The backup worth
buying together.

A backup right can appear worthless on its own and become essential when another right changes the worst disruption the buyer faces. Research on LNG procurement turns that interaction into a practical test for purchasing decisions.

A procurement committee asks a reasonable question before paying for extra flexibility: what does this option buy us? If every proposed right must justify itself independently, the committee can reject a combination that would materially improve its position. The difficulty arises when each purchase changes the problem the remaining purchases need to solve.

My paper, Buying Joint Deliverability: Awardable Exercise Rights in Emergency LNG Procurement, studies that decision. It connects the terms of a tender with the economics of a backup portfolio, asking when rights should be evaluated together and when the simpler standalone screen is reliable.

Start with what the buyer may award

An LNG buyer may need cargoes for several delivery windows. A supplier can appear in more than one column of the price table. Before treating those offers as a complete backup plan, the buyer needs to establish whether it may accept them together. The relevant object is a combination of commitments permitted by the tender terms.

The June 2026 LNGT67 documents provide a concrete setting. The financial evaluation contains BP and PetroChina quotations for both cargo windows, while Vitol quotes the first. The bid document permits a compliant bidder to submit separate unconditional bids for multiple cargoes and allows the buyer to accept more than one.

That establishes commercial two-cargo awardability during the applicable bid validity. Physical cargo control, route independence, actual awards and completed deliveries require their own evidence. In this paper, “joint deliverability” names the commercial right to award the offers together. Keeping that definition clear makes the subsequent comparison usable.

How the worst case changes

Consider the paper’s two-window, three-supplier model. Supplier A is the cheapest and quotes both windows. Supplier B also quotes both, while supplier C quotes only the first. Begin with a counterfactual rule limiting A and B to one cargo each, then consider rights that allow either supplier to serve both windows.

The buyer evaluates a declared set of scenarios in which one supplier is removed. It purchases rights before learning which scenario occurs, then selects the cheapest permitted basket afterward. The objective combines the worst procurement cost across those scenarios with the fees paid for the rights.

When A disappears, B’s one-cargo limit forces C into the first window and B into the second. Buying extra awardability from A does nothing to improve that case: A is absent. Its standalone benefit against the worst case can therefore be zero.

Allowing B to supply both windows relieves that constraint. A different contingency can then become the limiting one: B disappears while A remains restricted to one cargo. The right that initially appeared worthless now addresses the new worst case. Buying both can produce a gain that neither standalone assessment reveals.

A purchasing failure with observed price inputs

The paper calibrates this mechanism to the quoted LNGT67 menu. It uses documented two-cargo awardability as one configuration and the one-cargo restriction as a counterfactual. The comparison holds scenario prices fixed and uses specified supplier-removal assumptions, so its result is a conditional valuation of contract rights.

For stipulated equal-energy cargoes, the gross worst-case gain from the cheapest supplier’s right alone is zero. The alternative supplier’s right has a small standalone gain. Their combined gain is substantially larger because the pair resolves both limiting contingencies. At the manuscript’s illustrative fees, neither individual purchase passes a strictly positive net-gain screen, but the joint purchase does.

The gross worst-scenario cost gap between allowing both dual-window suppliers two cargoes and limiting each to one is 1.38885 US dollars per MMBtu in the stated equal-energy comparison, before acquisition fees. That figure describes the model evaluated at the observed prices. Establishing realised savings would additionally require actual exercise terms, awards, performance and delivery evidence.

Make the bundle auditable

A useful procurement rule must show why a proposed bundle is worth its fee. The paper supplies a scenario-by-scenario certificate. Start with the baseline’s worst cost. For each scenario, record how far its own cost lies below that maximum, then add the improvement the proposed rights would create in that scenario.

The smallest of those totals is the bundle’s gross improvement in worst-case procurement cost. The bundle is worth its fee only when every scenario clears the corresponding hurdle. This exposes the blocking contingency rather than hiding it inside a single portfolio score.

The calculation also explains why improving today’s worst scenario may be insufficient. Another scenario may already sit close behind it and become the new constraint. A committee can trace disagreement to the proposed fee, the applicable price, the award rule or the scenarios being tested.

Know when the simple screen is enough

The research gives standalone assessment a clear safe setting. If rights contribute additively within each scenario and the same scenario remains worst for every candidate portfolio, each right can be compared independently with its fee. Under those conditions, the standalone rule is exactly optimal.

When that structure cannot be defended, the buyer needs to consider combinations. The paper shows that hidden complementarity can extend beyond pairs: in a restricted model class, every smaller subset can have zero worst-case gain while the complete bundle creates value. Auditing a named bundle and discovering the best bundle are consequently different tasks.

Turn flexibility into an executable plan

Implementation begins with an evidence audit: prices, validity periods, joint-award rules, exclusivity and common constraints. A right already included in the baseline cannot be purchased again as an incremental improvement. Commercially permitted baskets must also pass a physical receiving assessment at the relevant operating state.

The stress-test family matters. If both suppliers that quote the second window disappear, the quoted menu cannot cover that window. Extra joint-award rights do not repair that infeasibility. Repricing or common-route losses can likewise change the recommendation and require a different scenario calculation.

The interdisciplinary contribution is a decision process that joins contract interpretation, operations modelling and resilience economics. Count the commitments that can be exercised together. Identify which contingency limits the portfolio. Test combinations when that limiting contingency can change. The value of backup supply then becomes a property of the plan the buyer can execute, rather than a score attached to each option in isolation.