This post uses hypothetical scenarios for illustrative purposes only. It does not describe any actual client, transaction, or representation, and is not legal advice.
Imagine a family-owned distribution company worth eight figures, held fifty-fifty by two siblings. Years ago their accountant told them every closely held business needs a buy-sell agreement, so they signed one: if either dies, the company redeems the deceased’s shares, and to fund the redemption the company bought life insurance on each of them. Everyone slept better. The structure sat in a drawer for a decade, doing what estate planning documents do, which is quietly go stale. Then one sibling dies, the company collects the policy, redeems the shares from the estate — and the IRS shows up with a position nobody in the drawer-stuffing era considered: the insurance payout made the company itself more valuable, so the estate owes tax on a bigger number than anyone modeled.
That is not a hypothetical risk anymore. It is the holding of Connelly v. United States, and two years on, most closely held companies with insurance-funded redemption agreements still have not reread their documents. If the company might also be sold someday — and every company might be sold someday — the buy-sell deserves a second look now, because the same document creates problems on both the estate side and the deal side.
Connelly held that the redemption obligation doesn’t cancel the insurance asset
The facts were as clean as tax cases get. Two brothers owned Crown C Supply, a building-supply company. Their buy-sell gave the survivor an option to buy the deceased brother’s shares; if he declined, the company had to redeem them. The company held life insurance on each brother to fund that obligation. When Michael Connelly died, his brother declined the option, the company collected the insurance and redeemed Michael’s shares for about $3 million, and the estate valued the company as if the insurance and the redemption obligation washed each other out. The IRS disagreed, and in June 2024 a unanimous Supreme Court sided with the IRS: life insurance proceeds payable to a corporation are a corporate asset that increases the company’s fair market value for estate tax purposes, and the company’s contractual obligation to redeem shares with those proceeds is not a liability that offsets them. Justice Thomas’s reasoning in the slip opinion is almost arithmetic: a redemption at fair market value doesn’t impoverish a corporation in a way that hurts the remaining shareholders — it shrinks the share count along with the cash — so the obligation to redeem isn’t a value-reducing liability in the ordinary sense. The estate’s tax bill went up by nearly a million dollars, and every insurance-funded redemption agreement in the country inherited the math.
The mechanical lesson planners have drawn since is structural. In a redemption agreement, the company owns the policies, so the proceeds inflate the very valuation the estate is taxed on. In a cross-purchase agreement, the owners hold policies on each other personally, the proceeds never touch the company’s balance sheet, and the Connelly problem largely evaporates — at the cost of more policies to manage as the owner count grows, which is why insurance-dedicated LLC arrangements have become the workaround of choice in multi-owner companies. None of this is exotic. It is a known problem with known fixes, sitting unfixed in thousands of file drawers.
The buy-sell is also a deal document, and diligence now reads it that way
Here is the M&A angle, which gets less attention than the estate planning angle and deserves more. When a third-party buyer looks at a closely held target, the buy-sell agreement is one of the first organizational documents in the data room, and it can complicate a sale in three distinct ways.
First, transfer restrictions. Buy-sells routinely contain rights of first refusal, consent requirements, and prohibited-transfer clauses that were drafted to keep shares inside the family, and they do not politely step aside because everyone now wants to sell. A sale process that gets to signing before anyone waives the buy-sell’s transfer machinery has a title problem dressed up as a formality — the target’s own shareholders’ agreement can make the shares undeliverable. The fix is procedural and cheap if done early: joinders or waivers from every party to the buy-sell, executed alongside the letter of intent, not scrambled during closing week.
Second, stale prices. Older agreements often carry fixed-price certificates or formula clauses — book value, a multiple frozen in a different decade — that no longer resemble reality. In an estate setting, the tax rules have long refused to let below-market buy-sell prices set the estate tax value unless the arrangement meets bona fide business-arrangement tests; Connelly itself flagged, without deciding, how far those protections reach. In a sale setting, a stale formula does something worse: it hands a disgruntled owner an argument about what their shares are entitled to receive in the deal waterfall, and intra-owner valuation fights are the deals’ quietest killers. If the certificate says one number and the buyer is paying five times that number, counsel should assume someone will notice.
Third, the policies themselves. If the target owns life insurance on its shareholders, the purchase agreement has to do something with it: distribute the policies to the insureds before closing, surrender them for cash value, or transfer them — and transfers of existing policies carry their own tax traps, because moving a policy for consideration can compromise the tax-free character of the eventual death benefit unless an exception applies. Buyers rarely want to inherit key-person policies on departing founders; sellers rarely think about the policies until the flow-of-funds memo forces the issue. It belongs on the term sheet, not the closing call. In S corporation targets these mechanics also brush up against the shareholder-level cash planning I’ve covered in the posts on S corp distribution covenants and the 338(h)(10) election, since policy distributions and deemed-asset-sale elections both change what actually lands in the founders’ accounts.
The exemption is high, which is exactly why people stop paying attention
The reflexive objection is that the federal estate exemption — roughly $15 million per person under current law — makes all of this academic for most owners. Two answers. The exemption is a political number with a history of moving, and a business worth comfortably under the exemption today can appreciate past it by the time anyone dies; buy-sells are signed years before they matter. And the deal-side problems — transfer restrictions, stale prices, orphaned policies — bite at every company size, exemption or no exemption. For Florida owners the state side is mercifully simple, since Florida imposes no estate or income tax of its own, which is part of the calculus in the post on changing domicile before a sale; but the federal rules in Connelly apply the same in Naples as in New Jersey.
The homework is short. Pull the buy-sell. Check who owns the policies and who the beneficiary is — company means redemption structure means Connelly exposure. Check the price mechanism against reality. Check what the agreement says about third-party sales, and whether its restrictions would slow one down. In most cases the whole review is an afternoon, the fix is an amendment or a restructured insurance arrangement, and the alternative is litigating valuation with the IRS or with each other at the worst possible moment. Documents age. This one has a Supreme Court case explaining exactly how.
If your company’s buy-sell agreement predates Connelly, or a sale process has surfaced one nobody has read in years, 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.
