> ## Content Index > Fetch the complete content index at: https://adikumar.co/llms.txt > Use this file to discover other available public pages before exploring further. # Deal structure, reps and warranties, escrow architecture - URL: https://adikumar.co/deal-structure-reps-and-warranties-escrow-architecture/ - Published: 2026-08-30T10:20:57.000Z - Updated: 2026-10-07T05:39:03.000Z - Description: How DD findings translate into the transaction paper. Structure choice, R&W with insurance, escrow with DD-informed carve-outs, earnout mechanics. - Author: Adi Kumar - Tags: Due Diligence for the AI Buildout, AI Infrastructure, M&A, #Essay [Due Diligence for the AI Buildout](https://adikumar.co/due-diligence-for-the-ai-buildout/) · Part 12 of 14 · Structural + paper · [Framework page](https://adikumar.co/framework-due-diligence/) Due Diligence for the AI Buildout · Part XII of XIV # Deal structure, reps and warranties, escrow architecture How the DD findings translate into the transaction paper. Structure choice, representations and warranties, escrow architecture, earnout mechanics, and indemnity carve-outs. The transaction paper is where DD findings from Stages 1 and 2 either land as protection or leak. Due Diligence for the AI Buildout · fourteen essays for practitioners 1. **Part I.** The scoping decision 2. **Part II.** Team, timeline, data room 3. **Part III.** Market and demand-side DD 4. **Part IV.** Customer, pricing, competitive DD 5. **Part V.** Product architecture and technology DD 6. **Part VI.** Manufacturing, IP, supply chain DD 7. **Part VII.** Quality of earnings and working capital DD 8. **Part VIII.** Capex, growth economics, stress-testing 9. **Part IX.** Organisational, key personnel, culture, systems 10. **Part X.** Legal and contract DD 11. **Part XI.** Regulatory, CFIUS/HSR, environmental, cyber 12. **Part XII.** Deal structure, reps and warranties, escrow *(you are here)* 13. **Part XIII.** Financing structure, covenants, closing risk 14. **Part XIV.** Post-close validation and the DD improvement loop Supplements 1. **A.** The Question Banks (downloadable) 2. **B.** Worked example: Eaton / Boyd Thermal end-to-end 3. **C.** Seller-side prep: the DD workbook you should have ready #### Executive summary - Deal structure choice (share vs asset, cash vs stock, all-in vs earnout) rests on the tax treatment, on how liability transfers to the acquirer, and on what the financing structure requires. Share deals are simpler but transfer all liabilities including undiscovered ones; asset deals allow selective assumption but require third-party consents on transferred contracts. - Reps and warranties allocate risk between buyer and seller for issues discovered post-close. The R&W insurance market has grown substantially since 2018 and now covers most mid-market and large-cap industrial-tech deals. Insurance-backed R&W changes negotiation dynamics because the buyer’s recovery is against the insurer, with recourse to the seller receding. - Escrow architecture determines how much money is in escrow, for how long, and what triggers release. Standard mid-market escrow: 10-15 percent of purchase price for 18-24 months. Deal-specific carve-outs for tax, litigation, environmental, and IP infringement extend beyond the general escrow. - Earnout structure works when the deal is priced on future performance the buyer cannot fully diligence. Structure that works: milestones that have binary triggers, use defined measurement periods, and are protected by explicit buyer covenants around operating the target. Structure that produces litigation: vague milestones, disputed measurement, buyer changes to operating plan that suppress earnout achievement. - Three recurring failures: structure chosen for tax without regard to liability transfer, R&W architecture that leaves specific DD findings uncovered, escrow sized against generic benchmarks; deal-specific risk sizing stays absent. ## DD does not eliminate uncertainty. It determines where the uncertainty sits after closing WS03 through WS11 identified and tested the risks in the target. WS12 is the workstream where those risks are allocated to specific parties through specific mechanisms. Absent that allocation, the buyer owns them all by default. **DD does not eliminate uncertainty. It determines where the uncertainty sits after closing.** Every material risk in the transaction ends up with a party: the seller (through indemnity or escrow), an insurer (through R&W policy), a lender (through financing structure), the target's management team (through retention structure), the customer or supplier (through contractual restructuring), or the buyer (through explicit acceptance in the underwriting). The workstream succeeds when every material risk has been consciously assigned. It fails when risks are noted; assignment to parties stays absent. **A risk that is neither eliminated, priced nor contractually allocated is simply being accepted by the buyer, whether the IC memo admits it or not.** ### Risk and deal protection measure different things DD identifies risk. WS12 selects the mechanism that transfers, prices, eliminates or contains it. The two are frequently conflated in conventional PE DD, and the conflation produces the standard failure mode where a serious finding results in an escrow that does not match the risk it purports to protect against. Suppose the DD workstream identifies IP ownership uncertainty. Seven distinct responses are available, and they stay non-interchangeable. 1. **Walk away.** If the risk is fundamental to the underwriting. 2. **Reduce price.** If the risk is economically quantifiable and the buyer is willing to take the expected loss at an adjusted valuation. 3. **Escrow.** If the exposure is quantifiable and time-bounded, and cash security is the appropriate protection. 4. **Specific indemnity.** If the risk is identifiable but the loss magnitude is uncertain and potentially uncapped. 5. **Broad representation or warranty backed by R&W insurance.** If the risk is one of the ordinary category of unknown historical issues. 6. **Closing condition.** If the risk must be resolved before ownership transfers. 7. **Covenant.** If the risk can be fixed post-close through defined management action. The DD finding does not determine the mechanism. The nature of the risk (known vs unknown, historical vs prospective, quantifiable vs uncertain, reversible vs irreversible, time-bounded vs open-ended) determines the mechanism. WS12 succeeds when the match is deliberate. ### The four risk buckets Every material DD finding falls into one of four risk buckets, and each bucket has a different appropriate protection mechanism. | Bucket | Definition | Appropriate mechanism | | ------------------------ | ----------------------------------------------------------------- | ----------------------------------------------------------- | | **A. Seller risk** | Historical facts, unknown liabilities, breaches of representation | R&W policy + specific indemnity + escrow | | **B. Buyer risk** | Future performance, market outcome, execution | Valuation adjustment + financing structure + operating plan | | **C. Shared risk** | Uncertain outcomes both parties can influence | Earnout / contingent consideration / mutual covenants | | **D. Deal-breaker risk** | Cannot reasonably be allocated to either party | Closing condition or walk-away | The classic failure mode is trying to insure a buyer risk (Bucket B) with a seller-risk mechanism (R&W or indemnity). If the underwriting depends on Customer X continuing to buy and Customer X stops buying because industry demand falls after close, that is business risk. R&W insurance does not make Customer X buy. You cannot shove every uncertainty into representations and warranties; the mechanism must match the bucket. ## Structure choice Share deals transfer the entity (and everything in it) to the buyer. Simpler execution, no third-party consents required for contract transfer. All liabilities (known and unknown) move with the equity. Tax treatment for the seller is generally capital gains; tax step-up for the buyer stays unavailable. Asset deals transfer specified assets and liabilities. More complex execution, requires third-party consents on transferred contracts, but permits selective liability assumption. Tax treatment for the seller is typically ordinary income on the intangibles; tax step-up for the buyer is available. The consent burden is often the practical deciding factor; tax analysis sits downstream: a target with a hundred material customer contracts, forty of which include anti-assignment language, will take four to six months of consent work to complete the transfer, and every consent conversation is an opportunity for the customer to renegotiate. Where the target is in a heavily contracted commercial environment, share deals get chosen despite the messier liability profile because the consent friction of an asset deal is worse than the liability inheritance. Hybrid structures (F-reorganisations, 338(h)(10) elections) combine features. Structure choice should follow the DD-identified risks: significant undisclosed liability exposure argues for asset structure; clean target with clean liability profile argues for share simplicity. Two variables that sit alongside the share / asset / hybrid choice: consideration mix and rollover equity. Consideration mix (stock, cash, seller notes, contingent value rights) allocates transaction risk between the parties, a share-for-share exchange keeps the seller exposed to combined-company performance and can be tax-efficient, but concentrates seller wealth in the buyer’s equity in a way many sellers resist. Rollover equity (management team rolls a portion of existing equity into the go-forward entity) is standard in sponsor-adjacent industrial-tech deals and does substantial work: it aligns key personnel with post-close performance, deferring tax on the rolled portion, and signals seller confidence in the plan. A typical rollover is in the 10-25 percent range for founder-led targets, higher for sponsor-to-sponsor secondaries where retention of the operating team is the deal thesis. #### Chart 1\. Structure choice: what each optimises for Share deals optimise for execution simplicity. Asset deals optimise for liability filtering. Hybrid structures balance tax efficiency with liability control. Structure trade-offs illustrative; specific deals require tax counsel. ## The seven-level protection hierarchy Protection mechanisms sit on a hierarchy ordered by how completely they insulate the buyer from the underlying risk. The strongest mechanism (fix before close) is preferred when available; escalation down the hierarchy is a conscious choice; default fallback stays weaker. Every DD finding should be classified against the highest level of the hierarchy that is achievable within the deal timeline. 1. **Level 1\. Fix before close.** Risk is resolved before ownership transfers. Best possible outcome; residual buyer exposure is zero. 2. **Level 2\. Closing condition.** Buyer refuses to close until specific evidence of resolution arrives. Preserves the option to walk if resolution does not materialise. 3. **Level 3\. Price adjustment.** Buyer compensated upfront through reduced consideration. Buyer bears the residual, but at an adjusted entry price. 4. **Level 4\. Specific indemnity.** Seller retains identified risk on defined terms with named survival period. 5. **Level 5\. Escrow.** Cash security held against specific claims or general R&W breach. Bounded amount; bounded duration. 6. **Level 6\. R&W insurance.** Third-party risk transfer. Subject to policy exclusions, retention and coverage limits. 7. **Level 7\. Covenant.** Post-close remediation obligation on the seller (or on the buyer). Depends on ongoing seller-side capacity and willingness. Below Level 7 sits a category that goes unlabelled: the buyer accepts the risk explicitly, either at the underwritten valuation or with an offsetting operational plan. That is a legitimate outcome when the mechanism cost exceeds the expected loss. It stays illegitimate when it happens by default because none of the higher-level mechanisms were pursued. ## Effective protection: amount × coverage × duration × enforceability × collectability The nominal amount of protection is one of five components. Effective protection is the compound of all five, and where any component is materially weaker than the others, the effective protection may be a small fraction of the nominal number. 1. **Amount.** How much money is nominally available (escrow amount, indemnity cap, R&W policy limit)? 2. **Coverage.** Does the mechanism cover the specific loss category the risk produces? Broad exclusions in R&W policies, coverage carve-outs in specific indemnities and scope limitations in escrow release mechanics can materially compress effective coverage. 3. **Duration.** Does the protection last as long as the underlying risk can crystallise? A five-year exposure with an 18-month escrow falls short of five years of protection. 4. **Enforceability.** Can the buyer recover through the mechanism? Onerous claim procedures, short notification periods, mandatory arbitration in unfavourable jurisdictions and broad seller defences all reduce enforceability. 5. **Collectability.** Who stands behind the mechanism? A $50M indemnity from a seller entity with no continuing assets falls short of $50M of protection. **Effective protection = amount × coverage × duration × enforceability × collectability.** Not a literal financial formula, but the reasoning framework the deal team should apply before crediting nominal protection against a material risk. ### The duration-mismatch veto The most consequential of the five components is duration matching. Every material risk has a specific expected crystallisation timeline (a regulatory claim can emerge 3-5 years after close; a tax exposure can surface at any audit within the statute-of-limitations period; a customer-CoC-consent breach materialises within months of close). The protection mechanism must survive the crystallisation window. **The duration-mismatch principle: if the protection expires materially before the risk can reasonably crystallise, the protection should not be credited as full risk mitigation.** A 24-month escrow against a 5-year regulatory exposure provides bounded protection during the first 24 months and zero protection during years 3-5\. Nominal $20M escrow, effective protection substantially lower once duration mismatch is priced in. The workstream flags every material risk where the protection duration falls short of the reasonable crystallisation window as a duration-mismatch item requiring escalation, either by extending the protection, restructuring to a different mechanism, or explicitly pricing the residual buyer exposure into the valuation. ### Enforceability and seller creditworthiness A $30M indemnity is worthless if the seller has no assets left post-close to satisfy a claim. This is particularly relevant when the seller is an individual founder extracting substantially all proceeds, a limited-life PE fund approaching wind-down, or a special-purpose vehicle designed to dissolve after distribution. Screen every material indemnity against seller-side collectability. Where the seller entity's post-close balance sheet is materially weaker than the indemnity commitment, the effective protection is limited to whatever mechanism stands behind the indemnity: escrow reserve, R&W policy backing, parent-company guarantee, or (in the limit) nothing. Nominal indemnity without collectability falls short of underwriting-grade protection. ## Known risk versus insured risk: R&W does not solve known problems R&W insurance is designed for unknown breaches of ordinary representations discovered post-close. It stays outside the scope of insuring known risks the DD workstream identified. Known risks (a specific pending litigation matter, a specific tax audit in progress, a specific customer relationship in dispute) are typically excluded from R&W coverage, must be addressed through specific indemnity or escrow, or must be priced into the valuation directly. The default posture that "we will just get R&W insurance" collapses when the underwriter's diligence phase excludes the specific risk the buyer most cares about. Known risk ≠ insured risk. Specific indemnity is the appropriate mechanism for known unknowns; R&W policy is the appropriate mechanism for unknown unknowns; the two operate in parallel (not as substitutes). ## Closing conditions are stronger than escrow When a risk can be resolved before ownership transfers, closing condition (Level 2) is materially stronger than escrow (Level 5). Escrow compensates the buyer for a loss that has already occurred; closing condition preserves the buyer's option to walk if the loss cannot be prevented. Illustrative cases where closing condition is the correct answer: - Unresolved customer change-of-control consent on a top-3 customer contract. Escrow does not preserve the customer relationship; consent does. - Unresolved FDI or national-security approval. Escrow does not make the transaction legal to close; approval does. - Unresolved supplier CoC consent on a critical single-source input. Escrow does not preserve production continuity; consent does. - Unresolved specific-indemnity item where the risk is loss of a specific operating licence. Escrow compensates for financial loss but does not restore the licence. Escrow is a compensation mechanism. Closing condition is a preservation mechanism. Where the underlying business asset cannot be restored by cash, closing condition is the required protection level. ## Reps and warranties architecture R&W insurance reduces friction with the seller, the buyer’s recourse for breach runs to the insurer; the seller’s equity sits further from reach, but the claims process with the R&W carrier is its own workstream. The carrier is a new counterparty with its own economic incentives; claims that involve interpretation of the DD file, of the disclosure schedules, or of the underlying breach itself frequently take twelve to twenty-four months to resolve and often involve legal costs the buyer pays through. R&W is friction reduction with the seller and risk transfer to a carrier; automatic recovery stays absent. Representations and warranties are seller statements about the state of the business. Breach entitles the buyer to indemnification. R&W architecture has three variables: which reps are made (scope), the survival period (how long the buyer can claim), and the recovery mechanism (escrow, insurance, or direct claim). R&W insurance has changed the market. Insurance policies commonly cover 10-15 percent of enterprise value, with survival periods around 6 years for fundamental reps and 3 years for general reps (deal-specific variation). The insurance market has expanded substantially since 2018 and now covers most deals over $50M. Cost per major broker guidance runs in the 2.5-4 percent of coverage limit range for standard risk, higher for complex or contested underwriting. Insurance-backed R&W changes the negotiation. Buyers recover from insurers; recourse to sellers recedes, which reduces post-close litigation friction with the seller. The trade-off is that the insurer becomes a new counterparty in the claims process, with its own underwriting incentives, and recovery stays non-automatic. Sellers can accept broader reps because breach doesn’t trigger clawback. Buyers can pursue smaller claims that direct-seller recovery would make uneconomic. #### Chart 2\. R&W insurance market: coverage and cost trends R&W insurance covers most mid-market and larger deals now. Cost has trended down as the market has matured. Coverage limits have expanded. Market data from Marsh, Aon, WTW published R&W insurance trends 2018-2026. ## Escrow architecture Escrow benchmarks depend on whether R&W insurance is in place: uninsured mid-market deals typically hold 10-15 percent of purchase price for 18-24 months, while insured deals compress to a retention layer of 0.5-1 percent plus the insurance tower. Both patterns then add deal-specific carve-outs. Deal-specific carve-outs (tax, environmental, litigation, IP infringement) extend beyond the general escrow for the specific risk periods. DD-driven escrow sizing accounts for identified risks. A deal with environmental Phase II exposure of $2-4M (sized in Part XI) needs an environmental carve-out escrow of at least that size, separately from the general 10-15 percent. A deal with pending patent infringement claim of $8M asserted damages (surfaced in Part X legal DD and stress-tested in Part VI IP audit) needs an IP carve-out escrow. The pattern is that Part VI and Part XI produce the risk numbers; Part XII sizes the escrow that transfers or retains them. #### Chart 3\. Escrow architecture: general + carve-outs General escrow covers unforeseen R&W breaches. Carve-outs cover DD-identified specific risks. Combined size varies by deal profile. Escrow sizing per typical mid-market industrial-tech deal profile. ## Earnout structure Earnouts allow deal pricing on future performance the buyer cannot fully diligence. Common structures: revenue milestone (X percent of deal price if revenue reaches Y in year 2), EBITDA milestone (similar on EBITDA), customer retention milestone (share of price contingent on top-N customers retained through year 2). Earnouts that work have three features: binary trigger (achieved or not, no partial), defined measurement (specific accounting standard for the metric; open interpretation stays weaker), and buyer covenants (buyer commits to operate the target consistent with the plan the earnout is measured against). Earnouts that produce litigation have the opposite: sliding scale triggers, disputed measurement, and buyer freedom to change operating plan in ways that suppress earnout achievement. The Bain / Proterial 2023 take-private is a reference on how a well-structured earnout can bridge buyer/seller price gap; poorly-structured earnouts in adjacent transactions have produced multi-year litigation. Operational adherence to the buyer covenants is harder than the drafting suggests. A buyer that commits to operate the target consistent with the pre-close plan is committing to a plan that will collide with post-close integration priorities, market shifts, and cross-portfolio strategic choices. The commitment to "consistent operation" is usually honoured in the specific investment decisions the covenant explicitly names (product roadmap protected, sales team compensation held, capital allocation to the target ring-fenced) and negotiated around at the edges. The sharpest earnouts name a small number of high-consequence covenants precisely and leave the rest to the general good-faith obligation, across the few mechanics that count; fencing every operating decision would overreach. #### Chart 4\. Earnout structures: what works, what litigates Binary triggers with defined measurement work. Sliding scale with disputed measurement litigates. Buyer covenants around operating plan are the third element. Structural features and their correlation with clean outcomes. ## Earnout case patterns: three sponsor deals and what the structure produced Earnout mechanics are easier to describe in the abstract than to structure in practice. Three recent sponsor deals in the AI infrastructure and adjacent industrial-tech universe illustrate the range of outcomes when the earnout meets year-2 reality. **Pattern A. Vertiv / PurgeRite December 2025.** Vertiv acquired PurgeRite (liquid cooling flush and fill services) for approximately $1.0 billion cash with an earnout of up to $250M tied to 2026 performance metrics: revenue growth against a defined trajectory and retention of the top-5 hyperscaler service contracts through year 2\. Binary triggers on both, defined measurement using pre-agreed accounting standard, and buyer covenants requiring PurgeRite to be operated as a distinct business unit with independent sales targets. Outcome (Dec 2025 close, earnout not yet crystallised): the structure is a template for how tight DD-milestone alignment can hold contingent consideration on services carve-outs. The structure held because the milestones matched what the DD had surfaced (revenue driven by hyperscaler contract flow, retention driven by service quality on named accounts) and the covenants prevented Vertiv from cross-subsidising or reorganising in ways that would suppress achievement. **Pattern B. TDK / Ferric (illustrative structure).** TDK's Ferric relationship runs as a partnership (not a disclosed acquisition) as of mid-2026 (Ferric remains independent per its August 2025 Fe1766 IVR launch). The structure below is illustrative of how a hypothetical technical-milestone earnout could bridge process-node maturity risk. A closing payment plus earnout structured against three technical milestones (production yield thresholds on named process nodes, design-win count against enumerated hyperscaler platforms, and revenue trajectory in years 2-3). Milestones are binary but the measurement window is long (three years) with defined but complex dependency chains (a design win only counts if the platform ships, which depends on hyperscaler roadmap timing that TDK cannot fully control). This illustrative structure stays on the reference shelf pending any actual disclosed acquisition. This structure is a sponsor-friendly precedent for how to bridge a strategic price where the target’s value depends on adoption events beyond either party’s complete control. But it required careful drafting of what "shipment" means and what happens if the hyperscaler delays. **Pattern C. Bain / Proterial 2023.** Bain took Proterial (formerly Hitachi Metals) private in a transaction reported at roughly $7.5B, structured through a tender offer. Public disclosures do not describe a formal contingent-consideration mechanism, so the earnout dimension of this pattern draws on the operational-covenant structure the sponsor negotiated around continued operating philosophy across the multi-line target; a disclosed earnout schedule stays absent. Cross-reference Investment Layer Part II (a separate series on capital allocation) for the broader context on Bain / Proterial as a template for European and Japanese take-privates in industrial materials. The pattern across all three: milestones matched to what DD surfaced, binary triggers with defined measurement, and buyer covenants that constrained the buyer’s freedom to change operating plan in ways that would prejudice earnout achievement. Cross-reference Essay X on legal DD which flagged R&W insurance treatment of earnout disputes (typically excluded from coverage), and Essay XIII on financing which addresses how earnout obligations interact with covenant packages. #### Chart 5\. Earnout case patterns: structure, DD alignment, outcome Vertiv / PurgeRite achieved clean earnout via tight DD-milestone alignment. TDK / Ferric structured against long-horizon technical milestones. Bain / Proterial used contingent consideration to bridge take-private gap. All three had binary triggers and buyer covenants. Structure features on 0-10 scales. Deal specifics from public disclosures; earnout terms illustrative of pattern. ## The AI-infrastructure-specific overlays The generic structure / R&W / escrow / earnout architecture above applies to any mid-market industrial-tech transaction. AI-infrastructure targets bring three overlays that drive a reset in how the standard components get sized and drafted. The overlays are worth naming explicitly because they change the negotiating positions on both sides. **Hyperscaler-contract-retention earnouts.** Where 40-70 percent of target revenue depends on two or three hyperscaler customers, and where those contracts include consent-on-change-of-control clauses, retention through the earnout window is the single most negotiated variable. Sellers push for the earnout to trigger on any contract renewal in the measurement period, on the theory that renewal proves retention. Buyers push for the earnout to trigger on capacity-commitment level maintained through the period, on the theory that a re-signed but reduced contract is a real loss. The tighter earnout drafts distinguish between the contract existing (binary) and the committed spend (level), and pay against the level not the existence. The hyperscaler procurement roadmap is the exogenous variable both sides are betting on, this is what the TDK/Ferric-style technical-milestone earnout does not cover. **R&W underwriting gaps on AI-adjacent IP.** R&W carriers have decades of underwriting history on industrial-IP breaches; they have months of history on foundation-model training-data indemnities, GPU-allocation contract disputes, model-output infringement claims, and hyperscaler master-services-agreement enforceability. What this means in practice: R&W policies on AI-infrastructure targets include explicit exclusions on foundation-model IP and model-output infringement, and the exclusions are drafted broadly enough that the buyer effectively retains those risks even with R&W insurance in place. The DD workstream should size these residual risks explicitly and either build them into escrow carve-outs or price them into the transaction; assuming R&W covers them is the pattern that catches acquirers. **CFIUS / export-control escrow carve-outs for advanced compute targets.** For wide-bandgap, advanced-packaging, and interconnect targets that fall inside CFIUS critical-technology definitions, retrospective national-security review remains a real risk even after clearance. The Wingtech / Newport Wafer Fab pattern (initial clearance, later forced divestiture) has raised acquirer sensitivity to this class of risk. Escrow architecture for these deals should include a specific CFIUS/export-control carve-out sized against the divestiture-risk scenario; sizing only against the closing-condition CFIUS review would understate it. This is a Part XI workstream output that has to land in the Part XII escrow build. ## Three recurring failures Structure chosen for tax without regard to liability transfer. The seller pushes for share structure to optimise capital-gains treatment; the buyer accepts without testing whether the target’s liability profile warrants selective assumption. Post-close, undisclosed liabilities (environmental, tax, employment) surface as buyer responsibility. R&W architecture leaves specific DD findings uncovered. The DD identifies a specific risk (customer CoC, environmental exposure, IP litigation) but the R&W insurance excludes it because insurers do not cover known risks. The escrow does not carve out for it either. Post-close, the identified risk falls on the buyer. Escrow sized against generic benchmarks. The deal escrow is 10 percent of purchase price because that is the market convention. DD-identified risks that would warrant larger escrow stay unreflected. Post-close claims exceed escrow capacity and litigation follows. #### Chart 6\. Three recurring deal-structure failures Tax-driven structure without liability testing is the most common. R&W gaps on identified risks is the second. Generic escrow sizing is the third. Frequency from post-close deal-structure retrospectives. ## A worked example Continue the magnetics acquisition. Deal structure scope: Structure: share deal (US Delaware holding company). Clean liability profile allows share simplicity. Tax step-up not available but seller-side benefit of capital gains treatment supports the structure. Malaysia operating subsidiary transfers with the parent equity. R&W: insurance policy at 10 percent of $400M enterprise value ($40M coverage), 6-year survival on fundamental reps, 3-year on general reps. Cost: $1.6M premium. Insurer requires exclusions on the two identified DD risks (patent infringement claim, Malaysia environmental exposure). Escrow: general escrow of 12 percent ($48M) for 18 months. Carve-outs: environmental Malaysia $5M for 36 months, patent infringement $10M for 24 months, tax indemnity $8M for 3 years. Total escrow: $71M or 17.75 percent of purchase price. Earnout: none. Deal priced on current earnings and clear growth trajectory; DD findings support the price at closing; earnout-bridge requirements stay absent. #### Chart 7\. Worked example deal structure summary, magnetics deal Share structure, $48M general escrow + $23M carve-out escrow, R&W insurance $40M coverage with two exclusions. No earnout because DD supports the price at closing. Structure summary; specific terms subject to negotiation. ## The risk-to-protection matrix The output of WS12 is a specific mapping from each material DD finding to a specific protection mechanism, with the mechanism sized against the exposure and matched to the exposure duration. The matrix is what deal counsel and the deal team use to build the transaction paper. | Risk | Loss magnitude | Timing / duration | Certainty | Preferred protection | Sizing basis | | -------------------------------- | --------------------------- | -------------------------------------- | ----------------------- | ------------------------------------------------------- | --------------------------------------------------------------------------------------------- | | Historical tax exposure | $20M | Up to 3 years (statute of limitations) | Medium | Specific indemnity + escrow | Expected loss + uncertainty premium; escrow duration matches statute | | Top-customer CoC consent | $80M annual revenue | Immediate on close | High | Closing condition | Consent obtained pre-close; walk if not | | IP infringement claim (pending) | $50M potential damages | 2-5 years to resolution | Low (uncertain outcome) | Specific indemnity + escrow + R&W carve-out | Escrow sized to legal-team expected-loss estimate; indemnity uncapped for the specific matter | | Revenue-forecast miss | $100M EV impact | 1-3 years post-close | Medium | Valuation adjustment or earnout on measurable milestone | Priced into entry multiple; earnout on specific customer-contract milestone if used | | Founder / key-person exit | Execution disruption | 6-12 months | Medium | Retention agreement + non-compete + holdback | Retention pool sized to replacement cost; holdback tied to continuity milestones | | Regulatory approval (post-close) | Site operational continuity | Triggered on scale threshold | Low | Covenant + specific indemnity for pre-close conduct | Seller covenant to cooperate; indemnity for pre-close acts | | Broad ordinary R&W breaches | Unknown | 12-24 months typical | Low | R&W insurance policy + escrow for retention | Policy limit calibrated to deal size; escrow to policy retention | The matrix is the single most useful WS12 artifact for the deal team. It converts every DD finding into a specific paper-side mechanic with a specific size and duration, and it is the direct input to deal-counsel drafting. ## Risk stacking: correlated cascades; independent-bullet treatment misses the mechanism Conventional DD reports treat risks as independent bullets. In practice, material risks are often correlated in ways that produce cascading failure modes the individual-risk analysis misses. Illustrative cascade for an AI-infrastructure target: **Founder departs → key hyperscaler relationship weakens → hyperscaler reduces qualified allocation → factory utilisation falls → gross margin compresses → EBITDA falls → debt covenant tightens → refinancing terms worsen.** Each individual risk (founder retention, customer concentration, utilisation, covenant headroom) may be within tolerance. The cascade falls outside those tolerances. The protection mechanism must therefore be tested not against each individual risk in isolation but against the specific correlated failure pattern the DD workstream identifies. A retention agreement that protects against founder departure does not protect against the downstream customer, utilisation, margin and covenant consequences. A separate mechanism is required at each downstream layer, or the cascade must be priced into the valuation as a compound stress case. WS08's compound-downside methodology and WS12's cascade analysis are the same exercise viewed from the two ends of the DD framework. ## The unallocated risk register The most important output of WS12 sits past the list of allocated risks; it is the list of risks that remain unallocated after all other mechanisms have been considered. Every material DD finding should end in one of six explicit allocation categories. 1. **Eliminated.** Resolved before close via Level 1 or Level 2 mechanism. 2. **Allocated.** Assigned to seller via indemnity, escrow, R&W or covenant. 3. **Priced.** Buyer accepts the risk at an adjusted valuation. 4. **Financed.** Debt or equity structure carries the risk (financing tenor, liquidity buffer, covenant headroom). 5. **Insured.** Third-party carrier bears the risk under specific policy terms. 6. **Unallocated.** Buyer owns the risk without offsetting mechanism or explicit price adjustment. **The unallocated register is the deal-killer category.** Where multiple material risks fall into the unallocated category, the deal team has produced diligence that identified the risks and a transaction that did not address them. This is the specific pattern that produces post-close surprises. The unallocated register should be surfaced explicitly to the IC before signing; post-close discovery arrives too late. ## Protection is capital-structure dependent The mechanism choice depends on financing structure, not isolated from it. A $100M escrow reduces seller cash proceeds by $100M, which affects the seller's willingness to accept the structure, the seller's rollover appetite, and the buyer's financing capacity through the equity commitment required. Earnout structures similarly affect seller-side cash timing and buyer-side leverage capacity. Illustrative interactions: - Large escrow reduces seller net proceeds and may require rollover restructuring or additional buyer equity. - Earnout deferred consideration reduces initial funding requirement but increases post-close governance complexity. - Specific indemnity that requires ongoing seller counterparty limits the ability to structure the seller as a wind-down vehicle. - R&W policy premium and retention affect deal economics on both sides. - Financing lender may require specific covenants or MAC protection around identified WS12 risks, feeding directly into WS13. **Transaction protection is part of financing capacity.** Every WS12 mechanism has a WS13 financing consequence, and the two workstreams should be run in coordination; sequential-draft treatment stays weaker. ## The ideal WS12 flow: 11 steps from DD finding to residual buyer exposure 1. **DD identifies risk.** Which upstream workstream surfaced the finding? 2. **Quantify exposure.** Maximum loss, expected loss, tail loss. 3. **Determine timing.** When can the loss crystallise? Over what window? 4. **Determine reversibility.** Can the underlying issue be fixed pre-close or post-close? 5. **Determine ownership.** Which of the four risk buckets applies (seller, buyer, shared, deal-breaker)? 6. **Select mechanism.** Which of the seven protection levels is appropriate? 7. **Size protection.** Nominal amount calibrated to exposure and uncertainty premium. 8. **Match duration.** Protection window covers the reasonable crystallisation timeline. 9. **Test collectability.** Who stands behind the mechanism? What is the effective enforceability? 10. **Calculate residual buyer risk.** Effective protection = amount × coverage × duration × enforceability × collectability. Residual = exposure minus effective protection. 11. **Feed financing.** Which WS12 mechanisms constrain or enable the WS13 financing structure? Every material DD finding runs through all eleven steps. Findings that stop at step 6 (mechanism selection without sizing, duration matching or collectability testing) are the pattern that produces nominal protection that fails to hold under stress. **The WS12 contrarian.** The nominal amount of protection is irrelevant if its duration, collectability or scope does not match the underlying exposure. A $50M indemnity from a wind-down seller with an 18-month survival period against a five-year regulatory exposure falls short of $50M of protection; it stays a headline the IC memo holds without an effective mechanism behind it. WS12 is the workstream where the whole DD framework converts from analytical output into transaction architecture. Every unallocated risk is a risk the buyer has accepted by default, whether the IC memo says so explicitly or not. ## Closing Deal structure is where DD findings translate into the transaction paper. Structure choice, R&W architecture, escrow with DD-informed carve-outs, and earnout mechanics where applicable together determine what the buyer acquires and what protection the buyer retains. Part XIII treats financing structure and closing risk. ### Glossary of terms used CFIUS Committee on Foreign Investment in the United States. US inter-agency review body that screens foreign acquisitions of US businesses for national security implications. CoC Change of Control. Contractual clauses that trigger renegotiation, consent requirement, or termination when a counterparty is acquired or its ownership changes. DD Due Diligence. The workstream discipline of testing a target's claims before committing to a transaction. EBITDA Earnings Before Interest, Tax, Depreciation, and Amortisation. The most commonly referenced operating-earnings metric in M&A pricing. HSR Hart-Scott-Rodino Antitrust Improvements Act. US merger notification threshold and review regime. R&W Representations and Warranties. Seller statements about the target's condition. R&W insurance covers buyer's post-close claims arising from breaches. For the full corpus glossary of acronyms used across all essays, see [adikumar.co/glossary](https://adikumar.co/glossary/). Related topic hubs - [AI Infrastructure M&A](https://adikumar.co/ai-infrastructure-m-a/) - [AI Infrastructure Financing](https://adikumar.co/ai-infrastructure-financing/) - [Browse all topic hubs →](https://adikumar.co/topics/) ### Sources and method R&W insurance market data from Marsh, Aon, WTW published reports 2018-2026\. Structure choice tax treatment per US federal tax code and common practice. Bain / Proterial reference per public deal disclosures. ### Series Due Diligence for the AI Buildout. Part XII (this piece): *Deal structure, reps and warranties, escrow*Written in a personal capacity. Nothing here is investment advice. *Related tool:* the [DD Evidence Matrix](https://adikumar.co/dd-evidence-matrix/) is where every material finding lands as an auditable row with provenance code, decision impact, and post-close observable. The chain from question to post-close monitoring runs through this artefact. Due Diligence for the AI Buildout. The workstream index 1. [Part I. Framework: how DD works for AI infrastructure deals](https://adikumar.co/due-diligence-for-the-ai-buildout/) 2. [Product, technology, and roadmap DD](https://adikumar.co/product-architecture-and-technology-dd/) 3. [Customer, pricing, and competitive DD](https://adikumar.co/customer-pricing-and-competitive-dd/)