A family's first direct deal
Going direct means giving up the fund's screening, diversification and operating bench all at once. What replaces each of them, and in what order.
Going direct is usually described as cutting out the fees. What you are actually cutting out is three things the fund was quietly providing: a screening machine that rejected hundreds of deals you never saw, diversification across a portfolio, and a bench of people who had done the first year after a purchase before. Each of those has to be replaced, and the fees were the cheapest of the three.
What the fund was actually doing for the money
The case for direct ownership is strong and I have made it elsewhere. Longer horizon, no forced exit at year seven, no layer of carry, and a proposition a founder-seller often prefers. None of that is wrong. It just describes the benefits without pricing the services being given up.
A sponsor's team looks at several hundred opportunities a year. By the time one reaches an investment committee it has already survived a filter that killed everything else, and the killing is the valuable part. A family going direct sees perhaps a dozen situations, most of them brought by people who are paid when a transaction happens. The sample is smaller and it is selected against you.
Diversification is the second loss and the more obvious one. A fund holds ten or twenty positions, so the two that fail are absorbed. A family that has put a meaningful share of its assets into one operating company has taken a position with no absorption at all, in an asset class where roughly a quarter of tracked acquisitions lose money.
The third loss is the one that actually decides outcomes and gets discussed least. Sponsors keep operating partners: people who have run the first hundred days somewhere before, who know what an integration looks like when it is going badly, and who can be dropped into a company that is drifting. A family buying its first business has a board seat, a lawyer, and an accountant, none of whom can do that job.
What has to replace each one
The three replacements are different in kind, and they are not equally hard.
| What you gave up | What replaces it | How hard |
|---|---|---|
| Screening across hundreds of deals | A written buy box, applied before anybody gets attached, and a deliberate habit of looking at targets you have no intention of buying | Easy to describe, hard to sustain |
| Diversification across a portfolio | Position sizing: cap the first deal at a share of assets you could lose entirely without changing how the family lives | Easy, if agreed before the deal appears |
| An operating bench | One experienced operator, hired or partnered with, before the first purchase rather than after | Genuinely hard, and the one people skip |
The first is a discipline problem rather than a resource problem. You cannot manufacture hundreds of deals to reject, but you can write down what you will buy before anything is in front of you, and you can look at ten businesses you will not buy in order to calibrate the one you might. Reading a target you have already decided against is the cheapest training available.
The second is arithmetic and it should be settled in the governance document rather than in the deal. A first direct position that could take a third of the family's assets with it is not a diversified portfolio with an interesting addition, it is a concentrated bet with a portfolio attached.
Doing that arithmetic properly produces a result that sounds wrong the first time. Taking the tracked total-loss rate of ten percent per acquisition and holding it constant, here is what different position sizes actually mean:
| Position size | One total loss costs | Positions to be fully deployed | Chance of at least one total loss across them |
|---|---|---|---|
| 33% | a third of everything | 3 | 27% |
| 20% | a fifth | 5 | 41% |
| 10% | a tenth | 10 | 65% |
| 5% | a twentieth | 20 | 88% |
The last column is the counterintuitive one. The properly diversified family is almost certain to suffer a total loss, at 88 percent. The concentrated one probably will not, at 27. Read quickly, that makes concentration look safer, and it is the exact reading that ruins people. The diversified family will lose a twentieth and carry on. The concentrated one has a roughly one-in-four chance of an event it does not recover from. A high probability of a survivable loss beats a low probability of a fatal one, every time, and the instinct runs the other way.
The trap underneath it
Which sets up the problem a family going direct cannot fully solve, and it is worth naming because most write-ups skip it. To match a twenty-position fund's absorption you would cap each deal at five percent of assets. For a family with a hundred million that is a five-million-dollar company. For a family with twenty million it is a million-dollar one, which is a job rather than an investment.
And small deals are not proportionally less work. That is the whole basis of Constellation's tiered hurdle rates: a small acquisition consumes almost as much management attention as a large one, which is why they demand a higher return from it. A family trying to build twenty positions is signing up for twenty integrations, twenty sets of accounts and twenty managing directors, with a bench of roughly one person.
The operating bench is the one you cannot buy from a firm, cannot borrow from an advisor, and cannot do without. It is also the one that never appears in the fee comparison.
The specific ways a first one goes wrong
The failures are patterned rather than random, and they mostly happen before completion.
The deal arrives through a relationship
A friend of the family, a private banker, an advisor with a fee attached. It is warmly introduced, it comes with implicit endorsement, and the social cost of rejecting it is high in a way that a sponsor's pipeline deal never is. That warmth is not evidence.
The committee is the family
Sixty percent of family offices run an investment committee. In a first-deal situation it is frequently three people who have the same information, the same advisor, and a shared reluctance to be the one who says no. That is not a committee, it is a group.
Diligence is outsourced entirely
Lawyers and accountants confirm that the numbers are the numbers and the contracts are the contracts. Neither of them will tell you whether the business runs without its owner, whether the customer relationships are personal, or whether the second qualified person exists. Those are operating questions and no professional firm is engaged to answer them.
Nobody has decided who runs it on Monday
The question gets deferred because the seller has agreed to stay for a transition, and a transition period is not a plan. It ends, usually earlier than agreed, and the answer is then chosen under pressure from whoever is available.
The price reflects the good version
Without a portfolio to average across, the temptation is to underwrite the case where it goes well, because the case where it does not is intolerable rather than merely unprofitable. Concentration should make you more conservative on price and frequently does the opposite.
Outsourced diligence is worth expanding on, because it is the most common of these and the most fixable. Legal and financial diligence answer whether what you were told is true. They do not answer whether the business will still work once the person who has been holding it together leaves. That second question is the one that determines the outcome, it is answered by sitting in the business rather than by reading about it, and it is nobody's engagement letter.
The sequence I would actually follow
Order matters more than any individual step, and the first two happen before a target exists.
- Agree what the money is for and what share of it can go into one illiquid position, in writing, before any deal exists
- Write the buy box: sector, size, what must be true, and the conditions that end a conversation outright
- Find the operator. One person who has run something like this, engaged, incentivised, and identified before you commit to anything
- Look at ten businesses you have no intention of buying, and write down why each one fails
- Only then engage with a live target, and run the operating diligence yourself alongside the professional kind
- Cap the first cheque below what you can comfortably afford, on the assumption that the second one will be better informed
Finding the operator is the step that gets deferred, and it should not be. Hiring an operator before you own anything feels premature and expensive, and it is the difference between a family that can buy companies and one that can buy a company. It also changes what you are able to look at: a target that is too messy for a family with no operator is a good target for a family with one, and that is where the price advantage lives.
On the fourth item, ten rejections sounds like busywork and is the highest-return use of time in the whole process. It builds the comparison set you would otherwise not have, it makes the buy box concrete instead of theoretical, and it establishes with everyone involved that no is a normal outcome, which is what makes the eventual yes mean something.
What direct ownership is genuinely good for
Having spent most of this on the risks, the case remains strong, and it is worth naming precisely what the advantage is rather than leaving it as a general preference.
The edge is not access and it is not fees. It is that a family with no fund life can underwrite a business on what it produces over twenty years, accept an improvement that takes three years to show up, and tell a seller truthfully that the company will keep its name and its people. A sponsor competing for the same asset cannot say any of that, and for the kind of owner who spent thirty years building something, it is frequently worth more than the last few percent on price.
That advantage is real and it is conditional. It only exists if the family genuinely holds, genuinely has somebody who can run the thing, and has decided in advance what happens when two family members disagree about it. Those conditions are the subject of the governance almost nobody builds, the structure and cost around them is in what a family office actually costs, and the screen itself is good company, bad ops. If you want the checks that end a conversation before any of this starts, they are on the buy page.
The short version
- Going direct gives up three things, not one: a screening machine, diversification, and an operating bench. The fees were the cheapest of them.
- A fund rejects hundreds of deals you never see. A family sees a dozen, mostly from people paid when a transaction happens.
- Replace screening with a written buy box and ten deliberate rejections. Replace diversification with a position cap agreed before any deal exists.
- At a 10 percent total-loss rate, twenty positions of 5 percent each make a total loss 88 percent likely and survivable. Three positions of a third make it 27 percent likely and terminal. Prefer the high chance of a small loss.
- Diversification and deal size pull against each other: matching a 20-position fund means twenty integrations with a bench of one. Below a certain scale you are choosing concentration, so size the first deal to survive losing it.
- Hire the operator before the first purchase. It is the hard one, the one everybody defers, and the one that changes which targets you can even consider.
- Legal and financial diligence confirm what you were told. Whether the business runs without its owner is an operating question that nobody is engaged to answer.
Questions I get on this
What do family offices get wrong on their first direct investment?
How much should a first direct deal be?
Do you need an operating partner to buy a business directly?
Investment committee prevalence from the UBS Global Family Office Report 2026. Loss rates on small-company acquisitions from the 2026 Stanford Search Fund Study.