sideways markets. Buys oversold bounces, exits at the mean.
The headline numbers below are the backtest — a historical simulation (2024-09-01 → 2026-10-28), not live results. What these same rules have produced on real sessions is in the Live forward record directly underneath.
Stocks by sector
The stocks this system currently holds or tracks, grouped by sector. Pick a sector to filter every table on this page — open positions, closed trades, replay positions, and backtest history. Pick All sectors to reset.
Live forward record — what the signals have actually produced
Every live Entry alert becomes a Rs 10 lakh paper trade at the signal close, scored against real forward OHLC. Loading…
Open paper positions
| Date | Symbol | Entry | Stop loss | Trail stop | Target 1 | Target 2 | Current | Unrealized | Progress to T2 | Sessions |
|---|---|---|---|---|---|---|---|---|---|---|
| Loading… | ||||||||||
Closed paper trades — newest first
| Symbol | Entry date | Exit date | Entry | Exit | Exit reason | P&L % | P&L Rs | Sessions |
|---|---|---|---|---|---|---|---|---|
| Loading… | ||||||||
Closed-trade P&L distribution
Trade profile — backtest (historical simulation)
Hypothetical replay of these exact rules on historical data — not trades anyone took. Real outcomes are in the Live forward record above.
| Metric | Value | Metric | Value |
|---|---|---|---|
| Total return | -20.51% | Capital multiple | ×0.79 |
| Benchmark (NEPSE buy & hold) | -1.04% / yr | Alpha | -9.07% / yr |
| Total trades | 254 | Win rate | 46.46% |
| Profit factor | 0.85 | Avg win / avg loss | 6.83% / 7.04% |
| Sharpe (monthly) | -0.17 | Sortino | -0.08 |
| Max drawdown | 34.77% | Avg holding | 9 days |
| Best trade | NBL 33.92% | Worst trade | HATHY -17.63% |
| Charges paid | Rs 18.75 L | Backtest window | 2024-09-01 → 2026-10-28 |
Equity curve — Rs 1 Cr through this system
Weekly marked-to-market equity curve — Rs 1 Cr start, cash + open positions − exit costs. Drawn from the published backtest series, not a forecast.
The rules (exactly as coded)
- Setup: RSI(14) below 30 and close below the lower Bollinger Band (20, 2)
- Trigger: bullish candle (close > open) closing in the top half of its range (rejection)
- Filter: 20-day avg turnover above Rs 10 lakh
- Stop: 1.5 × ATR(14) below entry (tighter — mean reversion). Trail: none; exit at the mean
- Targets: T1 = 20-day SMA (the mean), T2 = entry + 2 × risk (reference levels — not traded; see methodology)
Backtest — positions still open when the replay ended (hypothetical)
Positions the historical replay still held when it ended on 2026-10-28 — hypothetical, and not the system's live positions (live positions are the Open paper positions in the Live forward record above). A replay position exits only on its stop loss or a trail break, while the live forward record also exits at Target 1 or after 20 sessions — so the same signal can be closed in the live record yet still open here. Entry is taken at the signal day's close in the replay; "Current" shows the latest traded price for scale only. Progress shows how far price has moved toward Target 2.
| Date | Symbol | Alert | Entry | Stop loss | Trail stop | Target 1 | Target 2 | Current | Unrealized | Progress |
|---|---|---|---|---|---|---|---|---|---|---|
| Loading… | ||||||||||
Closed history — backtest trades
Finished trades from the mechanical backtest replay, newest first — hypothetical, not live trades (all 254 finished trades shown). The Live forward record above shows what the signals have produced since tracking began. P&L is after the same trading charges the backtest applies.
| Symbol | Entry date | Exit date | Entry price | Exit price | Exit reason | P&L % | P&L Rs | Hold days |
|---|---|---|---|---|---|---|---|---|
| Loading trade history… | ||||||||
About this system
Reversal Signal is one of Nepse Decode's transparent signal systems. Every rule above is public — there is no hidden model and no "AI prediction". Signals are generated mechanically from daily OHLCV data (corporate-action adjusted); a stock appears here only if it passes every filter, including a Rs 10 lakh average-turnover liquidity bar. This page is educational and is not investment advice.
Words on this page, in plain language
- ATR (Average True Range)
- How much a stock's price typically swings in a day. A "2 × ATR stop" means: if the price falls twice that typical daily swing below your entry, you exit. It sizes risk to how jumpy the stock is.
- RSI(14)
- A 0–100 gauge of how fast the price has been rising versus falling over the last 14 days. 55–80 means strong upward momentum that is not yet extreme.
- SMA (simple moving average)
- The average closing price over the last N days. Price above its 50-day average means the medium-term trend points up.
- Breakout
- The closing price prints higher than any close of the previous 50 sessions — the market just pushed into new territory.
- Risk (per trade)
- Entry price minus stop price — the most you plan to lose if the trade fails. Targets are measured in multiples of this.
- T1 / T2 targets
- Planned take-profit levels: entry + 1.5 × risk (T1) and entry + 3 × risk (T2).
- Win rate
- The share of decided (finished) trades that made money, shown with a Wilson 95% interval — the range of true win rates consistent with the sample. Open trades never count. A low win rate can still profit if winners are much bigger than losers.
- Profit factor
- Total money won divided by total money lost, from the forward ledger's closed trades net of 0.5% round-trip charges. Above 1.0 means the system took more from the market than it gave back.
- Too few trades (n < 30)
- Below 30 decided trades the badge on the Live forward record applies: the sample is too small for the win rate or profit factor to prove an edge, however good they look. n = decided trades only.
- Max drawdown ("deepest fall")
- The worst peak-to-trough loss the portfolio suffered in the backtest — the pain you would have had to sit through.
- Alpha vs NEPSE
- How much more (or less) this system earned per year than simply buying and holding the NEPSE index itself. Negative means buy & hold won.
- Sharpe / Sortino
- Return per unit of bumpiness. Higher means smoother gains. Sortino counts only the bad bumps (losses), ignoring upside jumps.
- Turnover filter
- Stocks trading under Rs 10 lakh a day on average are skipped — too thin to enter and exit cleanly without moving the price.
Data: daily NEPSE OHLCV, corporate-action adjusted — historical bars from the open scraper dataset, recent sessions from our own NEPSE feed. Generated 2026-10-29. Educational only — not investment advice.