Pinned to CAISO · View the full index
Grid Flexibility Index · V1 · 2026Q3
Where CAISO's flexibility is worth the most, and whether the grid can use it.
CAISO is Activated. Its Readiness is the highest of the seven, and its Opportunity is among the lowest.
Opportunity is how much a grid needs flexibility and what that flexibility is worth. Readiness is how far policy, markets, and regulation have developed to let that value be captured and flexibility scaled.
Quadrants: · · ·
Dot: the grid's score. Shaded area: the middle 80% of its simulated scores on each axis. Dashed lines: the quadrant boundary at 5. Axes show 2 to 8 of the 1 to 10 scale, and every grid's score and range falls inside that window.
Questions or feedback on the Index?
Grid Flexibility Index V1. Coming next: utility-level scores within each grid and customer-class lenses.
CAISO
PinnedActivatedOpportunity
Readiness
CertaintyHigh / Medium
Full ranking
Sort by any column. Click a grid to expand its pillars.
| Untapped | 7.01 | 4.81 | |||||
Pillar breakdown (1-10) Value8.89 Stress4.82 Momentum7.08 Infrastructure9.10 Rates2.60 Market access6.53 Programs3.76 Institutions4.43 Notes Quadrant consistencyUntapped in 66% of our sensitivity tests (Ready to scale in 34%). | |||||||
| Untapped | 6.29 | 4.86 | |||||
Pillar breakdown (1-10) Value7.76 Stress4.26 Momentum7.73 Infrastructure9.01 Rates4.24 Market access4.23 Programs5.34 Institutions3.13 Notes Quadrant lineMISO is within 0.14 of the quadrant line on the Readiness axis. Quadrant consistencyUntapped in 53% of our sensitivity tests (Ready to scale in 46%). | |||||||
| Untapped | 6.09 | 4.34 | |||||
Pillar breakdown (1-10) Value6.88 Stress4.88 Momentum6.75 Infrastructure8.65 Rates3.95 Market access3.25 Programs5.16 Institutions2.70 Notes Quadrant consistencyUntapped in 86% of our sensitivity tests (Ready to scale in 14%). | |||||||
| Untapped | 5.02 | 4.26 | |||||
Pillar breakdown (1-10) Value6.97 Stress2.80 Momentum6.37 Infrastructure3.88 Rates1.69 Market access6.53 Programs4.27 Institutions7.65 Notes Price historyUsing three years of prices (October 2023 through September 2026) instead of the last four quarters would place ISO-NE in Monitor. Quadrant lineISO-NE is within 0.02 of the quadrant line on the Opportunity axis. Quadrant consistencyUntapped in 64% of our sensitivity tests (Monitor in 28%). | |||||||
| Activated | 4.73 | 6.42 | |||||
Pillar breakdown (1-10) Value6.24 Stress2.99 Momentum5.14 Infrastructure8.65 Rates2.72 Market access7.75 Programs6.37 Institutions9.37 Notes Quadrant consistencyActivated in 56% of our sensitivity tests (Ready to scale in 43%). | |||||||
| Activated | 4.02 | 6.99 | |||||
Pillar breakdown (1-10) Value4.61 Stress3.40 Momentum3.37 Infrastructure9.73 Rates7.73 Market access8.80 Programs4.68 Institutions5.38 Related analysis | |||||||
| Activated | 3.84 | 5.04 | |||||
Pillar breakdown (1-10) Value3.76 Stress4.17 Momentum3.19 Infrastructure9.73 Rates2.19 Market access8.80 Programs4.22 Institutions4.10 Notes Quadrant lineERCOT is within 0.04 of the quadrant line on the Readiness axis. Quadrant consistencyActivated in 51% of our sensitivity tests (Monitor in 49%). Related analysis | |||||||
How to read the quadrant
Each cell answers a different reader's question.
The central result is the score produced from fixed scoring ranges and weighted geometric means of factors and category scores before any sensitivity run. The published quadrant comes from the central result. Quadrant consistency is how often a grid stays in the same quadrant across 20,000 Monte Carlo re-scorings, which vary the category weights, the scoring ranges, how factors are combined, and each input. How the index measures value and access
Opportunity: How much a grid needs flexibility and what that flexibility is worth. Readiness: How far policy, markets, and regulation have developed to let that value be captured and flexibility scaled.
Untapped
high opportunity · low readinessThe reform target. Value and need are high and the grid is not ready to use flexibility. A high Opportunity score meets a low Readiness score. Regulators and advocates should look here first.
Ready to scale
high opportunity · high readinessOperator entry territory. The value is large and the grid is ready to use it.
Monitor
low opportunity · low readinessWatch the momentum. The value and the means to capture it are both still small.
Activated
low opportunity · high readinessThe grid is ready to use flexibility, and its value and need are lower. A high Readiness score meets a low Opportunity score.
Key takeaways
- PJM leads the seven grids on Opportunity (7.01). Its Value score (8.89) is also the highest. On a gross, perfect-foresight basis over the scored year, energy is the largest of its four value streams, and capacity is the second largest. Priced on the same basis over the three years ending September 2026, energy falls to second, behind capacity at its scored value. PJM's capacity auction has cleared at its price ceiling for three straight delivery years: $329.17 for 2026/27, $333.44 for 2027/28, and $325.00 for 2028/29 per megawatt-day. The published PJM score uses capacity prices for the 36 months after September 30, 2026: the rest of the 2026/27 delivery year, all of 2027/28 and 2028/29, and the first four months of 2029/30. Those four months carry the 2028/29 price forward until the 2029/30 auction clears. Its Readiness (4.81) sits below the quadrant line, which places PJM in Untapped.#
- California has built the most of what flexibility needs. CAISO leads on Readiness (6.99). Its Infrastructure (9.73) and Rates (7.73) scores are the highest of the seven. California's three investor-owned utilities made time-of-use pricing the default residential rate. EIA reports that Southern California Edison's residential enrollment in time-varying pricing programs, time-of-use included, rose from about 20,000 in 2013 to 2.68 million in 2025.#
- In ERCOT's competitive market, time-of-use pricing is opt-in. ERCOT ties CAISO for the highest Market access score (8.80), yet its Rates score (2.19) is second lowest. Households in those competitive areas choose their own retail plan and pay a time-varying price only if they pick one. State rules have also barred indexed pricing plans for residential and small-business customers since February 2022.#
- New York rewards its utilities for growing demand response. NYISO's Institutions score (9.37) is the highest in the release, well above the next grid (ISO-NE, 7.65). New York regulators decouple utility revenue from electricity sales. They also pay earnings incentives tied to demand-response growth.#
- MISO is the Untapped grid closest to Ready to scale. It lands there in 46% of our sensitivity tests, more often than PJM (34%) or SPP (14%). It also has the highest Momentum score (7.73). The share of electricity used where congestion moves the price by more than $1 per MWh, in either direction, grew on six of the seven grids over the last three years (for ERCOT, each load zone's gap from the system-wide bus-average price). On MISO and SPP, the top 100 hours in each price zone also took a growing share of that zone’s positive real-time prices (weighted across zones by electricity use).#
How the index works
Grid flexibility means shifting the timing of power use, storage, and generation to match what the grid needs, away from the hours when it is strained and toward the hours when power is in surplus. Much of it comes from homes, businesses, and large facilities that shift or cut their electricity use. It can do some of the work of new power plants and power lines. It pays off only where the grid needs it and the rules let customers take part. The Grid Flexibility Index measures both for the seven US wholesale power markets. It is published as dated, versioned releases. This is the 2026Q3 release, and later updates are published as new releases.
Opportunity measures how much flexibility is worth on a grid: what its markets pay for capacity, energy, and grid services, how stressed the grid is, and which way those pressures are moving. Readiness measures whether that value can reach customers: smart meters, time-varying rates, market access for companies that pool customers, demand-response programs, and utility incentives that reward it. Both are scored from 1 to 10 and read as a pair. A grid can need flexibility badly and still be unable to use it.
Time-varying rates and Utility incentives use utility-level records. Rate access is averaged across utilities using residential-meter weights. Enrollment uses utility filings with estimates for missing records, and the Index scores Time-varying rates under each set of estimates and averages those scores. ERCOT uses a grid-wide enrollment estimate because its delivery utilities do not report the retail suppliers' enrollment. Utility-incentive steps are also averaged using residential-meter weights, including ownership-class estimates for utilities without an assessed incentive score. Smart meters, Demand-response enrollment, and Peak savings ratio enter as grid-level measures. Aggregator access and Retail market structure combine with equal weight into Market access. Demand-response enrollment and Peak savings ratio combine with equal weight into Programs. Smart meters form Infrastructure, Time-varying rates form Rates, and Utility incentives form Institutions. The five categories combine with equal weight in a geometric mean to form Readiness. A category with no scored factor withholds Readiness. Wholesale participation is shown for reference only.
Both scores are built from published sources and documented estimates: ISO market prices, EIA utility filings, NERC reliability assessments, regulator orders and tariffs, and FERC dockets. Where a market publishes no price, we use a benchmark or our own estimate. The factor list below says what each factor measures. Each factor is measured against the same fixed range for every grid, so scores are comparable across grids within a release. Energy coverage varies by grid. MISO, ERCOT, SPP, and CAISO each leave some load areas unpriced, and the Energy row of the value table names them. In MISO, the priced locations carry most but not all of the weight. The Index is released quarterly, each release effective at the end of its quarter. Publication timing can vary. Factors are being refined and may change from one release to the next.
To test how much the results depend on our own judgment calls, we run a Monte Carlo simulation. It re-scores every grid 20,000 times. Each run varies the category weights, the scoring ranges, and how factors are combined. It also shifts each input within a range set by its data quality or documented uncertainty. The shaded area around each grid spans the middle 80% of its simulated Opportunity and Readiness scores. Quadrant consistency shows how often a grid stays in the same quadrant.
Version 1 scores 19 factors. Wholesale demand-response participation is listed for reference while a replacement measure of market access is under review. The time-varying-rates factor counts residential customers, and the utility-incentives factor weights each utility by its residential meters.
The factors
Opportunity
12 factors · Value 3 · Stress 6 · Momentum 3
Value · 3 factors
- Capacity value. What flexible demand could earn by standing in for the capacity a grid pays generators to guarantee is available at peak.
- Energy arbitrage. What flexible demand could earn shifting electricity use from expensive hours to cheap ones, based on actual wholesale prices.
- Grid services. What flexible demand could earn providing the fast-response support services, like frequency regulation, that keep the grid stable minute to minute.
- Transmission avoidance. What flexible demand could earn by cutting usage during the specific peak hours transmission charges are based on.
Stress · 6 factors
Momentum · 3 factors
Readiness
8 factors, 7 scored in V1 · Infrastructure 1 · Rates 1 · Market access 3 · Programs 2 · Institutions 1
Infrastructure · 1 factor
Rates · 1 factor
Market access · 3 factors
Programs · 2 factors
Institutions · 1 factor
Time-varying-rates data comes from utilities serving 78.8% to 100.0% of the residential meters in each grid's utility roster. Utility-incentive scores are taken from each utility's rate orders where we have read them and from the average for similar utilities (investor-owned or public power) where we have not, with higher confidence for 6 grids and lower for ISO-NE.
Get the GFI briefing
The scores, methods, and ranges on this page stay open. Subscribe for each release's executive briefing: what changed, why it matters, and one market or utility example. Coming next: utility-level scores within each grid and customer-class lenses for residential, commercial, and large-load customers.
Questions or feedback on the Index?
Grid Flexibility Index V1. Coming next: utility-level scores within each grid and customer-class lenses.
Grid Flexibility Index V1 · 2026Q3 release · trailing 4 quarters ending 2026Q3 (price factors) · scores generated 2026-11-14
Corey Balgeman, "Grid Flexibility Index, 2026Q3 release: CAISO," Grid Flexibility, 2026. https://gridflexibility.fyi/gfi/caiso
License: CC BY 4.0