Calendar Effects (JP All Stocks)
slug: calendar-effects-jp-all
performance summary
Sharpe (median)
-0.71
Max DD
-30.2%
Win rate
—
Alpha (ann.)
—
p-value
0.41
backtest period: 2022-01-18 → 2026-06-08
Sharpe across 144 grid runs:median -0.71best 1.03 (not representative)worst -3.12p25–p75 -1.53–0.12
Hypothesis
Tests well-known calendar anomalies: Halloween effect, January effect, Turn-of-Month, and Day-of-Week effects on the full Japanese stock universe.
Why it didn't work
Across 144 grid combinations, median Sharpe=-0.71, best=1.03. Primary failure mode: no_edge.
Facts (5)
Best grid Sharpe ratio is 1.03 (p=0.6).
period 2022-01-18–2026-06-08; cost=5bps
cite ↗ /cite/calendar-effects-jp-all-fact-best-sharpevia r-calendar-effects-jp-all-0080
Median Sharpe across 144 grid runs is -0.71.
Full grid of 144 parameter combinations.
Median Sharpe in DOWN regime is -2.91 across 144 runs.
Regime=DOWN, N225 SMA20/SMA60 classification.
Median Sharpe in FLAT regime is -0.47 across 144 runs.
Regime=FLAT, N225 SMA20/SMA60 classification.
Median Sharpe in UP regime is -1.26 across 144 runs.
Regime=UP, N225 SMA20/SMA60 classification.
Runs (30)
| run | params | sharpe | UP | FLAT | DOWN |
|---|---|---|---|---|---|
effect=halloween, holding_days=1, n_stocks=5, cost_bps=5, regime_gate=False r-calendar-effects-jp-all-0000 | effect: "halloween" cost_bps: 5 n_stocks: 5 | -1.15 | -0.27 | -1.61 | -0.64 |
effect=halloween, holding_days=1, n_stocks=5, cost_bps=5, regime_gate=True r-calendar-effects-jp-all-0001 | effect: "halloween" cost_bps: 5 n_stocks: 5 | -1.55 | -1.98 | -1.61 | -0.64 |
effect=halloween, holding_days=1, n_stocks=5, cost_bps=10, regime_gate=False r-calendar-effects-jp-all-0002 | effect: "halloween" cost_bps: 10 n_stocks: 5 | -1.72 | -0.93 | -2.15 | -1.13 |
effect=halloween, holding_days=1, n_stocks=5, cost_bps=10, regime_gate=True r-calendar-effects-jp-all-0003 | effect: "halloween" cost_bps: 10 n_stocks: 5 | -2.11 | -2.67 | -2.15 | -1.13 |
effect=halloween, holding_days=1, n_stocks=20, cost_bps=5, regime_gate=False r-calendar-effects-jp-all-0008 | effect: "halloween" cost_bps: 5 n_stocks: 20 | -1.39 | -2.15 | -1.00 | -2.42 |
effect=halloween, holding_days=1, n_stocks=20, cost_bps=10, regime_gate=False r-calendar-effects-jp-all-0010 | effect: "halloween" cost_bps: 10 n_stocks: 20 | -2.06 | -3.02 | -1.63 | -2.98 |
effect=halloween, holding_days=5, n_stocks=10, cost_bps=5, regime_gate=False r-calendar-effects-jp-all-0020 | effect: "halloween" cost_bps: 5 n_stocks: 10 | -1.92 | -6.42 | -0.58 | -10.72 |
effect=halloween, holding_days=5, n_stocks=20, cost_bps=10, regime_gate=False r-calendar-effects-jp-all-0026 | effect: "halloween" cost_bps: 10 n_stocks: 20 | -2.27 | -8.48 | -0.78 | -11.77 |
effect=halloween, holding_days=20, n_stocks=5, cost_bps=5, regime_gate=False r-calendar-effects-jp-all-0032 | effect: "halloween" cost_bps: 5 n_stocks: 5 | -0.73 | -0.82 | -0.69 | -0.85 |
effect=halloween, holding_days=20, n_stocks=10, cost_bps=5, regime_gate=True r-calendar-effects-jp-all-0037 | effect: "halloween" cost_bps: 5 n_stocks: 10 | -0.95 | -1.20 | -0.88 | -1.28 |
effect=halloween, holding_days=20, n_stocks=20, cost_bps=10, regime_gate=True r-calendar-effects-jp-all-0043 | effect: "halloween" cost_bps: 10 n_stocks: 20 | -0.75 | -1.38 | -0.75 | +0.79 |
effect=january, holding_days=1, n_stocks=10, cost_bps=5, regime_gate=True r-calendar-effects-jp-all-0053 | effect: "january" cost_bps: 5 n_stocks: 10 | -0.31 | -2.29 | +1.51 | — |
effect=january, holding_days=1, n_stocks=20, cost_bps=10, regime_gate=True r-calendar-effects-jp-all-0059 | effect: "january" cost_bps: 10 n_stocks: 20 | -1.70 | -5.00 | +1.11 | — |
effect=january, holding_days=5, n_stocks=10, cost_bps=5, regime_gate=True r-calendar-effects-jp-all-0069 | effect: "january" cost_bps: 5 n_stocks: 10 | -0.11 | -20.04 | +4.30 | — |
effect=january, holding_days=5, n_stocks=20, cost_bps=10, regime_gate=True r-calendar-effects-jp-all-0075 | effect: "january" cost_bps: 10 n_stocks: 20 | -0.23 | -11.58 | +3.67 | — |
effect=january, holding_days=20, n_stocks=5, cost_bps=5, regime_gate=False r-calendar-effects-jp-all-0080 | effect: "january" cost_bps: 5 n_stocks: 5 | +1.03 | -0.14 | +1.58 | — |
effect=january, holding_days=20, n_stocks=10, cost_bps=10, regime_gate=False r-calendar-effects-jp-all-0086 | effect: "january" cost_bps: 10 n_stocks: 10 | +0.83 | -1.15 | +2.47 | — |
effect=turn_of_month, holding_days=1, n_stocks=5, cost_bps=5, regime_gate=False r-calendar-effects-jp-all-0096 | effect: "turn_of_month" cost_bps: 5 n_stocks: 5 | -0.70 | +0.37 | -0.61 | -4.47 |
effect=turn_of_month, holding_days=1, n_stocks=5, cost_bps=5, regime_gate=True r-calendar-effects-jp-all-0097 | effect: "turn_of_month" cost_bps: 5 n_stocks: 5 | -1.15 | -1.83 | -0.61 | -4.47 |
effect=turn_of_month, holding_days=1, n_stocks=10, cost_bps=10, regime_gate=True r-calendar-effects-jp-all-0103 | effect: "turn_of_month" cost_bps: 10 n_stocks: 10 | -1.70 | -3.75 | -0.84 | -6.72 |
effect=turn_of_month, holding_days=5, n_stocks=5, cost_bps=5, regime_gate=True r-calendar-effects-jp-all-0113 | effect: "turn_of_month" cost_bps: 5 n_stocks: 5 | 0.00 | -0.14 | +0.17 | -0.61 |
effect=turn_of_month, holding_days=5, n_stocks=10, cost_bps=10, regime_gate=True r-calendar-effects-jp-all-0119 | effect: "turn_of_month" cost_bps: 10 n_stocks: 10 | -0.73 | +0.06 | -0.67 | -2.19 |
effect=turn_of_month, holding_days=20, n_stocks=5, cost_bps=5, regime_gate=True r-calendar-effects-jp-all-0129 | effect: "turn_of_month" cost_bps: 5 n_stocks: 5 | +0.49 | -0.63 | +0.64 | +3.02 |
effect=turn_of_month, holding_days=20, n_stocks=10, cost_bps=10, regime_gate=True r-calendar-effects-jp-all-0135 | effect: "turn_of_month" cost_bps: 10 n_stocks: 10 | +0.43 | -0.41 | +0.57 | +0.77 |
effect=day_of_week, holding_days=1, n_stocks=5, cost_bps=5, regime_gate=False r-calendar-effects-jp-all-0144 | effect: "day_of_week" cost_bps: 5 n_stocks: 5 | -1.28 | -0.72 | -1.05 | -4.70 |
effect=day_of_week, holding_days=1, n_stocks=5, cost_bps=10, regime_gate=False r-calendar-effects-jp-all-0146 | effect: "day_of_week" cost_bps: 10 n_stocks: 5 | -1.87 | -1.40 | -1.61 | -5.23 |
effect=day_of_week, holding_days=1, n_stocks=20, cost_bps=5, regime_gate=False r-calendar-effects-jp-all-0152 | effect: "day_of_week" cost_bps: 5 n_stocks: 20 | -1.90 | -1.28 | -1.55 | -6.74 |
effect=day_of_week, holding_days=1, n_stocks=20, cost_bps=10, regime_gate=True r-calendar-effects-jp-all-0155 | effect: "day_of_week" cost_bps: 10 n_stocks: 20 | -3.12 | -6.09 | -2.20 | -7.42 |
effect=day_of_week, holding_days=5, n_stocks=5, cost_bps=10, regime_gate=True r-calendar-effects-jp-all-0163 | effect: "day_of_week" cost_bps: 10 n_stocks: 5 | -0.93 | -0.01 | -1.10 | — |
effect=day_of_week, holding_days=5, n_stocks=20, cost_bps=5, regime_gate=True r-calendar-effects-jp-all-0169 | effect: "day_of_week" cost_bps: 5 n_stocks: 20 | -0.94 | +2.37 | -0.89 | — |
Target universe
Tokyo Stock Exchange all listed (Prime/Standard/Growth), ~3747 stocks
Related strategies
source: qbuntu-analysis · kabu STATION REST API + J-Quants daily bars
computed asof: 2026-07-29T01:09:04