GammaBusters · PUBLIC SUMMARY · 2026-09-10 00:00 UTC

The C2 track record

What a subscriber's automation actually trades, measured on its own rules.

Get the signals — $299/mo Annual — save $598

C2 sells one SPX 0DTE vertical credit spread per signal, collecting a fixed net-credit range with a cheap deep-out-of-the-money protective wing. The stop is a StopMarket at 2.15× net credit (~3.15× the short leg). There is no profit target and no timed exit — a position is held to cash settlement unless the stop fires. Signals fire during regular trading hours only, and nothing is held overnight. Every figure on this page is stated per 1 contract, which is what the distributed trade rules specify (QtyDefault: 1) — the same rules whichever delivery you choose.

BacktestBacktest

$43,881
Net P/L, 1 lot
4,116 trades
76.0%
Win rate
3,127 winners
11.7%
Premium capture
on $376,235 collected
-$5,086
Max drawdown
-11.6% of peak
$10.66
Avg / trade
23.8% stopped
HYPOTHETICAL PERFORMANCE DISCLOSURE. The backtest figures on this page are simulated results, not actual trading. Simulated results have inherent limitations: they are prepared with the benefit of hindsight, do not reflect actual market liquidity or the impact of order flow, and no representation is being made that any account will or is likely to achieve profits or losses similar to those shown. The live-results section below reflects real fills in one account over a short period — individual results vary with broker, fills, timing and account size. All figures are stated per 1 contract and exclude the subscription fee, Trade Automation Toolbox fees, market-data fees and your own commissions. Past performance does not guarantee future results. Nothing on this page is investment advice. See Terms.

2022-05-16 to 2026-09-08 · 1,010 trading days · 1 lot · NFP days excluded · production live_now gate config · stop fills modelled at $0.10 of slippage. Our own live stops have averaged $0.21, at which this record reads $24,261 rather than $43,881 — see Execution sensitivity below before quoting the figure above.

YearTradesDaysP/L (1 lot) Win rateStoppedPCRAvg / trade
2022655152$10,10377.3%22.6%16.8%$15.42
2023917233$14,87677.4%22.4%17.6%$16.22
2024943235-$2,22071.9%27.8%-2.6%-$2.35
2025900230$12,31477.2%22.7%15.2%$13.68
2026701160$8,80876.7%23.1%13.9%$12.57
All4,116 $43,88176.0% 23.8%11.7% $10.66
Execution sensitivity — what the slippage assumption does to these numbers
Read this before the headline number. Roughly 24% of trades end at the stop. The backtest fills those at the stop price plus a fixed slippage allowance, and because the effect is linear in (slippage × number of stops), every extra $0.20 of real-world slippage removes about $19,000 per lot from the record. This single assumption swings the result by more than the strategy's entire profit.
Assumed stop slippageTradesP/L (1 lot)PCR Max drawdownAvg / trade
$0.10 ← headline · engine default4,116$43,88111.7%-$5,086$10.66
$0.30 ≈ our measured $0.214,116$24,2616.4%-$9,139$5.89
$0.504,116$4,6411.2%-$14,679$1.13

We can check this against reality, at least a little. Across the 38 stops taken live in the strategy creator's own account, the gap between where the stop was placed and where it actually filled was: mean $0.21, median $0.07, best $-0.20, worst $2.10. 7 of 38 (18%) came in worse than $0.10, 5 (13%) worse than $0.30, and 4 (11%) worse than $0.50.

Those last figures are not a typo: the counts past $0.10 and past $0.30 are the same because nothing landed in between. A stop either filled at essentially the modelled price or it blew straight through to $0.50–$1.70. That bimodality is the practical risk — not a gentle few cents of drift on every stop, but a small number of stops that cost several times what was expected. It is also why an average is the honest input and a median would mislead.

Set that against the headline. The figures at the top of this page use $0.10, the backtest engine's default. Our own measured mean is $0.21 — which corresponds to the $0.30 row above, or $24,261 rather than $43,881. The median stop does fill close to the modelled value, so $0.10 is not a fiction; but P/L is linear in total slippage, so the mean is what maps onto dollars, and the mean is dragged by a long right tail. If you want the figure this record would carry under our own observed execution, read the $0.30 line, not the top of the page.

Both caveats cut the same way and are worth stating: 38 stops is a very small sample, and that mean is sensitive to the single $2.10 outlier within it.

Per year, across the band

Year$0.10$0.30$0.50
2022$10,103$7,143$4,183
2023$14,876$10,776$6,676
2024 loses at every level-$2,220-$7,460-$12,701
2025$12,314$8,234$4,154
2026$8,808$5,568$2,328

Identical trades in every column — same entries, same strikes, same stop triggers. Only the assumed fill on a stopped spread changes.

Read the worst year, not the total. 2024 returned -$2,220 across 943 trades (-$2.35 per trade, 71.9% win rate). A multi-year headline hides that. Any year on this table is a plausible year to start in.
Drawdown curve

Live resultsLive

The strategy creator runs the identical 1-lot strategy in his own brokerage account, live since 2026-07-07. This is a 43-trading-day sample — far too short to be a track record on its own. It is here as a reality check on the backtest, nothing more.

$2,772.29
Net P/L, 1 lot
151 spreads
78.1%
Win rate
vs 76.0% backtest
16.5%
Premium capture
vs 11.7% backtest
$551.71
Commission paid
already deducted
43
Trading days
2026-07-08 to 2026-09-09
How the live number is derived
How the live number is derived. The live figures are reconstructed from the actual fill price of every leg traded in the creator's own brokerage account, settling every unstopped leg at intrinsic value against the official SPX close, net of the commission booked on each leg. This reconstruction has not been independently audited or reconciled against a broker statement.

The public Collective2 listingIndependent

The same strategy has been tracked on Collective2 since September 4, 2025 — a platform we do not control, so treat it as the independent cross-check on this page. As of 2026-08-30 it showed +10.4% cumulative, a 17.1% maximum drawdown, and an underwater stretch from Nov 20, 2025 – Aug 26, 2026. Context that chart cannot show: the listing's original data feed went dead in mid-December 2025 and no trades posted from then until late August 2026, when it was relinked to live brokerage fills — so the peak-to-valley dates to the Sep–Dec period on an earlier configuration, and most of the long underwater stretch is a dormant listing, not months of live losing. The drawdown itself is real, and we would rather you read it here than discover it there.

What this record does and does not include — the exclusions and why
  • Stop fills are modelled. Quantified in full above — at the published $0.10 assumption the record is $43,881; at $0.50 it is $4,641. This remains the largest single uncertainty here, and the live sample backing the chosen figure is only 38 stops.
  • No execution failures. The live book already shows what the backtest cannot: 35 rejected orders, 113 cancellations, 25 unwinds and 5 aborted spreads where the wing filled and the short did not. Subscribers have separately hit TAT margin rejections and expired entries against these exact templates.
  • Wing availability. The templates list fixed wing widths (100/125pt on puts, 50pt on calls). 3,304 of 4,116 trades (80%) found a wing at a listed width inside the $0.05–0.10 band; 812 (20%) needed the LongMaxWidth fallback to some other width.
  • On the most violent days the entry can simply fail. If no strike at a listed width prices inside the band, there is no qualifying wing and the trade does not open — TAT fails the entry rather than buying a wing at any price, and this record does the same. That happened on 3 days (2024-08-05, 2025-04-09, 2025-10-01). Small in count, but note what those dates are: they cluster on high-volatility sessions — exactly the days you would most want the protection. Expect the same behaviour live.
  • Put and call sides are not symmetric. The put template permits a short up to 20pt in the money with a 100/125pt wing; the call template requires out-of-the-money with a 50pt wing. Max loss per contract is roughly 2–2.5× larger on the put side for the same credit. This is almost certainly unintentional and is under review.
  • Coverage starts 2022-05-16 — a data-availability boundary, not a strategy start. There is no 2020–21 stress test, which for a short-premium strategy means the record has never seen a COVID-scale volatility event.
  • C2 re-entry is excluded (+$13,427 across the same period). The live engine cannot re-enter a stopped C2, so including it would inflate this record by roughly 25%.
  • Subscriber results will differ. Your fill prices, contract size and broker commissions are your own. The strategy creator legs into a spread manually with a credit floor and escalating limits; TAT sends a single spread order with its own retry ladder. Different mechanics, different fills.
  • Alternate gate configurations exist. Five were evaluated over this period; this page publishes only the one actually in production. The others are not shown and are not being selected from.
  • Manual (Telegram) execution is not modeled here. A manual trader enters minutes behind the signal, and the entry-timing sensitivity shown above means results are materially affected. This record assumes automated entry.
Alignment check — how this backtest relates to earlier ones

An earlier backtest of C2 used a different wing rule (nearest cheapest within 300pt) and a double-touch stop — neither of which matches what subscribers run. Over the same 3,955 overlapping trades that older method reports $54,260 at 76.3% win rate, against $43,881 at 76.0% here. The difference is the point: this page is generated from the shipped template rules so that the number describes the product. 18 MOC-session rows were dropped, since subscribers are gated off before the market-on-close window.

Convinced by the record?

Same signals, your account, cancel anytime. This page regenerates nightly from the same data.

← gammabusters.com

GammaBusters LLC · 30 N Gould St Ste 68150, Sheridan, WY 82801