we move to situations where binding agreements let players pool and reallocate payoffs
You run a $4M ARR SaaS product. A competitor with $6M ARR approaches you: 'We both use the same cloud vendor. If we negotiate together as a consortium, we can get volume pricing that saves $900K per year total. Split it 50/50?' Your Expected Payoff from negotiating alone is maybe $200K in savings. Theirs is $350K. Together you unlock $900K - but 50/50 means you get $450K and they get $450K. They brought more volume to the table. Should you take the deal, counter with a different split, or walk away?
Binding agreements let players in a Game Theory scenario commit to pooling their payoffs and redistributing the surplus - the value created above what each party could earn alone. Operators use them to structure mergers, consortium deals, cost sharing arrangements, and Vendor Negotiations where cooperation beats going it alone.
In basic Game Theory, every player acts independently. You pick your strategy, your competitor picks theirs, and the Expected Payoff falls where it falls. Nobody can make enforceable promises.
Binding agreements change the game. They let players:
The critical word is binding. A handshake promise to cooperate is just a promise without enforcement - the other side can walk away at no cost. A binding agreement has teeth: a signed contract with penalties for defection, a merged entity where separation means accepting Liquidation Discounts, or a consortium with committed capital.
Why does binding matter? Because without enforcement, each player has an incentive to defect. You agree to share infrastructure costs, your partner sees your architecture, then renegotiates from a stronger position. Binding agreements solve this by making defection more expensive than cooperation.
The two most common binding agreement problems an Operator faces:
Mergers. Two PE portfolio companies merge. The combined EBITDA is higher than the sum of the standalone numbers - shared overhead disappears, Expansion Revenue flows from selling across both customer bases. How do you split Enterprise Value between the legacy shareholders? The entire M&A due diligence process is a binding agreement problem: compute standalone values, project combined value, verify the surplus is real, and structure the deal so neither side can defect after signing.
Cost sharing. Three business units share a platform team. The Cost Center Budget is $1.2M. Each unit would pay $600K to build its own. How do you Allocate the shared cost fairly? Who pays what determines whether each unit's P&L shows a win - and whether any unit is tempted to leave the arrangement.
In both cases, cooperation creates surplus - value that didn't exist when players acted alone. Your P&L impact depends entirely on how that surplus gets divided.
Before any negotiation, calculate what each party earns without this specific agreement. This is their Outside Option - their best alternative to the deal on the table.
Important precision: the Outside Option is not simply your Expected Payoff from non-cooperative Game Theory in general. It is the best you can do absent this particular agreement. If you have other potential partners or alternative deals, those raise your Outside Option.
The surplus is the total value created by cooperation minus the sum of all Outside Options.
Surplus = Joint Payoff - (Your Outside Option + Their Outside Option)
Using the consortium example from the hook: $900K - ($200K + $350K) = $350K in surplus.
If the surplus is zero or negative, there's no deal to make. Walk away.
This is where Bargaining happens. Common approaches:
The agreement needs enforcement mechanisms that make defection more costly than compliance:
Use binding agreement analysis when all three conditions hold:
Skip it when:
You're an Operator evaluating a merger of two PE portfolio companies.
First, a note on Valuation: in this industry, Enterprise Value is calculated as a ratio of annual EBITDA. Comparable standalone companies trade at roughly 6 times EBITDA - meaning each dollar of recurring earnings translates to $6 of Enterprise Value. This ratio reflects how the market prices predictable earnings streams. Larger combined platforms often command a higher ratio because buyers pay a premium for scale.
Compute the surplus: $45.5M - $30M = $15.5M in Value Creation from the merger.
Compute each company's Outside Option: A walks away with $18M, B walks away with $12M. These are the floors.
Equal surplus split: A gets $18M + $7.75M = $25.75M. B gets $12M + $7.75M = $19.75M. Total = $45.5M. Check: passes.
Shapley value approach: Consider every ordering in which A and B could form the merged entity. A's marginal contribution when joining B = $45.5M - $12M = $33.5M. A's marginal contribution when going first (standalone) = $18M. Average = ($33.5M + $18M) / 2 = $25.75M. B by symmetry = $45.5M - $25.75M = $19.75M. Key insight: For two players, the Shapley value always equals the equal-surplus split. This is not a coincidence - it is a mathematical property of the Shapley formula when there are only two participants.
Structure the binding agreement: Merger agreement where A receives 56.6% of combined equity ($25.75M / $45.5M) and B receives 43.4%. Tie $2M of the purchase price to milestone-based payments contingent on hitting the $6.5M combined EBITDA target, so neither side can coast after signing.
Insight: The surplus ($15.5M) is what makes the deal worth doing. The split determines who captures it. Without binding agreement analysis, Company A might accept a naive 60/40 split ($27.3M / $18.2M) - overpaying by $1.55M relative to the equal-surplus benchmark. That's capital discipline.
You're the Operator responsible for a shared platform Cost Center. Three business units need the platform:
Determine which groups actually benefit from sharing. The platform costs $1M whether one unit uses it or three. Any single unit is better off building alone (every standalone cost is below $1M). For two-unit groups: X+Y standalone = $1.4M > $1M (saves $400K); X+Z = $1.2M > $1M (saves $200K); but Y+Z = $1M = $1M (break-even, zero savings). This asymmetry drives the entire Shapley calculation.
Naive equal split: $1M / 3 = $333K each. Unit Z's Outside Option is $400K, so Z saves just $67K while Unit X saves $467K. Technically no unit pays more than their standalone cost, but the surplus distribution is lopsided. Can we do better?
Shapley value: compute each unit's average marginal contribution across all 6 possible orderings. Define v(group) as the total savings that group captures by sharing the platform versus each member building separately.
v({X}) = $0, v({Y}) = $0, v({Z}) = $0
v({X,Y}) = $400K, v({X,Z}) = $200K, v({Y,Z}) = $0
v({X,Y,Z}) = $800K
Unit X - marginal contribution in each ordering:
Shapley(X) = ($0 + $0 + $400K + $200K + $800K + $800K) / 6 = $367K
Unit Y - marginal contributions across the 6 orderings: $0, $0, $400K, $0, $600K, $600K.
Shapley(Y) = $1,600K / 6 = $267K
Unit Z - marginal contributions: $0, $0, $200K, $0, $400K, $400K.
Shapley(Z) = $1,000K / 6 = $167K
Check: $367K + $267K + $167K = $800K. Matches total surplus.
Convert savings to cost Allocation. Each unit pays their standalone cost minus their Shapley-allocated savings:
Compare to proportional Allocation (split $1M by standalone cost ratios): X pays $444K, Y pays $333K, Z pays $222K. Shapley charges X $11K less and Z $11K more. Why? Because X is disproportionately critical - without X in the group, {Y,Z} generates zero savings. Shapley rewards X for being the unit that makes the shared platform worth building.
Make it binding: Internal cost sharing agreement with a 2-year commitment. Any unit that exits pays a penalty equal to one year of their Allocation. This prevents a unit from using the shared platform to defer building its own, then leaving once they have captured the institutional knowledge.
Insight: The Shapley value gives a different answer than proportional Allocation whenever one player is disproportionately critical to the group's success. Without computing all 6 orderings, you cannot see this. The full calculation revealed that X's presence is what makes the platform viable for everyone - and the fair Allocation reflects that.
Binding agreements create value by letting players cooperate in ways that non-cooperative Game Theory can't support - but only when the surplus is positive and defection is tempting.
Always compute Outside Options first. They set the floor for any deal. If your share of the joint payoff is less than your Outside Option, walk away.
The surplus split is the entire negotiation. Equal split, proportional Allocation, and Shapley value are frameworks - not answers. Which one applies depends on Leverage, contribution asymmetry, and what each side can credibly threaten.
Treating the total pie as the prize instead of the surplus. If a merger creates $45M in Enterprise Value but the standalone companies are worth $30M combined, you're negotiating over $15M - not $45M. Anchoring on the total pie leads to overpaying.
Assuming cooperation is self-enforcing. Without binding mechanisms (contracts, structural commitment, exit penalties), rational players defect whenever short-term gains exceed long-term cooperation value. If the agreement isn't enforceable, model it as non-cooperative Game Theory and plan accordingly.
You're an Operator at a Holding Company that owns two consumer brands. Both brands have been independently buying ad slots in the same industry publication's quarterly catalog. There are 20 premium slots per issue. When both brands bid, they drive each other's prices up: each wins about 10 slots at $8K per slot ($80K per brand per quarter).
You propose a binding internal agreement: each brand commits to bidding on 10 designated slots with no overlap. Without the internal competition, the price per slot drops to $5K. Each brand gets 10 slots for $50K per quarter.
Calculate: (a) Each brand's Outside Option, (b) the quarterly surplus from the agreement, (c) why this agreement needs a binding mechanism even though both brands report to the same parent.
Hint: The Outside Option is what each brand gets without the deal: their current cost for their current volume. The surplus is the total savings from eliminating the internal bidding war. For part (c), think about how business unit P&L incentives can diverge from parent-level Allocation goals.
(a) Outside Option: Without the agreement, each brand pays $80K/quarter for 10 ad slots. That is the baseline - what you get by walking away from the deal.
(b) Under the agreement, each brand pays $50K for 10 slots. Savings per brand: $30K/quarter. Total surplus: $30K + $30K = $60K/quarter. Both brands get the same slot volume at lower Cost Per Unit ($5K vs $8K per slot). The surplus is symmetric, so an equal split is the natural outcome.
(c) Even within the same Holding Company, business unit incentives diverge from parent incentives. Each brand's P&L owner wants maximum ad coverage. Without binding Allocation rules, Brand A's marketing team might quietly bid on Brand B's designated slots to grab more impressions - especially near the end of a quarter when they are chasing Revenue targets. The binding mechanism (a formal internal policy approved by the parent, with Budget adjustments as the enforcement tool) prevents this defection. Without it, you are back to the non-cooperative outcome where both brands overpay.
Three SaaS companies want to form a consortium to negotiate with a cloud vendor. Company A spends $2M/year, B spends $1.5M, C spends $500K. The vendor offers a 25% discount on the combined $4M spend if all three sign a 3-year contract. Assume no discount is available to any company negotiating alone, and a 15% discount is available to any two-company pair (applied to that pair's combined spend).
Using proportional Allocation and Shapley value, propose two different ways to split the $1M annual savings. Which split is more stable?
Hint: For proportional Allocation: split savings in proportion to each company's spend. For Shapley value: define v(group) as the savings each possible group achieves, then compute each company's average marginal contribution across all 6 orderings. The two-company discount tiers create asymmetric marginal contributions.
Proportional split: A gets $1M x ($2M/$4M) = $500K. B gets $375K. C gets $125K. Each saves exactly 25% of their spend. Simple, transparent.
Shapley value: Define the savings for each possible group:
Company A's marginal contribution in each ordering:
Shapley(A) = $2,300K / 6 = $383K
By same method: Shapley(B) = $346K (contributions: $0, $0, $525K, $300K, $625K, $625K). Shapley(C) = $271K (contributions: $0, $0, $375K, $300K, $475K, $475K). Check: $383K + $346K + $271K = $1M.
Which is more stable? The Shapley split. Under proportional Allocation, C saves only $125K but must commit to a 3-year contract - a thin surplus for a long commitment. C might walk away, collapsing the consortium. Under Shapley, C saves $271K because C is the player that unlocks the jump from the 15% to the 25% discount tier. That larger share correctly reflects C's marginal contribution and gives C enough incentive to stay.
A might resist the Shapley split ($383K vs $500K under proportional), but A's Leverage is limited: A's Outside Option is $0 (no solo discount), so A cannot credibly threaten to walk away.
You're acquiring a company for $10M. Your standalone EBITDA is $4M; theirs is $1.5M. Post-merger, you project $7M EBITDA from eliminating duplicate overhead and capturing Expansion Revenue across both customer bases.
In your industry, Enterprise Value is calculated as a ratio of annual EBITDA. Comparable companies trade at 8 times EBITDA - meaning each dollar of recurring earnings translates to $8 of Enterprise Value. This ratio reflects how the market prices predictable earnings in this sector.
The seller wants $14M. Should you pay it? What's the maximum you should offer?
Hint: Compute post-merger Enterprise Value using the 8x ratio, subtract your standalone Enterprise Value, and that gives you the maximum the acquisition is worth to you. The surplus is the gap between combined value and sum of standalone values. The Bargaining range runs from the seller's Outside Option to the point where you capture zero surplus.
Your standalone Enterprise Value: $4M x 8 = $32M. Theirs: $1.5M x 8 = $12M. Sum: $44M. Combined: $7M x 8 = $56M. Surplus = $56M - $44M = $12M.
The acquisition creates $12M in surplus for the combined entity. Their Outside Option is $12M (standalone Enterprise Value). Your maximum price = their Outside Option + the entire surplus = $12M + $12M = $24M. But paying $24M means you capture zero surplus - a terrible deal.
At $14M: You pay $2M above their standalone value. You capture $12M - $2M = $10M of the surplus. Good deal.
At $10M: You pay $2M below their standalone value. They would reject it - it is below their Outside Option.
Decision: The $14M ask falls within the Bargaining range ($12M to $24M). Counter at $13M to capture more surplus, but $14M is defensible. Never pay above $20M (that leaves you less than a third of the surplus you are creating). Tie $2M of the purchase price to milestone-based payments contingent on hitting the $7M combined EBITDA target, so the surplus projection is verified before full payout.
Binding agreements connect cooperative and non-cooperative Game Theory. The Outside Option links to Expected Payoff, but precisely - it is your best alternative to this specific agreement, not your Expected Payoff from some broader game with multiple equilibria. Surplus is Value Creation measured in P&L terms. Downstream: Bargaining covers how surplus gets divided under asymmetric Leverage. Shapley value provides the mathematically fair split when three or more parties must agree. M&A due diligence is the full application - compute standalone Valuation, project combined Enterprise Value, verify the surplus, and structure enforcement.
Disclaimer: This content is for educational and informational purposes only and does not constitute financial, investment, tax, or legal advice. It is not a recommendation to buy, sell, or hold any security or financial product. You should consult a qualified financial advisor, tax professional, or attorney before making financial decisions. Past performance is not indicative of future results. The author is not a registered investment advisor, broker-dealer, or financial planner.