Intergenerational Split Dollar: What the Regulations Settle, and Where the Real Risk Lives

A family has a real and permanent insurance need. The policy is sized correctly, the trust is drafted, and the trustee is the right person. Then someone runs the premium, and it turns out to be well beyond what annual exclusion gifting can absorb. The family would rather not spend lifetime exemption on payments that will continue for twenty years or more.

The planning stalls there, and in our experience it stalls there more often than it fails for any technical reason.
Split dollar exists to solve that specific problem. It is worth saying plainly at the outset that the problem is a real one, because this is an area where the technique has sometimes gotten out ahead of the need it was built for. That criticism is fair. The underlying difficulty is still genuine.

This article is about one version of the arrangement in particular. In intergenerational split dollar, the generation paying the premium is not the generation being insured. A grandparent funds coverage on a child’s life, owned by a trust for the grandchildren, and the person paying is expected to die decades before the policy pays out. Nearly everything that follows comes from that single fact. It means the person who funded the arrangement is likely to die holding a right to be paid back that nobody expects to be satisfied for another thirty or forty years, and somebody has to put a number on that right for estate tax purposes at a moment when no one can say when it will actually be paid.

Last month we wrote about split dollar as an executive benefit and a business planning tool. The same regulations cover this arrangement, but almost none of the questions carry over. Nobody is negotiating retention here, and no company is recovering an outlay. What is at stake is whether the arrangement does what the family intended, and what it looks like in the estate of the person who paid for it.

A note on our perspective: Summit Legacy’s expertise is in the insurance side of these arrangements; sizing the coverage to the actual need, policy selection and funding, in-force review, and planning the eventual unwind. The drafting, the trust structure, the valuation work, and the estate tax analysis belong to estate planning counsel, a tax advisor, and an independent appraiser. We write about the insurance and planning implications because that is where we work every day.

How the Arrangement Works

The mechanics are straightforward enough. One family member, or that person’s revocable trust, pays the premium on a policy the irrevocable trust owns. In exchange, the person paying holds a contractual right to be repaid out of the policy, generally the greater of the premiums advanced or the policy’s cash value at the time. Each year the trust receives something of value, namely life insurance protection it is not paying for, and that value is treated as a gift from the person funding the arrangement. Because the gift is measured by the cost of that protection and not by the premium, it is usually a small fraction of what the family would otherwise be giving away.

There is a second way to do this. Rather than the economic benefit structure just described, the family member can lend the premium to the trust under a written note bearing adequate interest, which puts the arrangement in what the regulations call the loan regime. We covered the mechanics of both last month and will not repeat them here.

The choice between them matters in this setting for a reason that has little to do with income tax. The two versions leave very different assets in the estate of the person who funded the arrangement. A promissory note is a familiar thing that estate practitioners and the Service both know how to value. A right to be repaid the greater of two numbers that are both still moving, at a date nobody can predict, is a considerably harder problem, and it is the problem the Tax Court has spent the last decade working through.

A Set of Facts

Some version of the following comes across our desk regularly. The figures are illustrative.

A widow in her early eighties holds an estate of roughly $60 million, most of it a manufacturing business her three children will eventually inherit. The children are in their fifties, and there are six grandchildren. The family’s concern is not really her estate, which planning already in place can largely absorb. It is the generation after hers. When each child dies, that child’s share of the business faces tax again, and the company has never carried the cash to fund a buyout.

Everyone agrees on the answer: permanent insurance on the three children’s lives, owned by a trust for the grandchildren. The combined premium runs about $900,000 a year.

Annual exclusion gifting does not come close. Even assuming the trust is structured so that gifts to it qualify for the annual exclusion, the 2026 exclusion of $19,000 per recipient produces six available exclusions, or $114,000, against a $900,000 premium. The remainder would consume lifetime exemption, or constitute a taxable gift, every year for as long as the policies are funded.

So she advances the premium instead, under a split dollar arrangement. The trust reports the value of the insurance protection as a gift each year, a figure that typically runs in the tens of thousands against a premium in the hundreds. She holds a right to be repaid out of each policy when that child dies.

What the Regulations Settle, and What They Don’t

The split dollar regulations are narrower than most people assume, and the two halves of them are narrow in different ways.

The economic benefit regulation announces its own scope. By its terms it provides rules for taxing a split dollar arrangement for income tax purposes, for gift tax purposes, and for employment tax purposes. That is the entire list, and the federal estate tax is not on it.

The loan regime regulation reads differently. It provides that a payment made under a split dollar arrangement is treated as a loan for Federal tax purposes, without naming which ones. That broader language gives the loan regime a more straightforward estate tax posture, and it is one reason practitioners may prefer it for family arrangements.

What neither regulation does is answer the questions that actually decide these cases. Neither tells you whether the person who funded the arrangement kept a string attached to the policy that pulls it back into her estate. Neither tells you whether the restrictions written into the split dollar agreement will be respected when the estate values what it holds. And in the economic benefit version, neither tells you what the right to be repaid is worth on the date of death.

Those questions run through ordinary estate tax law. Section 2036 reaches property a decedent gave away but continued to enjoy or control. Section 2038 reaches property the decedent could still alter or revoke. Section 2703 permits the Service to disregard certain restrictions in valuing property. General provisions, applied to a particular set of documents and a particular file.

Which is why every serious dispute in this area has been litigated there, and not under the split dollar regulations at all. It is also the clearest reason we know to treat the choice between the two regimes as an estate planning decision. In practice it usually gets made on the income and gift tax math, and the consequences run a good deal further than that.

The Right to Be Repaid, and What It’s Worth

Return to the widow. Suppose she dies eight years in, having advanced roughly $7 million in premiums. Her right to be repaid is still outstanding, and it will not be satisfied until her children die, which is likely another thirty years away.

That right is an asset of her estate, and it has to be valued. Nobody would pay $7 million today for the right to collect $7 million three decades from now, so the reported value is lower, sometimes dramatically lower. The discount is a real position with real support behind it. It is also one of the most heavily scrutinized numbers in the arrangement, and it has to be established rather than asserted. In practice that means an independent valuation, built on assumptions about how long the arrangement will actually run that the file can support.

A second consequence gets very little attention. Say the estate reports that right at $2 million against the $7 million advanced. Whoever inherits it generally takes it with a $2 million basis. When the $7 million eventually comes back, the $5 million difference represents taxable income to the recipient. How that income is characterized depends on how the arrangement was documented and is a question for counsel, but the shape of it does not change. A valuation discount that saves estate tax now can create an income tax bill later. Both numbers belong in the analysis, and in our experience only one of them usually shows up.

What the Tax Court Has Looked At

The Tax Court has examined these arrangements closely three times, and the pattern across them is more instructive than any single outcome.

In Cahill, decided in 2018, the court refused to rule for the estate at summary judgment. It held that Sections 2036, 2038 and 2703 could all apply, because the decedent, acting together with the trust, had the power to end the arrangement and reach the policies’ cash surrender value. The court never got to the merits; the case settled, with the estate giving up the valuation position it had taken.

Morrissette, decided in 2021, went well on one issue and badly on another. The arrangement survived the Section 2036 argument. But the court rejected the value the estate placed on the right to be repaid, in part because an amendment provided that the note would pass to the insurance trust after three years, which shortened the arrangement’s expected life and severely undercut the discount. The court also sustained a forty percent gross valuation misstatement penalty, and refused to let reliance on the appraisal excuse it.

In Levine, decided in 2022, the taxpayer won. An independent investment committee, not the decedent, held the power to terminate. The decedent could not surrender the policies. The trust owned them throughout, so the only thing in her estate was the right to be repaid.

Each of these turned on its own facts and its own paperwork, which is the lesson and not a hedge. All three arrangements were built by competent people. What separated them was who could end the arrangement, and what the file showed about why it existed.

Where We Come Out

Last month we wrote that in an executive arrangement, the exit is the design. In a family arrangement, control is.

Before anything is signed, we want to know who can terminate the arrangement acting alone. If the answer includes the person funding it, whether by herself or jointly with the trust, we want to hear the argument that she has retained nothing, and we want it to be a good one. We want to know how long the arrangement is expected to run, and what the annual gift looks like in its last ten years rather than its first ten, because the cost of insurance protection climbs steeply with age and an arrangement that is inexpensive on a sixty-year-old is not inexpensive on an eighty-year-old. And we want to know whether there is anything in the file, a memo, an amendment, an email, suggesting the parties expected to unwind it early. That is not housekeeping. It is what cost the taxpayer most of the discount in Morrissette, and it points at the test we actually apply: not what the parties would say if asked, but what a reader fifteen years from now would conclude from the documents in front of them.

This structure earns its complexity when a family has a genuine and permanent insurance need, when the premium is large enough that annual exclusion gifting cannot realistically carry it, when the person funding it has the cash flow to sustain it for its expected life and intends to, and when an independent trustee actually controls the arrangement rather than nominally holding the title.

We recommend against it more often than we recommend it. Sometimes the premium is modest enough that annual exclusion gifts or a straightforward loan to the trust will do the job without the machinery, and that is frequently the honest answer. Sometimes the need is not permanent, and an arrangement built to run for decades is the wrong container for it. Sometimes nobody has agreed to maintain it, and these arrangements need attention every year for as long as they exist. And sometimes the family’s real objective is the valuation discount rather than the insurance, which is the one case we decline outright. An arrangement whose main justification is the tax result stands on weaker ground than one built around a need, and the record eventually shows which it was.

That last point carries more weight now than it did two years ago. The 2025 legislation set the federal estate, gift, and generation-skipping transfer tax exemption at $15 million per person, indexed for inflation, with no sunset, and the deadline that drove so much planning activity through 2024 and 2025 came off the calendar. What did not go away are the reasons families do this in the first place: liquidity, business succession, keeping an estate intact through a transition, and, as of 2026, the estate taxes imposed by twelve states and the District of Columbia, many at thresholds well below the federal one.

What changed is the ratio of substance to strategy. An arrangement justified mainly by an expiring exemption stands on thinner ground today, while one built around an insurance need the family can describe without mentioning tax at all stands on considerably firmer ground under the cases. That is not a worse environment for this planning. It is a more honest one.

Summit Legacy works with businesses and tax-exempt organizations on split dollar design, in-force review of existing arrangements, and exit planning, often in collaboration with specialist partners on the most technically complexSummit Legacy works with families and their advisors on the insurance side of these arrangements: sizing coverage to the actual need, policy selection and funding, in-force review of arrangements already in place, and planning the eventual unwind. We work alongside estate planning counsel and tax advisors; we do not replace them. If you have an arrangement that has not been reviewed recently, or a situation you think may call for one, we would welcome the conversation.applications. If you have an arrangement in place that has not been reviewed recently, or a situation you think may call for one, we would welcome the conversation.

IMPORTANT DISCLOSURES

Split-Dollar Insurance is not an insurance policy; it is a method of paying for insurance coverage. A split-dollar plan is an arrangement between two parties that involves “splitting” the premium payments, cash values, ownership of the policy, and death benefits. These arrangements are subject to Split Dollar Final Regulations that apply for purposes of federal income, employment and gift taxes. Regulations provide that the tax treatment of split-dollar life insurance arrangements will be determined under one of two sets of rules, depending on who owns the policy.

The information provided in this article is for general informational purposes only and does not constitute legal, tax, financial, or investment advice, a solicitation, or a specific recommendation. The facts and figures used in the illustration are hypothetical, are provided solely to demonstrate the concepts discussed, and do not reflect any actual client, arrangement, or policy; actual results will vary. Split-dollar life insurance arrangements are governed by Treasury Regulations Sections 1.61-22 and 1.7872-15 and are subject to complex tax rules that vary based on arrangement design, ownership structure, and individual circumstances. The estate tax treatment of a split-dollar arrangement is determined under general estate tax provisions, including IRC Sections 2036, 2038 and 2703, and depends on the specific facts, documents, and retained rights involved. A discounted valuation of an interest in a split-dollar arrangement should be supported by appropriate valuation analysis and documentation; inadequate substantiation of a claimed value may result in penalties, including the gross valuation misstatement penalty under IRC Section 6662(h). Court decisions referenced are described in general terms for informational purposes only; outcomes depend on the specific facts, documentation, and record of each case and are not predictive of any other result. Estate, gift, and generation-skipping transfer tax exemption and annual exclusion amounts are current as of 2026, are subject to legislative change and annual inflation adjustment, and state estate and inheritance taxes may apply, in some states at considerably lower thresholds. Life insurance products involve costs, risks, and material considerations that vary by individual circumstance, and results will vary. Consult a qualified estate planning attorney, tax advisor, and financial professional before implementing any split-dollar or estate planning strategy. Securities offered through Integrity Alliance, LLC, Member SIPC. Integrity Wealth is a marketing name for Integrity Alliance, LLC. Summit Legacy is not affiliated with Integrity Wealth.

Our corporate calling of helping others, along with our life insurance and estate planning specialties, intersects with our client’s desire for ongoing financial security and protection.

Gary Bottoms, CLU, CHFC

Chief Executive Officer

770-425-9989

Like this article?

Share on Facebook
Share on Twitter
Share on Linkedin