The 5 clauses that kill LMM deals
TL;DR — Most LMM deals don't die at the LOI stage. They die four to eight weeks later, when a clause everyone glossed over in the term sheet turns into an unbridgeable gap in the definitive agreement. These are the five that do the most damage, and what to negotiate on each.
Every deal that's blown up on us — or nearly blown up — has died on the same handful of clauses. It's not the enterprise value that kills deals in the lower middle market. It's not even usually the multiple. It's the structural terms that get waved through as "we'll figure it out in the APA" and then don't.
Here are the five I watch hardest.
1. The working capital peg
What it is. A target level of net working capital the seller must deliver at close, with a dollar-for-dollar adjustment (up or down) at true-up if the actual delivered NWC differs from the peg. The idea is to prevent the seller from stripping cash and running down receivables in the weeks before close.
Why it kills deals. The peg is almost always negotiated late, and it's almost always negotiated badly. Sellers who haven't run a deal before massively overestimate what their normalized working capital looks like — they anchor on year-end, which is often the trough, and get shocked when your QoE provider comes back with a trailing-twelve-month average that's $600K higher. Now you're asking them to deliver $600K of working capital they weren't planning on, which effectively cuts the price. That conversation, six weeks in, is the one that breaks deals.
The other failure mode is asymmetric — the peg is set at a point estimate with no collar, so tiny movements in receivables timing between signing and close turn into meaningful price adjustments neither side priced in.
What to negotiate. Push for the peg conversation in the LOI, not the APA. Anchor on trailing-twelve-month average NWC, not year-end. If the QoE isn't done yet, at minimum agree on the methodology in the LOI ("TTM average of monthly NWC, excluding [items]") so the number itself can be plugged in later without re-opening the principle. And on any deal with lumpy receivables or seasonal inventory, negotiate a collar — a $50K–100K band inside which no adjustment occurs — to absorb noise.
2. Indemnification basket and cap
What it is. The seller's indemnity obligation for breaches of reps and warranties is bounded on both ends. The basket is a deductible — the buyer eats losses up to a threshold before indemnity kicks in. The cap is the ceiling on total indemnity exposure, usually expressed as a percentage of purchase price.
Why it kills deals. Two failure modes at opposite ends. Too broad — a 25% cap with a low basket and non-tipping structure — and the seller sees an unacceptable overhang on their proceeds; they walked into the deal thinking they were getting X and now they're really getting X minus 25% at risk for eighteen months. Too narrow — a 1% cap with a high tipping basket — and the buyer's counsel will kill the deal on your behalf because you have no real indemnification.
Where I've seen this actually blow up: a mid-diligence pivot where the buyer's counsel discovers something ugly (a wage-and-hour exposure, an ambiguous IP assignment from a former CTO) and suddenly wants to expand the basket structure to cover it specifically. If the LOI language was vague, this is now a re-trade. If the LOI language was specific, you have a fighting chance to keep the deal on the rails.
What to negotiate. Get specific in the LOI. "General indemnification basket of 0.75% of purchase price, tipping, with cap of 10% of purchase price for a survival period of 18 months. Fundamental reps and tax reps uncapped and surviving through the statute of limitations." That's a sentence. Put it in the LOI. It removes 90% of the fights that would otherwise erupt at APA time. If you're not sure what to anchor on for your deal size and segment, our AI clause library (see pricing) will pull comparable structures from anonymized closed deals in your segment.
3. Non-compete radius and duration
What it is. The seller's covenant not to compete with the acquired business for a specified time in a specified geography. Usually paired with a non-solicit on customers and employees.
Why it kills deals. Two reasons this one is uglier than it used to be. First, non-compete enforceability is now heavily state-dependent — California and North Dakota basically don't enforce them, Colorado and Illinois heavily restrict them, and the FTC's attempted federal ban is still being litigated through the courts as of this writing, so the landscape is genuinely uncertain. Second, sellers in the LMM segment often intend to stay operationally involved for a transition period, and the non-compete language interacts with the transition-services agreement in ways nobody thinks through until the seller's counsel starts asking sharp questions.
Where this actually kills deals: seller signs the LOI, retains counsel, counsel reads the non-compete, seller realizes they've just agreed to a five-year, national-scope, all-related-services non-compete that would prevent them from consulting to anyone in the industry ever again. Seller balks. You re-open. Deal loses momentum.
What to negotiate. Right-size the geography to the actual operational footprint of the business, plus a reasonable buffer. Right-size the duration to the earnout period plus 12 months — going longer than that on standard operating businesses is usually enforcement risk without much upside. And carve out narrow permitted activities (consulting to specifically-named non-competitors, board seats in adjacent industries) upfront rather than fighting them out in the APA. Given the FTC uncertainty, we've been layering in jurisdictional fallback language — if the primary non-compete is unenforceable in the seller's state, a narrower fallback kicks in automatically.
4. Earnout structure
What it is. A portion of purchase price contingent on post-close performance against defined metrics, usually paid out over one to three years.
Why it kills deals. The earnout is where sellers overestimate their own optimism and buyers underestimate their own optimism, and both sides agree on a headline earnout number without agreeing on how the metric is defined. That definitional gap is where 100% of earnout disputes live.
An EBITDA-based earnout without clear definitions of add-backs, capex allocation, transaction cost carveouts, corporate overhead allocations from the buyer, and treatment of new customer acquisition costs — that's not a deal term. That's a lawsuit waiting for a trigger. I've seen a $2M earnout dispute that consumed 18 months of management time on both sides and ended up settling for a fraction of what either side originally claimed, because the underlying definitions were ambiguous.
What to negotiate. The number matters less than the definition. In the LOI, name the metric ("Adjusted EBITDA, defined per Schedule X of this LOI") and attach the definition schedule to the LOI itself. Specify the add-back methodology. Specify whether buyer's post-close changes (new headcount, corporate allocations, capex acceleration) are excluded from the calculation. If the seller is staying operationally, define who has authority over decisions that would affect the metric — no earnout dispute has ever been resolved cleanly when the seller believed the buyer was intentionally suppressing earnout-year performance.
Better yet, replace an EBITDA earnout with a revenue-based or gross-margin-based one where the definitions are less contestable. Sellers usually resist this because they think EBITDA is a bigger number for them; often the opposite is true after add-back fights.
5. R&W survival period
What it is. How long after close the buyer can bring an indemnification claim for a breach of the reps and warranties in the APA. Typically 12–24 months for general reps, longer for fundamental and tax reps.
Why it kills deals. In the LMM segment, the survival period is the risk allocation on unknown post-close issues. Representations & warranties insurance (RWI) is broadly available but often uneconomic on deals below $30M enterprise value — the premium and retention structure doesn't pencil, so most LMM deals skip the policy and rely on the seller's indemnity backed by an escrow.
If your survival period is too short (say, 12 months), a breach that surfaces in month 14 leaves the buyer with no recourse. If it's too long (say, 36 months), the seller has an unacceptable overhang on their proceeds. The gap between "acceptable to buyer" and "acceptable to seller" here is where deals stall out at the APA drafting stage.
On the smaller LMM deals where RWI isn't feasible, the survival period isn't a legal detail. It's the deal's actual risk-transfer mechanism. Treat it that way.
What to negotiate. For most sub-$30M deals without RWI, we target 18 months on general reps, 36 months on tax reps, and open-ended (statute of limitations) on fundamental reps. Escrow sized to 5–10% of purchase price, released in tranches — half at 12 months, half at end of survival period. Get all of this in the LOI. If the seller pushes back hard on the survival period, that's often a signal they know something you don't — worth an extra diligence workstream before you accept a shorter tail.
The pattern
You'll notice the theme: get specific in the LOI. Every one of these clauses can be handled cleanly if the principles are agreed upfront, and every one of them turns into a deal-killer if it's punted to the APA.
The LOI is not just a term sheet. It's your one chance to align on the structural terms while both sides still have deal momentum. Once you're in APA drafting and lawyers are billing and momentum has stalled, every re-opened clause is a chance for the deal to die.
That's the discipline. That's also why we built an AI clause library into Arendel — every LOI you draft on the platform pulls comparable clause structures from anonymized closed deals in your segment, so you're not reinventing the peg language or the earnout definition schedule from scratch on every deal. If you want to see it, the pricing page has the details.
Five clauses. Get them right in the LOI. Watch how many fewer deals die in the last mile.