Two co-founders called me late last fall. They had sold their company to a strategic buyer eight months earlier in a deal that closed clean, ten sellers signing, escrow funded, RWI in place. The exit had been everything they wanted. Then one of their early-engineer shareholders — a person who had left the company in 2022 and was a small holder by the time of sale — sent a demand letter back through the seller representative confirming that he had personally signed an indemnification claim against the company’s earlier patent prosecution, and that the buyer’s IP counsel had concluded, after several months of diligence, that one of the foreground patents had been incorrectly claimed by the company without proper assignment from him at the time he left.
The dollar figure on the demand was not catastrophic. It was about $1.4 million against a deal where total seller proceeds had run to the high eight figures. What rattled the two founders was that the demand letter named all ten sellers, that the buyer’s claim cited the indemnification article without distinguishing among them, and that their counsel — me, on the phone — had to confirm that yes, under the merger agreement they had signed, the buyer’s first move was to pursue the full $1.4 million from the escrow, and that the escrow had been funded pro-rata across all ten sellers. The clean nine were about to fund the recovery against the dirty one.
This is the joint-and-several seller indemnification problem, and it is one of the most under-negotiated terms in middle-market M&A. The default in most buyer-form purchase agreements is joint and several seller liability for indemnification claims. The drafting move that flips the default to several-only is well known to deal lawyers, but it is rarely fought for at the term-sheet stage, and it is the kind of provision that founders learn about, post-closing, under demand-letter conditions.
The structural mismatch the default creates
The conceptual problem is straightforward. When ten sellers stand behind a single set of company reps and warranties, and one seller has knowledge that makes a particular rep false, the buyer’s loss is the same regardless of which seller knew. The buyer wants the cleanest possible recovery mechanism — a single pool of liability, jointly and severally backed by all ten sellers, where the buyer picks the deepest pocket and that pocket then sorts out internal contribution rights against the other nine. From the buyer’s perspective, this is rational. It eliminates the buyer’s burden of proving which seller knew what and shifts the allocation problem onto the seller side, where the sellers are presumably in a better position to know.
The problem from the seller side is that the buyer’s rational structure is the structurally wrong allocation. In any multi-seller deal, the sellers do not have symmetric knowledge of the company’s affairs. The founders typically know everything. The CFO knows the financial reps. The senior engineers know the IP and code reps. The investors know nothing operational. The early employees who left years ago know less. Yet under joint and several liability, the buyer can pursue the founder for a rep that turned out to be false because of something the CFO knew and the founder did not — and the founder’s remedy is to chase contribution against the CFO after the buyer is paid.
That chase is theoretical for most sellers. Contribution actions are expensive, the venue is often inconvenient, and the bad-actor seller is frequently the seller least able to fund a judgment. The clean sellers, in practice, eat the loss. The merger agreement’s allocation of liability among sellers — joint and several, several-only, several-only-up-to-pro-rata-share, or some negotiated hybrid — is the term that decides whether the clean sellers can avoid this outcome.
The four regimes you actually negotiate
The buyer’s first draft will almost always present joint and several liability for everything. The seller-side counter typically runs through a spectrum of four regimes, and understanding the spectrum is more useful than memorizing the names.
First, full joint and several. Every seller is liable for every claim, up to the overall cap. The buyer pursues whichever seller is most convenient and the sellers sort it out among themselves. This is the default that founders should resist.
Second, several-only on a pro-rata basis. Each seller is liable for its proportional share of every claim, based on its proceeds from the sale. A 30-percent founder pays 30 percent of any indemnification claim, regardless of which seller’s conduct or knowledge underlies the claim. This is the regime that protects clean sellers from carrying dirty sellers, and it is the seller-side ideal for the general rep set.
Third, the hybrid that has become the modern middle-market norm. Several-only on the general rep set; joint-and-several for specific items where the buyer’s loss is impossible to allocate cleanly among sellers — typically fundamental reps (capitalization, due authority, title to shares), and sometimes pre-closing tax indemnities and a small set of negotiated special indemnities. The seller-specific reps in the agreement — each seller’s own representation about title to its own shares, its own authority, its own no-conflicts — are several-only as to that seller alone, on an uncapped or specially-capped basis, since by definition only that seller knew the underlying facts.
Fourth, and the cleanest from the seller side, several-only across the board, with each seller’s exposure capped at its share of the escrow plus its share of the cap. The buyer’s recourse beyond the escrow is then several-only too. This is the regime John always pushes for when a seller has serious leverage; it is less common in the middle market but increasingly seen in PE auctions where multiple bidders are competing on terms.
What the seller representative is — and is not — for
One of the standard buyer responses to a several-only push is to point at the stockholder representative provision and say something like “the rep handles internal allocation among sellers; that is what the rep is for; the buyer should not have to deal with that.” This argument is partially correct and partially a sleight of hand. The stockholder representative provision does centralize seller communications and decision-making after closing, and the rep does handle the mechanics of distributing escrow releases and processing buyer notices. What the rep does not do is fix the underlying liability allocation. If joint and several liability is on the page, the rep cannot rewrite it. The rep merely administers the regime the agreement created.
Worse, the rep is typically empowered to settle claims on behalf of all sellers, which means a rep facing a joint-and-several agreement and a buyer-aggressive claim can authorize an escrow release that disproportionately disadvantages the clean sellers without their case-by-case consent. The rep’s fiduciary duties to the seller group are real but limited; the rep’s authority under most agreements is broad. The combination of joint-and-several liability plus a broadly-empowered rep is the structure that produced the call from my two founders. They had not paid attention to either term and the combination quietly handed the allocation problem to the rep, who handed it to the escrow agent, who paid the buyer out of pro-rata seller funds.
The drafting moves that actually work
The clean drafting fix is a three-sentence amendment to the indemnification article. First, a sentence stating that the sellers’ indemnification obligations are several and not joint, with each seller’s exposure for any claim capped at that seller’s pro-rata share of the cap and the escrow. Second, a carve-out listing the limited categories where joint and several liability will apply — fundamental reps, perhaps pre-closing taxes, perhaps one or two negotiated specials. Third, a clarifying sentence confirming that seller-specific reps and covenants are several-only as to the breaching seller alone, with that seller’s liability uncapped or capped at a separate amount, but not pulling in the other sellers.
The negotiation that follows is usually narrow. Buyers will press for joint and several on fundamental reps and pre-closing taxes; sellers will accept this for fundamental reps and push back on pre-closing taxes (the company knew its tax positions; the sellers as individuals generally did not). Buyers will press for joint and several on covenants; sellers will distinguish between company covenants (joint and several is workable if at all) and seller-specific covenants (must be several-only). The fight, if there is one, is about pre-closing taxes and the size of the joint-and-several carve-out list.
The fact pattern that should not happen — and that happens regularly because the term sheet didn’t cover it — is the one my two founders walked into. A several-only regime for the general rep set would have routed the $1.4 million claim to the early-engineer seller alone, would have allowed the escrow to be unblocked for the clean sellers, and would have shifted the buyer’s collection problem onto the seller who actually held the underlying knowledge.
The term-sheet move that prevents the closing-week scramble
The reason this term ends up un-negotiated is that the term sheet rarely addresses it. The LOI talks about cap, basket, survival, and escrow size — the dollar terms. The allocation regime is usually treated as a definitive-documents issue and gets pushed to the back end of the negotiation, where it loses to other priorities under time pressure. Founders who care about this should write it into the LOI as a single bullet: “Seller indemnification obligations will be several and not joint, with each seller’s liability allocated pro-rata to deal consideration received, except for fundamental representations and pre-closing taxes, which may be joint and several.”
That one sentence, in the term sheet, anchors the negotiation. It does not eliminate the buyer’s pushback, but it makes the buyer the party trying to walk back something the parties already agreed to, which is a different posture than the buyer’s counsel introducing a joint-and-several first draft in the merger agreement and forcing the seller to fight backwards.
The two founders I started with eventually settled with the buyer for less than the demanded amount, but the escrow was depleted across all ten sellers and the contribution litigation against the early-engineer holder never fully closed the gap. The cost of the joint-and-several default, in their deal, was somewhere between $700,000 and $900,000 of money the nine clean sellers paid that they should not have paid. The drafting fix would have taken three sentences. Cooley’s M&A practice writes regularly on the state of seller-friendly indemnification trends, and their broad directional message is consistent with the one I’d give: the seller-friendly drafting choices in M&A are mostly available to sellers who know to ask before signing the LOI, and mostly unavailable to sellers who learn the term names only after closing.
If you are a multi-founder, multi-seller team approaching an LOI, or an investor in a target where some of the founders are operationally close to the company and some have left, feel free to reach out to my firm manager, Magda, at Magda@montague.law, or fill out our contact form. Mention you read this post.
