The Last Tab

How one insurance limit decision travels through your entire capital stack.

In project finance, revenue projections get scrutinized cell by cell. Debt sculpting gets debated. Tax equity sizing weighed. Then someone says, “Insurance looks fine, next.”

But insurance, I argue, is where your capital structure is most exposed and least examined.

So I built a model to make the exposure visible.

The Setup: 150 MW Solar + 75 MW / 300 MWh BESS

The model includes a full cash flow waterfall - construction through operations, debt engine, tax equity partnership flip, and covenant testing.

Then I changed one variable: natural catastrophe limit set at $25M instead of $50M.

What happens in a 250-year event? What happens in a 500-year event?

The Cascade

A lower limit doesn’t just increase retained loss. It creates a sequence - and each step makes the next one worse:

1. Capital gap - the uninsured portion of loss exceeds reserves.
2. Liquidity stress - Debt Service Reserve Account depletion.
3. Covenant breach - DSCR compression below lender thresholds.
4. Distribution lock-up - months or years of trapped cash for the sponsor.
5. Equity cure - sponsor capital call. This is where IRR goes to die.

Same project. Same PPA. Different insurance limit. One survives. The other needs a rescue.

Scenario 1 | $25M Limit

250-Year Event. The Near Miss

The debt service reserve fund absorbs it. Barely. DSCR compresses - uncomfortable, but doesn’t breach the covenant threshold. Distributions blocked 1–2 years while reserves replenish.

On paper, the project survives. In practice, the safety net is gone. Any subsequent event - even a modest one - hits an unprotected structure. And in a world where natural catastrophe frequency is trending upward, “subsequent event” isn’t hypothetical. It’s a planning assumption.

500-Year Event. The Equity Destruction

Now the cascade is real: DSRA fully depleted, DSCR breaches covenant threshold, distributions locked 3+ years, Sponsor faces a $17.2M equity cure - 86% of their original $20M commitment.

That’s not a loss. That’s a near-total wipeout of the equity position.

Scenario 2 | $50M Limit

Same loss scenarios. Different structure.

250-year event. Structure holds.

Capital gap drops to $1.2M. Reserves intact. No distribution lock-up. The structure barely notices.

500-year event. Structure holds.

Capital gap drops to $4.5M. No equity cure. No default risk. DSCR stays above covenant. The structure holds.

The Strategy: Why It’s Not Just "Buy More"

The assumption: incremental annual premium of $150,000 - $200,000 for the additional $25M of natural catastrophe limit. Over 20 years, that’s roughly $3M - $5M.

That’s the trade-off to avoid a potential $17M+ capital call.

On its face, that’s not a cost decision. That’s a capital preservation decision.

But here’s the part most analyses skip: if this trade-off were always this obvious, everyone would buy the higher limit. They don’t. Why?

• Capacity. Some projects can’t obtain $50M of Natural Catastrophe coverage for a specific site or technology. The market won’t offer it, or the tower requires more carriers to participate where only a few are willing to write.
• Marginal pricing. The incremental cost isn’t linear. When capacity is rare, the higher the tower, the higher the rate-on-line because pricing isn't linear; the "top" of the tower can be the more expensive relative to the risk.
• Competing priorities. Every premium dollar competes with other uses such as higher business income limits, broader exclusion buybacks, lower deductibles. Limit isn’t the only lever.

So Why Still Show the Model?

The model isolates one variable to show the structural consequence. Real placement decisions involve all of these constraints at once. The point is that you should know what you’re choosing when you don’t.

What helps you choose consistently and over time? Risk tolerance. That is the distinction between picking 250-year event vs. 500-year event as your starting point, regardless of how the market behaves.

Same Deal, Different Concerns

The same insurance limit decision means different things depending on where you sit in the capital stack. And those different meanings can conflict.

Executives and Investment Committee: “What do we actually lose and when?” They want a number. Not “it depends.” Not a range. A specific downside scenario they can weigh against return. The model gives them that, but only if someone translates the insurance tab into the language of equity loss and IRR impact.

Lenders: “Is our collateral protected? Does DSCR hold?” Lenders care whether insurance proceeds are sufficient to rebuild the asset (their collateral) and maintain debt service. In a shortfall, the lender’s question becomes: who controls the insurance proceeds to rebuild or pay down debt? The answer is usually in the loan docs, and it usually favors the lender.

Tax equity: “Is my downside exposure what I was told it was?” Tax equity investors care about one thing above all else: are the tax benefits protected? A catastrophic loss that delays commercial operation, interrupts production, or triggers ITC recapture is an existential event for their position. The entire return is front-loaded into the first five years. They don’t necessarily want to rebuild it and operate. They want to harvest the tax benefits and exit. They want to know whether tax insurance covers their downside and their definition of “downside” is different from everyone else’s in the room.

Sponsors: “How much of my equity is genuinely at risk?” At $25M limit, the answer in a 500-year event is: almost all of it. That is material because the sponsor is interested in the project generating returns over 20 years. At $50M, the answer is: the structure absorbs it. Sponsors also carry a risk nobody else does: the reputational cost of a capital call. An equity cure doesn’t just destroy IRR. It destroys confidence in the next fundraise.

Risk advisors: “Can I translate PML into capital consequences my client understands?” This is the translation problem. Probable Maximum Loss is a modeling output. Capital consequences are what the deal team needs. The gap between those two conversations is where underinsurance lives.

Here’s The Tension:

The lender wants low deductibles and high limits (it’s not their insurance premium).

The sponsor wants to optimize premium spend (it’s their cash flow).

Tax equity investor wants certainty (it’s their tax basis).

These incentives structurally conflict and the insurance program is where the conflict gets resolved, whether anyone frames it that way or not.

Why Earthquake?

I chose earthquake deliberately.

For solar, earthquake isn’t a market-stressed peril. No significant loss history. No reactive pricing. No anxious underwriting cycles. That quiet is the point. The model isolates the structural relationship between one insurance limit decision and everything downstream. No noise. Just the math.

The model uses earthquake to prove the principle. Climate makes the principle urgent.

One important caveat: return period estimates assume stationary risk - the idea that a 250-year event has the same probability this year as it did a couple of years ago. For earthquake, that’s defensible. But when you substitute wildfire, wind, hail, or flood - the perils that are actually stressing portfolios right now - return periods aren’t stationary. They’re compressing. Which makes this analysis, if anything, conservative.

Climate change makes climate more unpredictable.

Substitute the peril that’s most relevant to your portfolio. The framework holds.

The Final Caveat: The Claims Cascade | What This Model Doesn’t Show (And Why That Matters)

The model traces the financial cascade with clean assumptions: loss occurs, claim is filed, insurance pays as modeled. In practice, there’s a second cascade that runs in parallel - the claims cascade.

Adjuster appointment takes weeks or months after a regional catastrophe event (everyone is filing at once). Proof of loss requirements create documentation burdens that delay payment. Business interruption measurement becomes a negotiation: carrier and policyholder rarely agree on lost revenue projections without dispute. Lender and sponsor may disagree on whether proceeds should rebuild the project or pay down debt.

A 500-year event with an 18-month claims resolution timeline and a lender who diverts proceeds to debt paydown doesn’t just deplete DSRA. It creates a liquidity crisis that the financial model may have never contemplated.

The model assumes the claim pays as modeled. The claims cascade is a workflow in and of itself. Factor that into your risk tolerance.

Download the Model

The full model is open-access. No email gate. No paywall.

Link: Clean Power Whisperer | Insurance Capital Model

Start with the Insurance & PML tab. That’s where limits become capital consequences. If your current financial model doesn’t connect insurance limits to DSCR, liquidity, and equity cure - it’s incomplete.

Because the tab nobody opens is the one that decides whether your equity survives.

-------- Clean Power Whisperer™ Perspectives Educational insights on risk, insurance, and capital for AI • Energy • Climate

← All Perspectives · Next: Stress Test Result from the World's Largest Sovereign Fund →