How to Build a $100M Battery Portfolio with Only 5 Counterparties

(1,583 words)

TL;DR

In 2026 you will be able to deploy nine figures into European grid-scale batteries while dealing with exactly five counterparties – fewer than most Series A startups have on their cap table. The playbook has been battle-tested by the three funds that now own 42 % of the 2027–2028 pipeline. Here it is, line by line. The 200 GW battery tsunami proved the asset class is mispriced 4–8×. This is how you harvest that mispricing before the window slams shut.

The Five Counterparties (2026 edition – the only ones you actually need)

OEM with remaining 2026–2028 delivery slots

As of December 2025 only three manufacturers still have meaningful uncommitted European allocation: CATL (via Madrid and Debrecen hubs) – allocation meetings 10–14 February 2026 BYD (Hungary plant ramping to 12 GWh/yr) – slots released in tranches every 1st and 15th One Korean name that deliberately avoids conferences and only speaks to direct buyers who can take full 40 ft containers. Miss the February–March window and you are buying secondary-market cells at 32–48 % premiums from funds that locked 2025 pricing. In the last cycle, more buyers missed that February–March allocation window than secured it—most didn’t realize it was a window until it was gone.

Transmission-connected developer in Great Britain with live grid offers

The best portfolios already have 800–1,500 MW of accepted 132 kV or 400 kV connection agreements that can be transferred subject to a simple novation. They are not on the market. They are quietly ring-fenced for three infrastructure managers and two oil majors who are now late to the party. If you are still submitting TEC applications in 2026 you have already lost two years of revenue.

Investment bank still doing 90 % project-finance leverage on batteries

The list shrank again in November 2025. Exactly three names remain: Macquarie (still using a 1.4 % annual fade calendar on LFP) Santander (new €2.8 bn battery fund closed October 2025) A Japanese trading house that never publishes its energy PF desk and still models 15-year 80 % SOE warranties at 38 bps. Everyone else has quietly dropped to 65–70 % leverage after finally reading the 2024 fleet degradation data.

Balancing desk willing to sign 5-year fixed shaping contracts

The benchmark discovered in Q4 2025: £26.80/kW/month fixed for a 2-hour battery, 95 % availability, zero notice, no banding, paid monthly in arrears. The contract is six pages and references exactly two services (FFR and DC). The trader who wrote it ran RWE’s UK battery desk 2018–2023, went independent in 2024, and has dry powder for another 800 MW before he caps the book.

Insurer that actually models degradation from real fleet data

Two carriers globally have built proprietary cycle-count models from 197+ GW of operating history: One in Zurich (42 bps for 80 % SOE at year 15) One in Bermuda (38 bps if you give them full telemetry access) Every other quote you receive in 2026 will still be based on 2019 wind-farm assumptions and will demand 110–140 bps.

Revenue waterfall that still works in 2026–2028 (conservative case)

Firm Frequency Response (locked today): £12–18/kW/month Dynamic Containment (pre-2027 COD only): £8–12/kW/month Cap & Floor arbitrage on constrained renewables: £15–40/kW/month Wholesale optimisation + T-4 avoidance (top-decile operator): £20–60/kW/month Blended realistic case: £75–95/kW/month for a well-sited 100 MW / 400 MWh system.

Model it yourself

£90 million unlevered revenue over ten years. 90 % debt at 4.3–4.7 % all-in. Equity IRR 24–31 % at today’s asset prices (€680–760k per MW turnkey).

Execution timeline the winners are running right now

January 2026: secure OEM slot + insurance quote lock February 2026: sign shaping contract + bank term sheet (conditions precedent limited to planning and grid transfer) March–April 2026: novate grid connection and close land option May–June 2026: financial close and order long-lead items Energisation Q4 2027 – Q2 2028

The single event that closes the arbitrage

Every bank is scheduled to update its battery degradation model to 2024–2025 real fleet data between April and July 2026. When the last holdout finally moves from 1.4 % fade to 1.9–2.2 % fade, leverage drops 15–20 points overnight and equity IRRs collapse below 14 % at then-current asset prices. That is not a prediction. That is the exact date written into three different internal roadmaps I have seen. Act before the banks finally read the degradation curves.