daxsnack is the fully automated trading stack i built for myself: it refreshes market data, reruns models, and turns the daily search into one setup i am comfortable backing.
under the hood, finance logic meets production software: candidates are ranked, risk is sized, and robustness is checked with sharpe-based scoring, holdout tests, drawdown limits, monte carlo simulation, and stability filters before anything goes live.
the result is not a stream of tips. it is one disciplined output per day: an automated trade plan with clear entry, clear stop, and readable context, built to favor consistency over noise.
start guide
daxsnack automates the search, not the responsibility. the system scans, ranks, and filters; you still decide whether the setup fits your account and your discipline.
think of it as a live decision engine: data refresh, model checks, risk caps, and portfolio filters run in the background so the final card stays compact.
your job is execution. use the same market if possible, respect the published stop, keep size consistent, and be prepared to do nothing when daxsnack does nothing.
execution environment
to match prices and order handling as closely as possible, use a free capital.com account. the same venue means fewer gaps between the published plan and your execution.
if you already have an account there, you can follow the setup immediately.
what the card shows
each setup card is built from a simple market-structure view: trend, signal bar, entry level, and invalidation point.
entry comes from the signal bar close or trigger logic. the stop sits at the opposite side of the signal bar or the structure break that would invalidate the trade.
volume and follow-through are used as confirmation and exit context, not as decoration.
- High
- Close
- Open
- Low
- High
- Open
- Close
- Low
risk framework
each plan is sized so a stop-out is limited to 1% of account equity. wider stops automatically reduce size.
open positions are capped at 4. if exposure is already full, daxsnack stands aside instead of forcing another trade.
notional exposure per trade is capped at 20% of equity, so the system stays liquid even when several signals cluster.
getting started
- read the start guide
-
subscribe to the setup of the day (free)
- then scroll to today’s plan ↓
setup dashboard
from automated market scan to one executable plan
every day, daxsnack pulls fresh prices across the full universe, reruns the active strategy stack, and narrows thousands of theoretical combinations down to the few candidates worth serious attention.
those candidates are then checked like software in production, not like marketing copy: risk is sized, holdout performance is reviewed, drawdowns are capped, monte carlo paths are simulated, and unstable models are filtered out.
only after that does daxsnack compare the remaining instruments as a portfolio, account for skipped entries and live constraints, and publish the single plan that survives both the financial and technical checks.
349instruments tracked
8strategies live
2,792daily combinations
frequently asked questions
daxsnack starts with automation, not opinion. it refreshes market data, reruns the active models across the tracked universe, and reduces thousands of theoretical combinations to a much smaller candidate set.
that short list is then filtered much harder: risk is sized, holdout performance is reviewed, drawdowns are capped, monte carlo paths are simulated, and unstable models are rejected before anything is considered for publication.
only the survivors are compared as a portfolio. that means skipped entries, live constraints, and competing signals are considered before one final plan is published.
the public card shows the result of that pipeline, not the full internal recipe.
because “no trade” is a valid output. daxsnack is built to prefer silence over a weak signal.
if current candidates fail the risk framework, recent quality checks, stability checks, or portfolio constraints, nothing is promoted just to fill the slot.
this matters more than constant activity. a system that is willing to stand aside is usually more trustworthy than one that must always produce a trade.
when there is no setup, daxsnack is still doing its job: filtering noise, protecting capital, and waiting for conditions that are worth acting on.
daxsnack does not promote a model just because one backtest looks good. candidates are checked against holdout data, drawdown limits, recent behavior, and monte carlo path stress before they are allowed to stay in the live stack.
the point is simple: high returns without stability are not enough. the system is built to prefer models that keep their shape when conditions change.
the search, ranking, and filtering are automated. execution is still yours. that means matching the market if possible, respecting the published stop, choosing the right account size, and deciding whether the setup fits your own rules.
daxsnack is designed to compress the decision process, not to remove responsibility from the person placing the trade.