The Everglades and the Exoskeleton: Rethinking De-Risking in Project Finance

West of Moscow, I once stood in a sand trap that wasn’t a sand trap. It was 1990, in Nakhabino, where the Soviet Union was attempting something almost as improbable as perestroika itself: building a championship golf course. I had gone on a journey not permitted to journalists at the time, for a story that, when the tanks rolled into Kuwait that week, became peripheral and never saw daylight.

But I remember the bunker. Not as in golf, as in a battlefield.

Nearly half a century earlier, this had been part of the ground across which the German drive on Moscow finally ran out of road. Robert Trent Jones Jr. was designing Russia’s first championship course through the same landscape—and the wartime position was being left in the course.

There are worse monuments to persistence.

In 1941, an army that had advanced astonishing distances encountered something its victories had steadily consumed: room for error. Supply lines stretched. Men and machines were exhausted. Winter closed in. Soviet reserves counterattacked. The bunker survived. The army didn’t.

And that brings us, improbably enough, to project finance. We spend enormous effort making projects stronger: fixed-price EPC contracts, warranties, covenants, guarantees, locked feedstocks and offtakes. Each is a defense against something that can go wrong.

But what if enough defenses against Shock A leave the project unable to move when Shock B arrives? What if a project can advance toward its milestones while quietly consuming the very maneuvering room it will later need? And what if some of the things spreadsheets call inefficiency—cash reserves, schedule contingency, spare capacity, alternative pathways—are actually financial wetlands, places where uncertainty can go without taking the project with it?

In this column, let’s explore that idea in depth. The Everglades and the Exoskeleton, offers a different way of thinking about fragility, resilience, and the strange virtue of leaving a little room for the future, and I hope you enjoy it.

The Everglades and the Exoskeleton: Rethinking De-Risking in Project Finance

Engineers did not set out to destroy resilience in South Florida when they worked on the Everglades. They set out to solve specific, urgent problems—flooding, drainage, agricultural development, and water supply. In many respects, the massive mid-century civil works succeeded. But success at directing water systematically altered the larger landscape’s capacity to store, spread, and slowly release it.

The Everglades was valuable not merely because water could flow through it, but because millions of acre-feet of water could temporarily occupy it. When unscripted deluge met rigid channelization, the water didn’t disappear. It simply ran out of allowable space, cascading into surrounding systems.

We have spent a generation doing something remarkably similar in project finance, particularly for First-of-a-Kind (FOAK) infrastructure, industrial decarbonization, and complex energy assets. We approach risk by attempting to contract, covenant, warranty, and specify uncertainty out of existence.

In doing so, we often mistake structural rigidity for system resilience. Each de-risking measure is locally rational and genuinely effective against its intended threat. Yet collectively, they can strip the asset of its ability to deform, dissipate, and absorb the unexpected.

To build assets capable of surviving a changing world, we must move beyond static de-risking and ask the governing question of system architecture: When something happens that we didn’t protect against, where is the load allowed to go?

The Manifold Has Current: Resistance vs. Dissipation

In project development, we frequently treat environmental volatility—market shifts, supply chain disruptions, regulatory pivots—as the primary source of failure. But volatility itself is merely background forcing.

Think of external environmental disturbance as SWELL and structural vulnerability as FRAGILITY. SWELL is not the load itself; load is generated when SWELL meets the project’s architecture:

SWELL × FRAGILITY → LOAD

Once load is generated, the architecture determines where it can travel and where adaptation can occur. This is where modern project finance often miscalculates. Risk cannot be contracted out of a system; contracts simply reshape the resistance field through which adaptation flows:

LOAD + Resistance Field → Adaptive Flow

The viability manifold is not neutral geometry. It has current. Some directions are cheaper or harder for the project to move through than others. Fixed-price EPC contracts, restrictive covenants, debt-service requirements, and rigid warranties increase adaptive resistance in specific dimensions. Cash, schedule buffers, operational modularity, and contractual discretion decrease it.

Crucially, constraining one trajectory does not eliminate adaptation; it displaces adaptation onto another trajectory. Consequently, change goes where change is allowed. A feedstock price spike can remain a localized procurement issue—or, if contractual channels close off adaptation, it can be forced down remaining paths, becoming a liquidity problem, a covenant default, a sponsor dispute, and ultimately an existential crisis.

When an architecture specifies precisely where everything must go under ideal conditions, a fully optimized project may eliminate the very places where uncertainty was supposed to go.

Flying Buttresses, Exoskeletons, and Shock Substitution

Conventional project finance relies heavily on building structural defenses around the asset: long-term offtakes, fixed-price EPC contracts, strict debt ratios, performance guarantees, and rigid operating parameters.

The goal is not to remove these protections. A well-designed contract acts like a flying buttress—reinforcing a predictable, high-load vector while leaving the broader structure room to flex. The danger arises when every buttress is fused together until the architecture becomes an exoskeleton.

The more completely an exoskeleton constrains permissible deformation, the more load must be transmitted elsewhere when an unanticipated shock arrives.

This dynamic was demonstrated on the field of Agincourt in 1415. French plate armor represented the pinnacle of contemporary defense, meticulously engineered to resist weapon impacts (Shock A). On the constricted, rain-softened field, however, protection came with a severe mobility cost. As heavily equipped men-at-arms struggled forward through churned ground and increasingly compressed formations, the French system lost its capacity to maneuver.

The armor didn’t have to fail for the architecture to fail. Protection against Shock A helped create vulnerability to Shock B: exhaustion, crowding, and loss of adaptive freedom.

We see this shock substitution routinely in project finance:

• A fixed-price EPC contract suppresses price volatility (Shock A), but if inflation spikes, it can drive contractor distress or change-order rigidity (Shock B).
• Higher leverage can protect modeled equity returns against a high cost of capital, while the resulting debt-service burden may eliminate the financial wetland needed to absorb a delayed ramp-up.

Every defense against Shock A must be accompanied by the question: What Shock B does this defense make harder to survive?

The Cannae Trap: Shrinking Reachable Sets

Fragility does not always manifest as immediate failure; it frequently arrives disguised as rapid forward progress.

At the Battle of Cannae, Hannibal’s bowed center yielded under the Roman advance, as his deployment was designed to accommodate. The Roman legions pressed forward, recording tactical success at every step. Their operational indicators were entirely green: territory taken, momentum sustained, enemy retreating.

Yet with every forward step, the Romans were pressing deeper into the geometry of Hannibal’s double envelopment. Their forward movement progressively consumed their flank maneuverability. They entered a viability trap: a trajectory that remains locally successful while progressively destroying globally viable alternatives.

Many scale-up and FOAK projects follow this exact dynamic. Equipment is ordered, debt is drawn, capacity is built, and milestone payments are triggered. The dashboard shows green. Yet, every step increases commitment velocity while shrinking the set of reachable alternative trajectories.

Mathematically, the derivative of the reachable viable set turns negative:

d|𝒪(t)| / dt < 0

A project can be improving in present-state metrics while deteriorating in future-state optionality. Optionality isn’t what else a project could do in theory; it’s what else it can actually reach before liquidity, permit windows, or debt covenants expire. A project can be fully compliant today, yet already doomed by the derivative of its shrinking maneuverability space.

Four Archetypes of Project Fragility

Mechanism	Archetype	Core Failure
Absorption Fragility	Everglades	Nowhere for disturbance to reside
Rigidity Fragility	Agincourt	Insufficient permissible deformation
Trajectory Entrapment	Cannae	Success consumes reachable alternatives
Identity Fragility	Theseus	Adaptation requires becoming a different project

Identity fragility represents the deepest contractual trap. It occurs when an alternative trajectory that would save the enterprise is one the financing documents no longer recognize as the same project. A different feedstock, customer, product, or operating configuration might be economic—but not covenant-compliant, warranty-covered, or tax-qualified.

When the viability corridor V(t) and the project’s contractually defined identity corridor I(t) cease to overlap:

V(t) ∩ I(t) → ∅

This is identity overfitting: specifying the project so precisely around the world in which it was financed that it loses the ability to remain itself in the world that actually arrives.

Structured Slack: Space, Time, and Information

If an exoskeleton creates rigidity, the alternative is not chaos or unconstrained flexibility—it is structured slack.

Structured slack acts as a system shock absorber, operating across a clear mechanics chain:

Slack → Space + Time → Better Information

• Space to move: Room to adjust operational profiles, switch inputs, or pivot off-takers.
• Permission not to move yet: Time to allow uncertainty to resolve before making irreversible commitments.

A six-month schedule buffer or reserve account does not merely absorb six months of stress. It buys management time to learn whether a shock is temporary, structural, local, or systemic before spending the project’s remaining viability delta-v. Slack preserves maneuvering capacity precisely because it makes immediate maneuver unnecessary.

Structured slack is not unallocated waste or management inefficiency; it is temporary viability volume. A reserve account is a financial wetland. Schedule buffers are temporal wetlands. Excess utility capacity or dual-fuel flexibility are operational wetlands. They allow SWELL to arrive, occupy space, and dissipate without instantly triggering default cascades.

The New Architecture of Diligence

Optimization and resilience solve fundamentally different objective functions:

• The optimized project asks: How efficiently can we execute the expected trajectory?
• The resilient project asks: How many plausible departures from that trajectory can we survive?

Traditional project finance diligence focuses heavily on the first question, asking whether enough uncertainties can be contracted into acceptable risks. For complex assets facing volatile environments, diligence must expand to encompass three deeper inquiries:

• The Load Question: When the world changes, where does the load travel?
• The Displacement Question: If this part of the project cannot move, what must move instead?
• The Viability Question: After it moves, is there still a reachable state in which the project remains both viable and itself?

A buttress supports load. An exoskeleton suppresses deformation. A wetland absorbs disturbance. A resilient project needs all three in the right places.

The purpose of resilience is not to prevent the project from moving. It is to preserve enough places for load to go, enough degrees of freedom for the project to move, and enough time for management to learn which movement the new world requires.

We have spent a generation learning how to contain uncertainty in project finance—transferring it, fixing it, guaranteeing it, reserving against it, specifying where it must go. But uncertainty was never eliminated. We merely changed the channels through which it could move. The ultimate test of project architecture isn’t whether the storm arrives. It is whether, when it does, we have left the asset an Everglades.

Category: Top Stories