Backtest Summary (2020‑01‑01 – 2026‑02‑22)

Interpretation

• The strategy generated a modest 19.6 % return over six years, far below the NIFTY 50 benchmark (+108 %).
• Sharpe 0.37 indicates low risk‑adjusted performance; the high drawdown (‑37 %) reflects a single large open trade that never closed before the backtest end.
• Zero win‑rate stems from the exit rule never triggering a sell; the position remained open at the final bar, inflating the unrealised P&L but not delivering realized profits.
• Exposure is near‑full capital, increasing risk while transaction costs consumed ~1.5 % of gross P&L.

Recommendations

1. Add a realistic exit – use a stop‑loss (e.g., S.fixed(pct=0.04)) or trailing stop to close losing trades and free capital.
2. Cap position duration – enforce a maximum holding period of 20 days (e.g., exit='(C >= EntryPrice()*1.08) | (C <= EntryPrice()*0.96) | (DayOfWeek() > 20)').
3. Refine volume filter – replace the rolling 20‑day mean with a shorter look‑back (e.g., 5‑day) to capture spikes without lag.
4. Run a parameter sweep – test variations of the price‑rise threshold (1.2 %–2 %) and volume multiplier (1.5×–3×) to identify a more robust configuration.
5. Validate with walk‑forward – after a promising variant, execute walk_forward_test to assess over‑fitting.

Next step – I can run a parameter sweep across entry thresholds and volume multipliers, or add a stop‑loss and re‑backtest. Let me know which direction you prefer.

Backtest Summary (2020‑01‑01 – 2026‑02‑22)

Interpretation

• The strategy resulted in a net loss of 4.73 %, significantly underperforming the NIFTY 50 benchmark (+80.9 %).
• A single breakout trade triggered but failed to reach the 12 % target or 5 % stop, remaining open with an unrealized loss at the test end.
• High drawdown (‑37.1 %) and low Sharpe (0.09) highlight the extreme risk of buying 52‑week high breakouts without secondary confirmation.
• Transaction costs further eroded the negative return by ~6 %.

Recommendations

1. Add volume confirmation – only enter breakouts if volume is > 1.5× the 20‑day average (e.g., (C > HHV(H, 252).shift(1)) & (V > 1.5 * V.rolling(20).mean())).
2. Implement a time‑based exit – force close after 30 days if targets aren't hit (as requested, but not triggered in this single‑trade sample).
3. Use a trailing stop – protect gains on partial moves with S.trailing(atr_mult=2).
4. Research regime dependency – check if this strategy works better during specific Nifty regimes using detect_market_regime.
5. Optimize targets – a 12 % target may be too aggressive for a 30‑day window; try testing 5–8 % targets.

Next step – I can run a monte_carlo_test to see if this loss was an outlier, or use run_parameter_sweep to find better profit/stop combinations.

Backtest Summary (2021‑01‑01 – 2026‑02‑22)

Interpretation

• Adding a 1.5× volume multiplier did not improve results for AETHER; the strategy still produced a single trade that resulted in a negative return of 4.73 %.
• The volume confirmation actually filtered out potential entries but the one that remained was a "bull trap" that failed to reach the profit target.
• High drawdown (‑37.1 %) remains the primary risk, as the stock has struggled to maintain momentum after hitting new highs.

Recommendations

1. Shorten the 52-week lookback – try a 3‑month or 6‑month high (e.g., HHV(H, 63)) to capture more frequent momentum shifts.
2. Tighten the stop-loss – a 5 % stop is wide for a volatile stock; try 2–3 % with a trailing component.
3. Use RSI filter – only enter if RSI(14) is between 60 and 75, avoiding "overbought" extensions.
4. Compare with peers – run scan_universe on the Chemicals sector to see if this breakout setup performs better on other stocks.
5. Analyze regime – use detect_market_regime to see if AETHER is currently in a "trending" or "mean‑reverting" phase.

Next step – I can run a parameter_sweep to optimize the volume multiplier and target levels, or scan the NIFTY_PHARMA universe for similar breakout opportunities.