Incorporator The Journal of U.S. Formation
Start Free Quiz
Editorial 8 MIN READ

Writing an operating agreement that actually holds up

The eight clauses that do the work, and the four failure modes that keep litigators fed

The Editors
Staff, Incorporator.org
8 min read Updated
Contents 5 sections
  1. The statute gives you the pen
  2. The eight clauses that do real work
  3. Four failure modes that keep repeating
  4. What "holds up" actually means
  5. Sources

n LLC operating agreement is a private contract, not a corporate charter, and Delaware's statute treats it that way. Under 6 Del. C. § 18-101(9) the "limited liability company agreement" means any agreement of the members, written, oral, or implied, concerning the affairs of the LLC and the conduct of its business. Everything the statute permits you to contract around, the operating agreement governs; everything you leave out falls back to the statutory default, which is almost never what you would have drafted if you had thought about it.

That is the mental model. The document does not describe the LLC; it constructs it. Members who skip the drafting exercise are not "going with the simple version." They are consenting, sight unseen, to a default the legislature wrote for strangers.

The statute gives you the pen

Delaware's enabling language is unusually generous to drafters. 6 Del. C. § 18-1101(b) declares it the policy of the chapter to give maximum effect to the principle of freedom of contract and to the enforceability of LLC agreements. Subsection (c) goes further: an LLC agreement may expand, restrict, or eliminate duties, including fiduciary duties, provided the agreement does not eliminate the implied contractual covenant of good faith and fair dealing. The waiver must be express, and Delaware courts read it strictly. Generic boilerplate disclaiming "all duties" is weaker than a clause that names the duty of loyalty and the duty of care and states, in terms, that they are eliminated.

If you want the benefit of a fiduciary-duty waiver, you write it plainly and you write it once. If you want to keep the common-law duties and simply clarify how they apply to a specific transaction, you say that instead. What you cannot do is leave the question ambiguous and hope for the best. Ambiguity in a fiduciary-duty clause, under Delaware law, resolves toward the duty surviving.

The same philosophy runs through the rest of the chapter. § 18-402 lets you pick member-managed or manager-managed, or invent a hybrid. § 18-702 lets you restrict transfers, including by absolute prohibition. § 18-801 lets you set your own dissolution events in addition to the statutory ones. The statute is a buffet. An operating agreement that reads like a form off the internet is usually one that left most of the food on the counter.

The eight clauses that do real work

A sound operating agreement has eight load-bearing sections. The order varies; the content does not.

First, capital contributions and capital accounts. Who put in what, when, valued at what, and how future contributions get called. This is the section founders hate writing because it forces a conversation about unequal commitment. It is also the section a court reads first when members fight over who owns what.

Second, member rights and voting. What decisions require unanimous consent, what requires a majority by units, what the manager can do alone. Delaware's default under § 18-302 is per-capita voting in proportion to contributions unless you say otherwise. If you want one-member-one-vote, or a supermajority for a sale, or a class of nonvoting units, this is where it goes.

Third, management structure. Member-managed is the default under § 18-402. Manager-managed requires an election in the certificate or the agreement. If you are manager-managed, the agreement must describe how managers are appointed, how they are removed, and what authority they hold between meetings. Silence on removal is how a bad manager stays for a decade.

Fourth, distributions. Two broad models: pro rata to units, or a waterfall with tiers (return of capital first, then a preferred return, then a catch-up, then residual profit sharing). The waterfall is standard in anything investor-backed. Pro rata is fine for a genuinely equal partnership. Mixing the two without the drafting discipline to keep the tiers clean is the most common source of distribution disputes we see.

Fifth, tax allocations and the § 704(b) language. IRC § 704(b) and its regulations require that partnership allocations have "substantial economic effect" or otherwise follow the partners' interests in the partnership. In practice this means the agreement needs qualified-income-offset language, a minimum-gain chargeback for nonrecourse deductions under Treas. Reg. § 1.704-2, and capital-account maintenance consistent with Treas. Reg. § 1.704-1(b)(2)(iv). Skipping these clauses does not save you drafting time; it invites the IRS to recharacterize your allocations to match the economics, which is almost never what you want.

Sixth, transfer restrictions and a right of first refusal. 6 Del. C. § 18-702 lets the agreement restrict assignment of membership interests however the members choose. A ROFR with a defined notice period and a valuation mechanic keeps the cap table from becoming a jump ball. Without one, a member's stake can be pledged, sold, or inherited in ways that introduce strangers to your operating table.

Seventh, buy-sell triggers. Death, disability, divorce, personal bankruptcy, termination of employment, involuntary transfer by court order. For each trigger the agreement states whether the interest must be sold, to whom, at what price, and on what terms. In a two-member LLC without buy-sell language, one member's death hands the business to a probate estate, and one member's divorce hands it to a judge dividing marital property.

Eighth, dissolution and winding up. When the LLC dissolves, how the affairs get wound up, who signs the certificate of cancellation under § 18-203, and in what order assets get distributed under § 18-804 (creditors first, then members with respect to unreturned capital, then members in proportion to distributions). You can adjust the intra-member ordering by agreement; you cannot subordinate creditors, and nobody has ever tried and won.

Four failure modes that keep repeating

The first is the California template dropped onto a Delaware LLC. California's RULLCA and Delaware's LLC Act reach different defaults on material questions: fiduciary-duty waiver language, voting by capital versus per capita, the consequences of a member's dissociation, the treatment of a charging order. A clause that works in the California form because it mirrors the California default can be silently wrong, or silently weaker than it needed to be, under Delaware law. Form banks are fine as a starting sketch. They are not fine as an unread final draft.

The second is the two-member LLC with no buy-sell. Fifty-fifty deadlock is not a hypothetical. When two members disagree on whether to sell, whether to take a distribution, or whether to fire an employee, there is no tiebreaker. Delaware's statute offers judicial dissolution under § 18-802 on a showing that it is "not reasonably practicable to carry on the business in conformity with" the agreement, which is an expensive way to solve a problem the agreement should have solved in a paragraph. A buy-sell with a shotgun clause or a neutral appraisal process turns a deadlock into a transaction.

The third is the missing § 704(b) language. Most operating agreements we audit either copy the regulatory language correctly or omit it entirely, and the second group is surprisingly large. If your allocations do not have substantial economic effect and do not demonstrably follow the partners' interests in the partnership, the IRS reallocates. The risk is not theoretical; it shows up on partnership audits under the centralized regime enacted by the Bipartisan Budget Act of 2015, which since 2018 has pushed adjustments down to the entity level by default unless the partnership properly elects out or pushes out to the partners.

The fourth, newer failure mode concerns the Corporate Transparency Act and its beneficial-ownership-information rule. As of 2025, CTA enforcement has been rerouted through rulemaking and litigation more than once; the current posture under FinCEN's March 2025 interim final rule narrows the BOI reporting requirement to entities formed under foreign law and registered to do business in the United States, and exempts most domestic entities from reporting. That is the state of the rule today. It is not a guarantee that the rule will stay there. An operating agreement drafted in 2025 should contain a confidentiality clause that covers BOI disclosures and a member covenant to supply whatever information the company needs to comply with reporting law, current or future. Drafting for the policy as it stands is fine. Drafting as if the policy will not move is not.

What "holds up" actually means

An operating agreement holds up when it survives three tests. It survives a dispute between members: every decision that has been made, or could be made, resolves by reading the document. It survives an IRS partnership audit: the capital-account math works, the allocations have economic effect, the agreement names a partnership representative under IRC § 6223 and sets out that person's authority. It survives an outside event: a death, a divorce, a bankruptcy, a subpoena, a reporting obligation that did not exist when the document was signed.

None of those tests is passed by a three-page template. All of them are passed by a document whose drafters thought, clause by clause, about what could go wrong and wrote the answer. The cost difference between a well-drafted agreement and a form-bank agreement is a few thousand dollars at formation. The cost difference between surviving a dispute and not surviving one is the business.

The rule of thumb: if a clause would be expensive to litigate, write it; if it would be cheap, rely on the statute.

Sources

Keep reading

More from the journal.