00 / Relay command center

Sign in to the
signal path.

RELAY PAPER DRY RUN
Live Performance Backtest Accounts
--:--:-- local --:-- Chicago

  1. 01 TradingViewTV -- --
  2. 02 TunnelTunnel -- --
  3. 03 RelayRelay -- --
  4. 04 IB GatewayGateway -- --
  5. 05 IBKRIBKR -- --

02 / Book · today

--

Daily P&L

Net liquidation · this session
Net liq
--
Unrealized
--
Realized
--
Available
--
Margin
--

03 / Control

Orders today

0 of 0

Today's P&L

-- limit off

Manual target

Runs through the same risk checks as a TradingView alert. Respects halt and dry run.

Cancels working orders and closes every position at market. Works even when halted or in dry run.

04 / Positions

Instrument Position Avg Mark Unrealized Roll in Actions

05 / Instruments & limits

Instrument On Max pos Max order

Changes apply instantly and survive restarts. A disabled instrument rejects every alert, including exits.

06 / Tape

    07 / Fills · this session

    Time Contract Side Qty Price Realized
    --

    08 / Net P&L · --

    --

    --

    09 / Calendar

    10 / Equity curve · net, cumulative

    11 / Daily P&L

    12 / Trades

    Closed Instrument Side Qty Entry Exit Hold Net

    13 / Breakdown

    Group Trades Win % Net

    14 / Backtest · MGC on IBKR history

    Loading the latest report…

    No backtest report on the server yet

    1. On the Mac, run the backtest with --json backtest/data/report.json.
    2. Upload that file to the server as data/backtest/report.json. The command is in the Runbook under “Backtest strategies”.
    3. Reload this tab. The bot only reads the file; it never runs a backtest itself.

    15 / Strategies · whole period, 1 contract, real costs

    Strategy TradingView net Real net Months up Worst 6 mo Trades Per day Win PF Max DD Costs Can't open from

    16 / Picking by backtest · tuned, then unseen

    Family Settings Made money tuned Made money unseen Beat holding unseen Best when tuning Tuned net Then unseen

    17 / Every setting tried · unseen period

    Setting Tuned net Unseen net Vs holding Trades Win PF Max DD Older yrs

    18 / Backtest vs your real fills

    19 / Strategy search · --

    Family Exits Settings 3+/day, 40%+ win Met targets Up every quarter Best net Best setting

    20 / Finalists on the unseen months · walk-forward

    Finalist Picked-on net Unseen net Unseen, 2-tick slip Unseen win Unseen per day All bars Targets unseen

    Walk-forward: each month, trade the best setting of the last N months

    Lookback Months traded TradingView net Net Per day Win Unseen months net

    21 / Every setting tested

    Setting TradingView net Net Unseen net Older yrs net Per day Win PF Max DD Quarters

    22 / Older years · spot gold, --

    a second unseen test

    Every setting Tested Made money Met targets Median net Median before costs Rank corr. with picked-on Picked-on winners: made money … median net

    Every round's finalists

    Finalist Picked-on Unseen Older yrs Year 1 Year 2 Older, 2-tick slip Per day Win

    Headline strategies

    Strategy TradingView net Net Months up Worst 6 mo Trades Per day Win PF Max DD Can't open from

    23 / Steady-curve search · --

    Finalist Picked on, per year Months up Worst 6 mo Straightness Never seen, per year Months up Worst 6 mo Straightness Never seen, 2-tick slip

    Families, on the years picked on

    Family Settings Made money Worst 6 mo above $0 Steadiest: per year Worst 6 mo Steadiest setting

    The 40 steadiest settings, on the years picked on

    Setting Per year Months up Worst 6 mo Worst 3 mo Straightness Max DD Per day

    Long from each New York hour: dollars a year, by hours held (no stop, no filter)

    24 / Accounts

    Each account has its own IBKR connection, history, limits and webhook

    25 / Add an account

    New accounts start halted and in dry run. Each IBKR login needs its own IB Gateway running on the server. A live account also has to be confirmed below before it can send a single order.