Electricity Markets: Non-Storability & Price Dynamics

Published: June 23, 2026

Article by Ryan O'Connell, CFA, FRM

Electricity is unlike any other commodity. While oil can be stored in tanks, natural gas in underground caverns, and metals in warehouses, electricity must be consumed the instant it is generated. This fundamental characteristic — non-storability — shapes everything about how electricity markets function, from real-time price formation to the dramatic price spikes that can send wholesale power costs from $30 to over $1,000 per megawatt-hour within minutes.

Understanding electricity markets is essential for energy traders, utility analysts, infrastructure investors, and anyone preparing for the CFA exam’s alternative investments curriculum. This guide covers the unique mechanics of power markets, how prices are set through locational marginal pricing (LMP), the structure of ISOs and RTOs, and why electricity behaves so differently from other commodities.

The Unique Nature of Electricity (Non-Storability)

When demand for oil exceeds supply, the shortfall can be met by drawing down inventories. When electricity demand exceeds generation capacity, the grid experiences frequency deviations, voltage drops, and potentially cascading blackouts. This real-time balancing requirement fundamentally changes how electricity markets operate.

The implications of non-storability include:

• No arbitrage via storage — The classic commodity arbitrage (buy spot, store, sell forward) is impossible. Standard cost-of-carry pricing models break down completely.
• Extreme price volatility — Without inventory buffers, prices respond immediately to supply-demand imbalances with spikes that can exceed 100x normal levels.
• Real-time markets — Electricity must be traded in time intervals as short as 5 minutes, with continuous grid balancing by system operators.
• Location matters — Electricity cannot be easily transported around transmission constraints, creating price differences across nodes.

While battery storage and pumped hydro are growing, they remain a small fraction of total grid capacity. As of 2025, utility-scale battery storage in the US represents roughly 5% of peak demand by capacity — meaningful for short-duration balancing but insufficient to create the inventory arbitrage channel that exists for storable commodities like natural gas or crude oil.

Real-Time Balancing and Grid Operations

Electricity grids operate at precise frequencies — 60 Hz in North America, 50 Hz in Europe. When generation exceeds load, frequency rises; when load exceeds generation, frequency falls. Deviations of just 0.5 Hz can damage equipment and trigger automatic shutdowns.

Grid operators must continuously match supply and demand through several mechanisms:

Electricity flows follow Kirchhoff’s laws — physical equations governing current and voltage in networks. Power flows along the path of least impedance, not necessarily the contracted path. This means a transaction from Generator A to Load B affects flows across the entire interconnected grid, potentially causing congestion on transmission lines hundreds of miles away.

Wholesale vs Retail Electricity Markets

Electricity markets operate on two distinct levels with very different price dynamics:

Retail prices lag wholesale prices because utilities average their procurement costs over months or years, and regulators smooth rate adjustments to protect consumers. This insulation means most consumers never see real-time price signals — a factor that contributes to demand inelasticity and price spikes during shortage conditions.

Day-Ahead and Real-Time Markets

Most organized electricity markets operate two parallel market structures that clear at different times:

Day-Ahead Market (DAM) — Clears 24-48 hours before delivery. Participants submit supply offers and demand bids for each hour of the next day. The market operator runs a security-constrained unit commitment optimization to determine which generators will run and at what levels. Day-ahead prices provide price discovery and allow participants to lock in positions before real-time.

Real-Time Market (RTM) — Clears every 5-15 minutes during actual delivery. Settles differences between day-ahead schedules and actual conditions. Real-time prices can diverge significantly from day-ahead when load forecasts miss, generators trip offline, or transmission constraints bind unexpectedly.

Ancillary Services Markets — Alongside energy, ISOs procure services needed to maintain reliability: regulation (following rapid load changes), spinning reserves (online backup), non-spinning reserves (quick-start backup), and in some markets, ramping products and black start capability. These services are essential for grid stability and represent additional revenue streams for flexible generators.

Locational Marginal Pricing (LMP)

Locational Marginal Pricing is the standard pricing mechanism in organized US electricity markets. LMP represents the cost of serving the next megawatt-hour of demand at a specific location on the grid.

Where:

• Energy Component — The marginal cost of generation system-wide (same across all nodes)
• Congestion Component — The cost of transmission constraints that prevent cheap power from reaching a location (can be positive or negative)
• Losses Component — The marginal cost of electrical losses in transmission (increases with distance from generation)

LMP creates price signals that incentivize efficient behavior: generators are paid more to locate near load centers, demand response is more valuable in constrained areas, and transmission investments can be evaluated based on congestion relief value.

Price Spikes and Mean-Reversion

Electricity prices exhibit a unique pattern: long periods of stable, low prices punctuated by extreme spikes that collapse rapidly. This behavior stems directly from non-storability and demand inelasticity.

Why spikes occur:

• Generator outages remove supply suddenly
• Extreme weather drives heating/cooling demand above forecast
• Transmission constraints prevent power flows from uncongested regions
• No inventory buffer exists to absorb the mismatch

Why spikes mean-revert quickly:

• Emergency reserves are dispatched
• Peaker plants come online within 10-30 minutes
• Interruptible loads are curtailed
• Adjacent regions export power if transmission allows
• Weather extremes are temporary

This spike-and-revert pattern creates challenges for standard commodity models. Monte Carlo simulation with mean-reverting jump-diffusion processes is often used to capture these dynamics for risk management and derivatives pricing.

Negative Electricity Prices

In markets with high renewable penetration, prices can go negative — meaning generators pay to produce. This occurs when:

• Wind or solar generation exceeds demand during low-load periods
• Nuclear and baseload plants cannot economically shut down for short periods
• Renewable generators have production tax credits that exceed negative prices

CAISO and ERCOT both experience frequent negative pricing events, particularly during spring months with high wind/solar output and mild temperatures.

Load Duration Curves

A load duration curve ranks all hours of the year by demand level, from highest to lowest. This visualization reveals the structure of electricity consumption and explains why different types of power plants exist.

Peaker plants may only run a few hundred hours per year, but they are essential for meeting demand during the highest-load periods. Their economics depend on capturing high prices during scarcity — they lose money most of the time but profit enormously during price spikes.

ISO/RTO Structure (PJM, ERCOT, CAISO)

In the United States, organized wholesale electricity markets are operated by Independent System Operators (ISOs) or Regional Transmission Organizations (RTOs). These non-profit entities manage the grid, operate markets, and ensure reliability.

Not all US electricity is traded through ISOs. The Southeast (served by utilities like Southern Company and Duke Energy) and much of the West operate under bilateral contract structures with limited organized markets. These regions represent roughly one-third of US electricity consumption.

Electricity Futures and Forwards

Despite non-storability, electricity derivatives markets are active and essential for hedging price risk. However, the pricing dynamics differ fundamentally from storable commodities.

For storable commodities, the cost-of-carry relationship links spot and forward prices:

This relationship collapses for electricity because you cannot buy power today, store it, and deliver it next month. Instead, electricity forward prices reflect:

• Expected spot prices — Market participants’ forecasts of future supply/demand conditions
• Risk premiums — Compensation for bearing spot price uncertainty (typically buyers pay a premium)
• Load shape — Peak vs off-peak pricing based on expected demand patterns
• Capacity and fuel costs — Expected marginal generation costs

Major electricity futures products include peak (7am-11pm weekdays) and off-peak blocks, traded on CME (NYMEX) and ICE. For detailed trading structures and contract specifications, see Peak vs Off-Peak Power Trading.

PJM vs ERCOT vs CAISO: Market Design Comparison

The three major US electricity markets take fundamentally different approaches to market design, reliability, and pricing.

ERCOT’s energy-only design relies on scarcity prices to signal the need for investment. During the February 2021 winter storm, prices hit the then-$9,000/MWh cap for days, highlighting both the risk and the design intent. The cap was subsequently reduced to $5,000/MWh in January 2022. PJM’s capacity market, by contrast, pays generators to be available regardless of energy prices, providing more stable revenues but potentially dulling investment signals.

Common Mistakes in Understanding Power Markets

Several widely-held beliefs about electricity markets are either oversimplified or incorrect:

“Electricity prices simply follow natural gas prices” — While natural gas is often the marginal fuel, the relationship is non-linear. When gas prices rise, more coal or renewables may set the marginal price. Transmission constraints, demand levels, and renewable availability all affect which fuel is marginal at any given time.

“Battery storage will eliminate price spikes” — Current battery storage can provide 2-4 hours of discharge at rated capacity. This helps with short-duration price spikes but cannot address multi-day weather events or sustained supply shortages. The scale of storage needed to fully buffer electricity markets remains far beyond current deployment.

“Deregulation always lowers electricity prices” — California’s 2000-2001 crisis, where wholesale prices spiked from $30 to over $1,000/MWh while retail prices were frozen, demonstrates that poorly designed markets can produce worse outcomes than regulation. Market design details matter enormously.

“Wholesale electricity is one national market” — The US has multiple regional grids with limited interconnection, plus bilateral regions outside organized markets. Prices in Texas (ERCOT) can be $200/MWh while neighboring MISO prices are $30/MWh, with no ability to arbitrage the difference.

“Low average prices mean low risk” — Electricity prices are low or moderate 99% of the time. But the 1% of hours with extreme prices can determine profitability for generators and create massive losses for exposed buyers. Tail risk, not average prices, drives electricity market economics.

Limitations of Electricity Price Forecasting

Key forecasting limitations:

• Weather dependence — Heating and cooling demand are primary price drivers. Weather forecasts beyond 7-10 days have limited accuracy, making price forecasting inherently uncertain.
• Generator outages — Forced outages are random and can remove thousands of megawatts instantly. Planned maintenance is known but creates vulnerability to other unexpected events.
• Transmission constraints — Congestion depends on generation patterns and load distribution. Line outages or de-ratings can create price separation with little warning.
• Regulatory changes — Price caps, capacity mechanisms, emissions rules, and reliability requirements can change market dynamics significantly.
• Renewable intermittency — Wind and solar output varies with weather, adding another source of forecast uncertainty.

Because of these challenges, electricity price forecasting typically combines fundamental models (supply stack, load forecasting), statistical time series methods, and Monte Carlo simulation for risk quantification.