Intraday mean reversion
Bounce
Buy the tick off the bottom, cap loss with a stop, trail the recovery, and review every fill.
price pathtrail active
Midas
Screen quant ideas. Backtest what survives.
A single-user strategy lab for turning filters into Bounce or Gravity systems, then testing them with assumption-aware fills before risking a dollar.
3,430
stocks
13F
signals
Paper
only
Strategy Lab
paper / backtest only
No live execution promises. No hidden optimistic assumptions.
Built for traders who distrust pretty equity curves.
From 3,430 names to one testable idea.
Screen by sector, volume, market cap, moving averages, and list membership. Save the result as a strategy universe.
Morning Bounce v1
rules tab
Universe
Saved screen: liquid US names
Entry
Drop greater than 6%, first recovery tick
Exit
Trailing stop follows the bounce
Risk
Initial stop 1.05% below entry
Every strategy is a set of rules you can inspect.
Bounce is not a black box. Entry, exit, risk, universe, and trailing-stop behavior all sit in the open before the run begins.
The summary is only the beginning.
Every result resolves into dates, tickers, entries, exits, P/L, and outcome tags. Pretty curves do not get to hide the trades.
Trades
auditable ledger
| Date | Ticker | Entry | Exit | P/L | Outcome |
|---|---|---|---|---|---|
| 2026-06-18 | AVGO | $1,734.20 | $1,759.80 | +1.48% | trail exit |
| 2026-06-18 | CRWD | $389.10 | $384.95 | -1.07% | stop |
| 2026-06-17 | NVDA | $142.80 | $146.32 | +2.46% | trail exit |
| 2026-06-17 | TSLA | $327.40 | $324.02 | -1.03% | stop |
One loop from market noise to a testable system.
Midas keeps the work inside one surface: screen the universe, define the rules, backtest honestly, then watch the strategy in paper mode.
01
Screen
Filter liquid US names by list, sector, volume, volatility, and moving averages.
02
Define
Turn a saved screen into a Bounce or Gravity strategy with explicit entry, exit, and risk.
03
Backtest
Model commission, slippage, intrabar stops, and filing-date-aware signals before you trust it.
04
Paper monitor
Track passing strategies in a paper account view. Live execution stays gated.
Gravity timeline
signal to universe
T+0
13F filed
No look-ahead signal enters the test.
T+1
CIK resolved
Recorded before simulation.
T+2
Holdings dated
Recorded before simulation.
T+3
Screen saved
Recorded before simulation.
T+4
Backtest queued
Recorded before simulation.
Designed against the backtest traps.
The page is calm because the product is skeptical. Every attractive curve needs assumptions, dated signals, drawdowns, and a reason to survive out of sample.
Optimistic fills
Commission, slippage, and intrabar stop behavior are modeled by default.
Cherry-picked curves
Losers, drawdowns, trades, and assumptions stay visible in the review surface.
Look-ahead signals
13F and X inputs become dated signals before they enter a backtest.
Black-box verdicts
The scorecard shows why a strategy passed, warned, or failed.
Two starting points. Both inspectable.
Bounce watches sharp intraday overreactions. Gravity turns public signal sources into a dated, backtestable universe.
13F and X signal feed
Gravity
Resolve a fund, date the holdings, fold in tracked accounts, and test the resulting pressure.
1Situational Awareness
dated2@aleabitoreddit
dated3@CKCapitalxx
datedRuns on the data you already pay for.
Connect the providers that power the loop. Broker surfaces are paper/connect UI only in v1.
source
Polygon
source
financialdatasets.ai
source
SEC EDGAR
source
twitterapi.io
source
Alpaca paper
source
Robinhood UI
Straight answers.
Midas is deliberately narrow in v1: one user, honest testing, paper monitoring, no live order execution.
Start with the assumptions visible.
Open Midas, build a screen, and make the strategy prove itself before it earns your attention.