DCCH SENTINEL

What comes preset

DCCH Sentinel comes with two strategies already set up: the cross and the Bollinger "eyes". Here is how they did, measured on the past with each option's real price, so you get an idea of how they would work. You can change them, turn them off or build your own.

Read this before trading. These results are simulated on past data: they are not real trades, they do not guarantee future results and they are not a recommendation or an invitation to trade these strategies or these companies. The companies and exits come preset because that is what measured best, but everything is editable. Simulated results are calculated after the fact and do not include everything that happens live (delays, fills, data outages). With options you can lose everything you invest. Trading is your decision and your risk.

Real history, not a sample. Every number comes from trades the strategy would have made on past days, with the option's real price at that second: it buys at the real ask at the exact time of the signal and closes at the exact time of the exit, at the real bid or at the TP price.

Tap a company (for example, SPY) in the tables and you will see every trade: the date, the contract, the time it entered with its real ask, the time it exited with its price, how it exited and what it made or lost.
Winners and losers alike: none were removed.

1 · The cross (CPM)

It enters when the 9 EMA crosses the Bollinger mid line (20) on 5-minute candles, from 9:45 to 16:00, with price on the right side of VWAP and the cross continuing. One entry per company per day.

Preset: SPY, MSFT and IWM · 2026

Side and days	Exit	January–May	June–October
CALL and PUT, every day	TP +8% / stop −25%	33 · 9.1 of 10 · +$355	22 · 9.5 of 10 · +$241
CALL and PUT, every day	TP +8% / stop −50%	31 · 9.0 of 10 · +$390	32 · 8.8 of 10 · +$147
PUTs only, Tuesday to Thursday	TP +10% / stop −50%	11 · 9.1 of 10 · +$59	8 · 10 of 10 · +$78

Entries · winners out of 10 · net result with 1 contract. Tap a company to see its trades.

What you should know

The companies were chosen with January–May only; June–October was not used to choose and serves as the check.
With a small TP and a wide stop you win small often and, when you lose, you lose more: MSFT's worst entry was −$227 with 1 contract.
Measured in 2026 only (January 2 to October 7). IWM had 19 entries: the smallest sample.
The cross on all 53 companies together did not pass 6.6 of 10 with any stop and lost money. That is why only SPY, MSFT and IWM carry their measured exit; if you add others, they use the general TP and stop from Settings.

2 · Bollinger "eyes" (BB)

In the morning, a 15-minute candle that opens inside the Bollinger Bands (20, 2): the bands open up, the candle fakes against and then breaks in favor. It enters on the break only if volume is strong (1.5 times normal for CALLs, 1.2 times for PUTs). The stop is on the stock (the fake-out + 30% of the band width); at +22% it locks in +15% and then trails 20% (TSLA has its own ladder). It closes at 15:55.

Preset: TSLA, SPY, PLTR, COIN, QQQ, IWM, AMZN and META (CALL only)

Since October 17, 2026 it also has the turnaround (if after the first entry the stock leaves the Bollinger on the other side, it buys that side too, once a day per company) and the premium net (on top of the stock stop, if the option stays at −70% for ten seconds in a row before reaching +22%, it is sold). Since October 21, 2026 it has the clean floor: if the option jumps below the locked-in floor, it does not sell in that second; it waits up to 5 minutes for it to come back to the floor and sells there (if it does not come back, it sells at market). The tables below are for the 7 without those changes.

Year	All 7 together
2026 (through September 28)	142 · 8.4 of 10 · +$6,586
2025	190 · 7.5 of 10 · +$1,656

Company	2025	2026
	23 · 7.8 of 10 · +$329	16 · 9.4 of 10 · +$863
	27 · 8.5 of 10 · +$586	16 · 8.8 of 10 · +$637
	32 · 6.6 of 10 · −$188	28 · 9.6 of 10 · +$2,094
	22 · 6.4 of 10 · +$196	20 · 7.0 of 10 · +$382
	27 · 7.4 of 10 · +$253	16 · 8.1 of 10 · +$541
	35 · 8.6 of 10 · +$726	18 · 8.3 of 10 · +$588
	24 · 7.1 of 10 · −$249	28 · 7.5 of 10 · +$1,481

Entries · winners out of 10 · net result with $300 per entry (first out-of-the-money contract). Tap a company to see its trades.

What you should know

2026 is its best year. In 2025 it won 7.5 of every 10 and made little (+$1,656 over 190 entries), with PLTR and AMZN in the red.
META left on October 16, 2026 and came back on the 17th CALL only: its CALLs won 8.8 of every 10 in 2025 and 2026 (24 entries, +$406); its PUTs, 6.1 of every 10 and in the red.
With META on CALL, the turnaround and the net, from January 2, 2025 to October 16, 2026: 365 entries, 7.8 of every 10, +$7,926 (before 338 entries, 7.8 of every 10, +$7,368); the worst entry goes from −$280 to −$227. The turnaround happened only 3 times in that period.
With real prices from 2022 to 2024, under almost the same rules, it lost money. Doing well in 2026 does not mean it will keep doing well.
It signals roughly one day in three. Because the stop is on the stock, a losing entry can lose 40% to 60% of what the option cost.

3 · Both together and an account example

The cross and the BB together, in 2026: 267 entries, 8.7 of every 10 won, with an entry on about 3 of every 4 market days.

Example computed with that history (January 2 to September 28, 2026, real bid and ask): a $2,000 starting account, $300 entries and $100 more per entry after every week that ended green. By September 28 the account reached about $53,692 (35 green weeks and 4 red; the entry ended at $3,800).

2026 was a favorable year. In 2025 the same calculation came out much worse: with this step the account ran out of money.
The calculation does not apply daily entry caps; with large entries, the real fill price can be worse than measured.
This is not a recommendation and we are not encouraging you to trade: these are statistics of the past computed with real market data. The decision and the risk are yours.

4 · How to test your own strategy

Build it in the strategy builder (dcch.us → Strategies): from blocks or in your own words, and the AI builds it. It is measured with minute-by-minute history for 53 stocks and ETFs since 2024, buying at the real price and paying fees.
Don't trust a single period. Choose the companies and the exit with one part of the time (for example, January–May) and check it on another part you did not use (June–October). That is how the cross was done: whatever scored 10 of 10 in January–May and fell to 5–7 of 10 afterwards was left out.
Look at both numbers together: how many of every 10 win and the net result. A 9 of 10 with a wide stop can lose money if the losses are big.
Practice in the Blind Replay (estadisticas.dcch.us): past days, candle by candle, without seeing what comes next.
Paper test in the app: trade simulated with real prices. Recommended: at least three weeks and 20 entries before using real money.

How it was measured

The same rules the app uses, on second-by-second history, with real quotes (bid and ask) from the options market: bought at the real ask of the signal's second, sold at the real bid of the exit's second (or at the TP price with a limit sale), and $0.65 per contract per side in fees.