Strategy Analytics

Corn Donchian Breakout

Pure Donchian channel breakout optimization on Corn futures — which lookback N fits the market's history best.

Instrument

Corn Futures (ZC=F)

History

1978-01-04 → 2026-09-22

Lookback Grid

N = 20–80 step 3, then 85, 90, 95, 100

Profit Unit

price points (cents per bushel, 1 contract)

Total Profit By Lookback N

System A (Long Only) vs System B (Long & Short, stop-and-reverse). Price points per 1 contract.

System B — Long vs Short Profit Split

The same stop-and-reverse system, with the long leg and short leg reported separately (long + short = total).

Results By Lookback

Entry: close above the N-period high. System A exits to flat on a close below the N-period low; System B reverses into a short there (and back to long on the next upper-band break). No ATR stops, volatility filters, or profit targets. 12,273 daily bars. Generated 2026-09-24 13:34:12 UTC.

N	A: Profit	A: Trades	A: Win Rate	B: Profit	B: Long P&L	B: Short P&L	B: Trades	B: Win Rate
20	+130.0	134	33%	+793.0	+130.0	+663.0	268	41%
23	+202.8	121	31%	+947.0	+202.8	+744.3	241	41%
26	+271.8	102	31%	+1085.0	+271.8	+813.3	203	42%
29	+199.0	95	32%	+939.5	+199.0	+740.5	189	43%
32	+146.5	90	30%	+834.5	+146.5	+688.0	179	42%
35	+78.3	84	31%	+698.0	+78.3	+619.8	167	41%
38	+78.0	79	32%	+697.5	+78.0	+619.5	157	41%
41	+140.0	73	34%	+821.5	+140.0	+681.5	145	43%
44	+202.3	68	35%	+952.5	+202.3	+750.3	135	43%
47	+133.5	66	30%	+820.8	+133.5	+687.3	131	40%
50	+130.0	62	31%	+823.5	+130.0	+693.5	123	39%
53	+221.5	59	32%	+1010.8	+221.5	+789.3	117	40%
56	+98.5	58	31%	+764.8	+98.5	+666.3	115	40%
59	+164.0	56	32%	+901.0	+164.0	+737.0	111	41%
62	+99.3	55	31%	+771.5	+99.3	+672.3	109	40%
65	-15.3	55	29%	+542.5	-15.3	+557.8	109	39%
68	+49.3	52	31%	+671.5	+49.3	+622.3	103	43%
71	-21.3	51	29%	+530.5	-21.3	+551.8	101	42%
74	+70.8	48	29%	+714.5	+70.8	+643.8	95	41%
77	+37.0	48	27%	+647.0	+37.0	+610.0	95	40%
80	+122.0	45	29%	+817.0	+122.0	+695.0	89	43%
85	+134.3	43	30%	+841.5	+134.3	+707.3	85	44%
90	+97.5	42	31%	+768.0	+97.5	+670.5	83	43%
95	+136.3	38	34%	+845.5	+136.3	+709.3	75	48%
100	+38.3	37	32%	+655.3	+38.3	+617.0	73	47%

Monthly Donchian — Does A Slower Channel Beat The Daily One?

The same pure channel logic on monthly bars (open = the month's first open, high/low its extremes, close its last close), lookback N = 3–20 months over 585 monthly bars. Note the horizons are not the same thing: the daily grid's 20–100 bars is roughly 1–5 months, while 3–20 monthly bars is 3–20 months, so this is a far slower channel rather than the same channel on other data. Sharpe and CAGR annualize with 12 periods here against 252 on the daily grid.

Answer: no — and not close. Median Sharpe across the grid is -0.096 on monthly bars against 0.039 on daily, and 7 of 36 monthly cells have a positive Sharpe at all (the daily grid manages 31 of 50). The best monthly cell is N = 3 at Sharpe 0.08, against the daily grid's best of 0.30 at N = 26.

The reason matters more than the result. A monthly channel of this length simply cannot generate enough trades on 26 years of data — the median monthly cell takes 19 trades in total, and the best one takes 61. Those numbers cannot support a conclusion in either direction, so the honest reading is not “monthly breakouts do not work” but “this history is far too short to test a 3–20 month channel”. Judging a strategy on four trades is how you end up fitting noise.

N (months)	A: Profit	A: Trades	A: CAGR	A: Max DD	A: Sharpe	A: Sortino	A: Calmar	B: Profit	B: Trades	B: CAGR	B: Max DD	B: Sharpe	B: Sortino	B: Calmar
3	+139.3	31	-1.2%	-79.1%	-0.07	-0.11	-0.02	+858.8	61	1.9%	-65.2%	0.08	0.13	0.03
4	+115.0	29	-1.5%	-80.5%	-0.09	-0.14	-0.02	+810.3	57	1.2%	-66.7%	0.06	0.09	0.02
5	+75.0	22	-1.5%	-84.8%	-0.09	-0.14	-0.02	+681.8	44	1.0%	-79.4%	0.04	0.07	0.01
6	+45.5	19	-1.1%	-80.0%	-0.07	-0.11	-0.01	+622.8	38	1.1%	-80.2%	0.05	0.08	0.01
7	-52.0	17	-1.8%	-82.4%	-0.11	-0.17	-0.02	+427.8	34	-0.2%	-84.4%	-0.01	-0.02	-0.00
8	+60.0	16	-1.4%	-78.0%	-0.09	-0.13	-0.02	+670.3	31	0.2%	-76.9%	0.01	0.02	0.00
9	+103.0	14	-1.5%	-78.0%	-0.09	-0.14	-0.02	+758.5	27	0.4%	-75.0%	0.02	0.03	0.01
10	+9.8	14	-2.1%	-79.7%	-0.12	-0.18	-0.03	+599.8	27	-0.4%	-76.2%	-0.02	-0.03	-0.00
11	+95.5	12	-1.6%	-80.6%	-0.10	-0.15	-0.02	+771.3	23	0.9%	-81.6%	0.05	0.07	0.01
12	-3.3	12	-2.3%	-86.3%	-0.15	-0.21	-0.03	+573.8	23	-0.6%	-90.3%	-0.03	-0.04	-0.01
13	-22.5	11	-2.3%	-85.0%	-0.14	-0.21	-0.03	+480.0	22	-0.7%	-90.3%	-0.04	-0.05	-0.01
14	-95.0	11	-2.7%	-87.4%	-0.17	-0.24	-0.03	+335.0	22	-1.5%	-91.3%	-0.08	-0.11	-0.02
15	-295.5	10	-3.7%	-91.5%	-0.24	-0.33	-0.04	-66.0	21	-3.7%	-96.4%	-0.18	-0.25	-0.04
16	-354.8	10	-4.1%	-92.7%	-0.26	-0.36	-0.04	-184.5	21	-4.3%	-97.4%	-0.21	-0.29	-0.04
17	-430.3	9	-4.1%	-92.0%	-0.27	-0.36	-0.05	-335.5	19	-3.9%	-95.4%	-0.19	-0.26	-0.04
18	-362.5	8	-3.6%	-88.7%	-0.23	-0.32	-0.04	-218.0	17	-2.8%	-90.9%	-0.14	-0.19	-0.03
19	-362.5	8	-3.6%	-88.7%	-0.23	-0.32	-0.04	-218.0	17	-2.8%	-90.9%	-0.14	-0.19	-0.03
20	-425.5	8	-3.9%	-90.6%	-0.26	-0.35	-0.04	-344.0	17	-3.5%	-93.4%	-0.17	-0.23	-0.04

Monthly bars 1978-01-31 → 2026-09-30. Profit is price points per contract, directly comparable with the daily table above since it is the same instrument and contract size. Trade counts are not comparable — that is the point.

EMA-170 Direction Filter On The Breakout

The same daily channel and the same N grid, with one extra precondition: a long may only open while the 170-period exponential moving average is rising, and a short only while it is falling. The band break is still required on top — the filter removes trades, it never creates them. Exits are untouched, so a trade opened with the trend still closes on the opposite band break even if the EMA has since turned; otherwise the filter would be changing two things at once and this comparison would not isolate it. The unfiltered grid stays above so “does it help” has something to be measured against.

On Corn: no, it makes things slightly worse. Across the 50 cells the filter beats the unfiltered channel on Sharpe in 42 and on Calmar in 42. Medians move the wrong way on all three measures: Sharpe 0.055 against 0.039, Sortino 0.078 against 0.055, Calmar 0.013 against 0.010.

Worth reading with one thing in mind, though: Corn has been the weakest market on this site all along — its base trend system sits near a zero median Sharpe, so there is little for a filter to improve. The same filter is applied to Cocoa, Coffee and Crude Oil on their own cards, and those are the more informative tests.

N	Direction	Trades (plain → EMA)	Sharpe plain	Sharpe EMA	Sortino plain	Sortino EMA	Calmar plain	Calmar EMA	Better?
20	A — long only	134 → 89	-0.01	0.03	-0.01	0.05	-0.00	0.01	EMA
20	B — long & short	268 → 208	0.16	0.23	0.22	0.33	0.05	0.07	EMA
23	A — long only	121 → 83	0.01	0.04	0.02	0.06	0.00	0.01	EMA
23	B — long & short	241 → 191	0.20	0.26	0.28	0.36	0.06	0.07	EMA
26	A — long only	102 → 76	0.08	0.06	0.12	0.09	0.02	0.01	plain
26	B — long & short	203 → 169	0.30	0.30	0.43	0.42	0.08	0.07	plain
29	A — long only	95 → 70	0.04	0.04	0.05	0.06	0.01	0.01	EMA
29	B — long & short	189 → 158	0.24	0.26	0.34	0.37	0.07	0.07	EMA
32	A — long only	90 → 67	0.01	0.03	0.02	0.05	0.00	0.01	EMA
32	B — long & short	179 → 151	0.20	0.23	0.28	0.33	0.06	0.06	EMA
35	A — long only	84 → 65	-0.01	0.02	-0.01	0.03	-0.00	0.00	EMA
35	B — long & short	167 → 144	0.16	0.18	0.22	0.26	0.04	0.05	EMA
38	A — long only	79 → 63	-0.02	0.01	-0.03	0.01	-0.00	0.00	EMA
38	B — long & short	157 → 138	0.14	0.16	0.20	0.22	0.04	0.04	EMA
41	A — long only	73 → 60	0.02	0.02	0.03	0.02	0.00	0.00	plain
41	B — long & short	145 → 129	0.20	0.21	0.28	0.30	0.05	0.06	EMA
44	A — long only	68 → 58	0.01	0.00	0.01	0.01	0.00	0.00	plain
44	B — long & short	135 → 124	0.19	0.17	0.27	0.24	0.06	0.05	plain
47	A — long only	66 → 56	-0.01	-0.01	-0.02	-0.01	-0.00	-0.00	EMA
47	B — long & short	131 → 120	0.16	0.16	0.23	0.22	0.05	0.04	plain
50	A — long only	62 → 53	-0.03	-0.02	-0.04	-0.02	-0.01	-0.00	EMA
50	B — long & short	123 → 114	0.15	0.13	0.21	0.19	0.04	0.03	plain
53	A — long only	59 → 51	-0.01	0.00	-0.01	0.00	-0.00	0.00	EMA
53	B — long & short	117 → 109	0.19	0.18	0.27	0.25	0.06	0.05	plain
56	A — long only	58 → 50	-0.05	-0.03	-0.07	-0.04	-0.01	-0.01	EMA
56	B — long & short	115 → 107	0.12	0.12	0.17	0.17	0.04	0.04	EMA
59	A — long only	56 → 48	-0.05	-0.02	-0.06	-0.03	-0.01	-0.00	EMA
59	B — long & short	111 → 103	0.13	0.14	0.19	0.20	0.04	0.05	EMA
62	A — long only	55 → 48	-0.07	-0.05	-0.10	-0.08	-0.01	-0.01	EMA
62	B — long & short	109 → 102	0.10	0.11	0.14	0.15	0.03	0.03	EMA
65	A — long only	55 → 48	-0.12	-0.09	-0.16	-0.12	-0.02	-0.02	EMA
65	B — long & short	109 → 102	0.03	0.05	0.05	0.07	0.01	0.01	EMA
68	A — long only	52 → 45	-0.08	-0.05	-0.11	-0.07	-0.01	-0.01	EMA
68	B — long & short	103 → 96	0.09	0.10	0.12	0.15	0.03	0.03	EMA
71	A — long only	51 → 45	-0.11	-0.08	-0.15	-0.11	-0.02	-0.01	EMA
71	B — long & short	101 → 95	0.04	0.06	0.06	0.09	0.01	0.02	EMA
74	A — long only	48 → 43	-0.07	-0.05	-0.10	-0.07	-0.01	-0.01	EMA
74	B — long & short	95 → 90	0.08	0.09	0.11	0.13	0.02	0.03	EMA
77	A — long only	48 → 43	-0.09	-0.06	-0.13	-0.08	-0.02	-0.01	EMA
77	B — long & short	95 → 90	0.06	0.08	0.08	0.11	0.02	0.02	EMA
80	A — long only	45 → 41	-0.06	-0.04	-0.09	-0.06	-0.01	-0.01	EMA
80	B — long & short	89 → 85	0.11	0.12	0.15	0.17	0.03	0.03	EMA
85	A — long only	43 → 39	-0.07	-0.03	-0.09	-0.04	-0.01	-0.01	EMA
85	B — long & short	85 → 81	0.11	0.13	0.16	0.19	0.03	0.04	EMA
90	A — long only	42 → 38	-0.08	-0.04	-0.11	-0.06	-0.01	-0.01	EMA
90	B — long & short	83 → 79	0.09	0.12	0.13	0.17	0.03	0.04	EMA
95	A — long only	38 → 37	-0.07	-0.06	-0.09	-0.08	-0.01	-0.01	EMA
95	B — long & short	75 → 74	0.12	0.12	0.17	0.18	0.04	0.04	EMA
100	A — long only	37 → 36	-0.10	-0.09	-0.14	-0.13	-0.02	-0.02	EMA
100	B — long & short	73 → 72	0.06	0.07	0.09	0.10	0.02	0.02	EMA

“Better?” compares Sharpe at the same N and direction. Sortino divides by downside deviation only, so it does not penalise the large winning days a breakout system is built to catch; Calmar is CAGR over the worst peak-to-trough drawdown.