DA/RT positioning, hedge timing, and risk

Decision packets for the day-ahead and real-time power desk.

Metis reads the same grid, weather, outage, and price state your desk watches, then turns the forecast into a size, direction, hedge, and risk-bounded trade rationale.

Horizon

Hours to weeks

Markets

ISO hubs and nodes

Output

Blotter-ready calls

Proof path

Replay the desk call before you wire it in.

Run a season of market history through Metis, grade the calls against settled outcomes, and integrate only after the desk trusts the calibration.

  1. 01

    Start with one hub or node

    Pick the desk's hardest repeated call and replay a full season of DA/RT outcomes.

  2. 02

    Grade on settled P&L drivers

    Compare direction, magnitude, tail behavior, and avoided bad trades against the desk baseline.

  3. 03

    Integrate the packet

    Deliver into the blotter, ETRM, risk feed, API, or agent workflow once the desk trusts calibration.

Why it matters

The expensive miss happens before execution.

The desk is usually not short on forecasts. It is short on a repeatable way to turn conflicting signals into a sized position.

The published ISO forecast is fine on normal days and least useful on heat onset, scarcity, outage, and congestion regimes.

The highest-value edge is not a prettier chart. It is knowing when the curve has underpriced a regime shift and how much risk to take.

Twin runtime

Model the system. Forecast the grid. Plug the answer back in.

Metis builds a live twin around the decision surface you already operate: the assets, desk state, constraints, market context, and software systems. The runtime forecasts what changes next, scores feasible moves, and writes the packet back to the tools that execute.

01Model the book
02Forecast the grid
03Score candidate trades
04Write to desk systems
TwinGraphERCOT North book - system

ERCOT North book

8 signals9 flows

Weather regime

+9 F

Net load ramp

+8.4 GW

ERCOT North hub

$71/MWh

Constraint state

2 active

Metis runtime

+$18 edge

Risk limits

42% open

Trade packet

draft

ETRM writeback

ready

Integration surface

Around the systems energy teams already run.

Example integration targets, not partner badges: Metis can pull from data systems, read operational state, and return bids, setpoints, or decision packets through APIs and control interfaces.

Surface

Market data

Examples

ISO/RTO feeds, Yes Energy, weather and outage APIs

Metis reads

Prices, awards, outages, load, renewable shape

Metis returns

Forecast features and settlement trace

Surface

Site systems

Examples

EMS, SCADA, AVEVA PI, inverter telemetry

Metis reads

SOC, limits, alarms, site load, generation state

Metis returns

Advisory setpoints and operator rationale

Surface

Control and bidding

Examples

Fluence Mosaic, Wartsila GEMS, Tesla Autobidder

Metis reads

Bid state, controller constraints, execution windows

Metis returns

Dispatch or bid packet once approved

Surface

Asset operations

Examples

Power Factors, warehouse, maintenance systems

Metis reads

Availability, derates, work orders, performance history

Metis returns

Constraint updates and asset-level traces

Surface

Desk and risk

Examples

PCI, ION, ETRM, blotter and risk systems

Metis reads

Positions, contracts, limits, exposure

Metis returns

Trade packet, hedge frame, risk-bounded call

Surface

Automation

Examples

API, MCP, Python and agent workflows

Metis reads

Typed context and workflow state

Metis returns

Decision packets, audit trace, model outputs

Runtime

Metis turns market state into a tradeable recommendation.

The runtime reads desk feeds and Metis signals, forecasts the price distributions that matter, scores candidate positions against desk limits, and returns a call with drivers and a settlement trace.

01

Desk state

Current positions, limits, hub and nodal prices, load and net-load shape, generation stack, and internal curves.

02

Metis forecast core

Weather regimes, outage notices, scarcity signals, renewable ramps, congestion history, and pretrained time-series models.

03

Trade packet

Long or short, size, hedge, confidence, downside frame, drivers, and API delivery into blotter or risk tooling.

Decision loop

Set the day-ahead position and hedge before the market clears.

For tomorrow's peak block at this hub, do I go long or short day-ahead versus real-time, how big, and where do I set the hedge?

Data in

Yours

  • LMP by hub and node (energy, congestion, loss)
  • System load and net load, evening ramp shape
  • Generation stack, wind/solar, reserve margin

Metis adds

  • Weather across load and renewable zones
  • Outage and derate notices, scarcity alerts
  • Settled DA/RT prices, congestion history

Plus cross-ISO history and pretrained time-series models.

Metis decision core

  • 1Read
  • 2Forecast
  • 3Score
  • 4Recommend

Same core, every desk.

Decision packet

Why

Go long day-ahead at North across the HE18 to HE20 ramp (expecting RT to settle above DA), sized to a moderate fraction of the block limit, and hedge the wings with a short on the soft midday HE13 to HE14 block. A shape trade, not an outright long.

Medium-high on direction, lower on magnitude (scarcity tail)

Trader blotterETRM ticket · APIRisk system feedMCP / agent

Calls Metis scores

From forecast to control.

Set the DA position

Translate load, renewables, outages, and reserve margin into a directional view before the market clears.

Hedge the wings

Shape the position around soft intervals, congestion exposure, and scarcity tails instead of treating the block as one bet.

Size against risk limits

Express confidence and downside as a trade size the desk can actually put on.

Prove it on one hard call.

Bring one asset, hub, node, or desk baseline. Metis will replay it, grade the lift, and show where the runtime changes the decision.