Market operator
Investing is assessed against market expectations: system operation and the future prices each developer expects, with congestion and zonal decoupling.
A new planning paradigm
ABEX
Generation expansion simulated agent by agent
Are we planning the system that will actually get built?
Why it fails
A model that optimizes the system cannot anticipate someone chasing returns
Planning has rested on centralized models: a plant enters when its investment is justified by the reduction in system operating costs. The problem is not the data, nor how those models are solved, but the assumption of who decides and why.
What the centralized model assumes
- Investment is justified by the reduction in system operating costs.
- Only what belongs to the least-total-cost solution gets built.
- Entry happens at the optimal moment defined by the planner.
VS
What actually happens in the market
- Nobody invests to minimize system cost: they invest for expected profitability.
- Without a PPA there is no financial close, however «optimal» the project is for the system.
- Timing is set by auctions, financing and portfolio strategy.
On top of that come assumptions of perfect rationality, full coordination and no market failures; and a technological and contractual homogeneity that ignores the diversity of preferences, contracts and practical constraints in the market.
How ABEX decides
It does not minimize costs: it simulates each agent's decision
The works plan is built by simulating each player's decision, with its own policies, instead of solving a single least-cost problem.
Generators and retailers
Each project is assessed by its effect on the agent's portfolio — profit and risk — not by its contribution to system cost. Asymmetric profitability across companies determines who builds.
Contracts as the driver
The PPA is the bankability condition: projects materialize when they secure contracts, not when an optimizer declares them efficient.
Real preferences
Technology, banking and auction preferences, learned from a decade observing the market.
Comparison
What should be built, or what will be built
Neither replaces the other. The centralized model remains the efficiency benchmark; ABEX tells you the system you will have to operate, contract and compete in.
| Centralized model | Agent-based model | |
|---|---|---|
| Who decides? | A central planner | Market operator, generators and demand |
| Investment driver | Least system cost | Price expectations and contracts |
| Decision inputs | Costs and availability | Contracts, portfolios, expectations, competition |
| Entry condition | Optimal moment for the system | Contract award and financial feasibility |
| Results | Works plan and costs | Works, contracts, prices, tariffs and commercial balance |
→ Swipe the table to see the agent-based column.
The centralized model answers what should be built. The agent-based model answers what will be built.
Methodology
Three integrated modules
They reproduce the interaction between system operation, investment decisions and contract formation.
1
Electricity market simulation
An investment model produces the price and generation projections. This is where the market expectations that guide the agents are set.
2
Bid creation through the optimal portfolio
Each agent builds its energy and price bids with a portfolio optimization model. This is where individual strategic decisions on investment and contracting come in.
3
Supply contract award
Bids are matched, contracts awarded, and the new works enter the expansion plan. That is how the endogenous evolution of the market emerges.
Electricity market simulation
Price and generation projections
Portfolio model
Bid curves
Matching model
New works
Rolling horizon
Every auction is assessed with the information available at its time
For each auction, the Optimal Portfolio Model computes every company's bid: the minimum price (break-even) that covers the supply and keeps the project viable. If the company can cover it with energy already available, 5 USD/MWh are discounted.
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
20262045
Click the track to pin an assessment; click again to resume the sweep.
Once awarded, the company's available energy is updated and, if new projects entered, so are the market projections. Only then does the next auction begin. The scheme shows the mechanism, not an auction calendar.
What it delivers
Not just a works plan
Modelling contracts, portfolios and expectations also yields the commercial and price trajectory: one model for commercial, investment and regulatory decisions.
Commercial position
- At what price will the next auctions be awarded?
- How does my contract balance and spot exposure evolve?
- When is it worth contracting and when to stay open?
Competition and investment
- Which technology and which location are most competitive at each moment?
- How do the others' strategies affect the contract price?
- Where in the investment cycle are we entering?
Regulation and public policy
- How does investment respond to a regulatory change?
- What trajectory do tariffs follow?
- What signal does the capacity market give?
What will get built in your market?
Tell us what decision you have ahead — an auction, an asset purchase, a regulatory change — and we will show you what ABEX projects.