COLUMN 07 — Revenue & Investment Decision
20-Year Cash Flow Structure for Battery Business
— Understanding IRR & DSCR
Understanding the 20-year cash flow structure is essential for battery business investment decisions. The LDA type and full merchant type fundamentally differ in both IRR (internal rate of return) and DSCR (debt service coverage ratio). This article explains the investment decision framework rather than specific project numbers.
Two Business Models, Two CF Structures
The cash flow (CF) character of grid-scale battery business fundamentally differs depending on the chosen business model. As a starting point for investment decisions, you need to correctly understand the CF characteristics of these two models.
| Item | LDA-Type (Fixed Revenue) | Full Merchant-Type (Market Trading) |
|---|
| Revenue Character | 20-year fixed (capacity payment) | Market price linked (variable) |
| CF Forecast Accuracy | High (contract amount is fixed) | Low (depends on market fluctuations) |
| IRR Estimate | 2-4% (bond-like returns) | 8-15% (large range) |
| DSCR Structuring | Easy (fixed CF) | Difficult (uncertain CF) |
| Suitable Financing | Project Finance (PF) | Corporate Finance (CF) |
| Risk Profile | Low-risk, Mid-return | High-risk, High-return |
LDA-Type 20-Year Cash Flow Structure
When awarded in the LDA (Long-term Decarbonized Power Source Auction), you receive fixed capacity payments over 20 years. In addition, a portion of market revenue (approximately 10%) remains as additional income.
LDA-Type CF Structure (Conceptual Model):
[Revenue Side]
- LDA capacity payment (20-year fixed): contract unit price x capacity (kW)
- Retained market revenue: approximately 5-15% of total market revenue
[Expenditure Side]
- Debt service (for PF)
- O&M costs: approximately 1-2% of CAPEX per year
- Insurance premiums
- Aggregator fees
- Wheeling charges, RE surcharge, etc.
- Battery replacement costs due to degradation (years 10-15)
Three-Tier Refund Rate Structure
LDA-awarded operators are obligated to refund a certain percentage of market revenue. This refund rate changes across three tiers.
Tier 1 (95% refund rate): Up to the business profit built into the bid price. Operator retains 5%.
Tier 2 (90% refund rate): Above the business profit up to the difference between LDA contract price and main auction clearing price. Operator retains 10%.
Tier 3 (85% refund rate): Excess profit above the above. Operator retains 15%.
Source: TMI Associates commentary / OCCTO system detailed materials (September 2024)
Importantly, only market trading profits are subject to refund; the LDA capacity payment itself remains 100% with the operator. KPMG analysis indicates that LDA-type IRR converges to approximately 3.2% "bond-like returns" from capacity payments alone.
Full Merchant-Type 20-Year Cash Flow Structure
Without bidding in the LDA, the full merchant type captures 100% of revenue from three markets: JEPX, balancing market, and capacity market. Without refund obligations, returns can significantly exceed the LDA type when markets are favorable.
Full Merchant CF Structure (Conceptual Model):
[Revenue Side - 3-Market Stacking]
- JEPX arbitrage: charge during low-price hours, discharge during high-price hours
- Balancing market: compensation for both delta-kW and kWh
- Capacity market (main auction): annual fixed income (award rate over 96%)
[Expenditure Side]
- Debt service (for BS finance)
- O&M costs: 1-2% of CAPEX/year
- Aggregator fees: 10-15% of market revenue
- Wheeling charges, RE surcharge, etc.
- Battery replacement costs due to degradation
Uncertainty of Full Merchant CF Forecasts
When forecasting 20-year CF for the full merchant type, the following variable factors must be considered:
-
JEPX spread: Risk of spread compression due to mass battery deployment
-
Balancing market: Price cap reduced from 19.51 to 15 yen from March 2026. Further reductions possible
-
Capacity market: Regional price differences widening. Hokkaido, Tohoku, Tokyo are high; Kansai, Chugoku, Shikoku are low
-
Battery cost: Cell prices are declining, but replacement timing cost assumptions significantly impact revenue
METI MRI warns that "revenue outlook uncertainty is high."
Source: METI MRI Materials (August 29, 2024) / ANRE Price Cap Reduction (January 23, 2026)
IRR — Understanding Internal Rate of Return
IRR (Internal Rate of Return) is the most common metric for measuring investment profitability. IRR is the discount rate at which the present value of 20-year cash flows equals the investment amount.
| Model | IRR Estimate | Character | Comparable To |
|---|
| LDA type | 2-4% | Bond-like returns | Competes with government/infrastructure bonds |
| Full Merchant (Conservative) | 8-10% | Equity-like returns | Competes with real estate/PE investments |
| Full Merchant (Optimistic) | 12-15%+ | Venture-like returns | High-risk investment category |
The variable with the greatest impact on IRR is "grid connection construction cost"
Even for the same 50MW project, the difference between 200 million yen and 2 billion yen in connection costs creates an 1.8 billion yen gap in total investment. This directly impacts the denominator, making the effect on IRR enormous.
For example, with identical annual CF:
- Total investment 10 billion yen → IRR 10%
- Total investment 11.8 billion yen → IRR 7.8%
The simple arithmetic that lower construction costs lead to higher IRR determines the success or failure of battery business.
DSCR — Understanding Debt Service Coverage Ratio
DSCR (Debt Service Coverage Ratio) is the metric most valued by financial institutions in project finance. The formula is "annual net CF / annual debt service." A DSCR below 1.0 means that year's debt service cannot be met.
DSCR levels generally required by financial institutions:
- Minimum DSCR: 1.1-1.2 (must not fall below this throughout the entire period)
- Average DSCR: 1.3-1.5
- Solar PF track record: Average DSCR of 1.3-1.4 is standard for FIT projects
LDA-type batteries can achieve DSCR structuring similar to solar FIT due to the 20-year fixed capacity payment. The full merchant type risks years where DSCR falls below 1.0 due to market fluctuations, making PF structuring difficult.
Why LDA-Type is Suited for PF
The CF structure of LDA-type batteries is structurally similar to solar FIT projects.
| Comparison Item | Solar FIT | Battery LDA |
|---|
| Basis of Fixed Income | FIT purchase price (20-year fixed) | LDA contract unit price (20-year fixed) |
| Variable Factors | Solar irradiance (annual variation +/-5-10%) | Capacity payment is fixed. Market revenue varies but only ~10% is retained |
| SPC Formation | GK + Tokumei Kumiai (TK-GK) is standard | Same scheme can be applied |
| Lender Understanding | Established (thousands of precedents) | Developing (few precedents but structural understanding progressing) |
Sensitivity Analysis Framework
When building a 20-year CF model, sensitivity analysis on key variables is essential. We recommend examining three scenarios - base case, optimistic, and conservative - for the following variables.
Key variables for sensitivity analysis:
(1) Grid connection construction cost (directly impacts CAPEX)
The confirmed amount from the connection review response is the base case. Factor in +/-10-20% variation risk after site survey.
(2) Battery cell price decline rate (impacts replacement costs)
BloombergNEF projects $69/kWh by 2030. Prices may be even lower at the time of replacement in years 10-15.
(3) JEPX spread (primary revenue for full merchant type)
Risk of spread compression from mass battery deployment. Conservative case assumes 30-50% reduction from current levels.
(4) Balancing market clearing rate and unit price
Factor in gradual price cap reduction and procurement volume reduction (3-sigma to 1-sigma).
(5) Capacity market clearing price
Regional price differences widening. Refer to historical trends in the installation area.
(6) Interest rates
For PF, consider fluctuations in base rate (TIBOR, etc.) plus spread.
Investment Decision Framework
Q1: Is project finance needed?
→ YES → Prioritize LDA type.20-year fixed CF makes DSCR structuring easy。
→ NO (self-funding/BS utilization) → Proceed to Q2
Q2: Do you have market trading expertise (or an aggregator)?
→ YES → Full merchant type becomes an option. Higher IRR achievable.
→ NO → LDA type is safer.Secure fixed income while avoiding operating risk.
Q3: What is your target IRR?
→ 3-4% is sufficient (stability-focused)→ LDA type
→ 8%+ desired (growth-focused) → Full merchant type
Common premise: In either model, projects with lower connection costs have higher IRR and more DSCR headroom. Site selection is the biggest lever.
Summary
The 20-year CF of battery business has fundamentally different characteristics between LDA and full merchant types. LDA type offers "predictable, stable CF" while full merchant type offers "high-return but variable CF."
To improve investment decision accuracy, individual CF models using project-specific connection costs and grid conditions are needed. Please use the framework explained in this column as a starting point for your decisions.
Sources & References:
- KPMG "Decarbonized Power Source Auction Analysis" https://kpmg.com/jp/ja/home/insights/2024/10/decarbonized-power-auction.html
- TMI Associates "LDA Commentary" https://www.tmi.gr.jp/eyes/blog/2023/15106.html
- OCCTO System Detailed Explanation Materials (September 2024)
- METI MRI Materials (August 29, 2024)
- Agency for Natural Resources and Energy, Price Cap Reduction Decision (January 23, 2026)
- BloombergNEF Battery Pack Price Survey (2024)
- METI Stationary Battery System Expansion Study Group, 5th Meeting Materials (January 30, 2025)