October 1, 2026Copy link

Earning Permanence: When to Have a Permanent Capital Structure

Mary Graif McDevitt was a secretary in Binghamton, New York, in the early 1900s. Her boss, A. Ward Ford, sat on the board of a small company called the Computing Tabulating Recording Company. The board hired a salesman named Tom Watson to run it, and Watson renamed it International Business Machines. Mary borrowed $125 and bought shares (borrowing to buy your boss’s stock usually ends badly). She kept reinvesting.

Her son Robert inherited the shares. He graduated from Georgetown in 1940 and ran a funeral home in Binghamton. He never sold. He died in 2008, at 90, as IBM’s largest individual shareholder, with more than $250 million in his personal account. He left $75 million of it to Georgetown, the largest gift in the school’s history at the time.

Nothing forced him to sell, so he didn’t.

Warren Buffett, Mark Leonard, and Henry Singleton are among the investing heroes we all know, yet mimicking their permanent capital vehicles without understanding why they worked is a dangerous game. A few permanent holding companies will join the rarefied air of Constellation, TransDigm, Danaher, Lifco, and Berkshire. Most won’t. You’ve probably met the founder of one, or heard one on a podcast. What does “building the next Berkshire” mean when there’s so much more to Buffett’s genius than the legal structure he used? Copying Berkshire’s visible legal structure is easier than replicating Buffett’s investment edge. By edge, I mean something specific a firm can do that others can’t or won’t. Constellation tracks tens of thousands of tiny vertical-software companies and stays in touch with their owners for years, which nobody can replicate in short order. It also has a culture that builds patience and discipline into the investing process. Danaher has the wherewithal to carefully implement the Danaher Business System. The right structure naturally aligns with an edge and multiplies it.

A typical fund with a ten-year life must return capital when the contractual and external clock runs out. Continuation vehicles can stretch out the timeline, but eventually the fund must dissolve to fulfill its fiduciary and legal obligations to return what is left to investors. A permanent vehicle can remove the legal clock, but does little for the other temptations an investor might face: boredom after a long stretch with no deal, new growth areas of the market you are unprepared to attack, and buyers willing to overpay for your assets.

Berkshire Hathaway is the classic permanent vehicle. It buys and sells companies and stocks when it wishes. But Berkshire itself is about as permanent as a vehicle gets. It’s a C Corp, controlled culturally and through Buffett’s voting stake. After the initial capital went in, Berkshire kept what it earned and built on it. It hasn’t paid a dividend since 1967. The capital is trapped, and that’s a positive when you have the greatest investor at the helm. Berkshire also had an amazing insurance business. Its float was cheap money to invest, and few personal accounts ever get that. And Berkshire could move cash without selling anything. It paid $25 million for See’s Candies in 1972. By 2007, See’s had earned $1.35 billion pretax and needed only $32 million of it back. Buffett sent the rest to other businesses.

Family offices are permanent for the same reason a personal account is. A family office exists to serve the family. You can trade what’s inside a personal account, but you can’t change whose account it is.

§The three reasons people give

§1. Lifestyle

Lifestyle reasons come down to one of two things:

• Ego. You want to say you have a permanent holdco, even if it doesn’t matter. “We never sell” does well at cocktail parties, but don’t expect the high to last.
• Belief that it leads to the life you want. Usually you want something else, like fewer reporting requirements or more time to think, but there are more direct ways to get there. Atypical fund structures are on the table when you work with investors who share a similar vision. Be careful, though. Sometimes this is ego posing as humility.

§2. Marketing

You may want to raise money for the vehicle itself. Or you’re in an auction, and “permanent capital” is a strong selling point. Sellers care about who owns their company, so a buyer who promises to hold for the long term can have a differentiated bid. After close, though, it’s critical the “permanent” strategy aligns with the subsequent operating, capital allocation, and cultural approaches the company takes.

§3. Duration: the edges that need decades

Every sale leaks somewhere. You pay tax on the gain, the cash sits around while you look for the next deal, and the next deal is rarely as good as the one you just sold. Hold something compounding at 15% for 20 years and sell once in the middle, and you give up about a fifth of what you’d have ended with, assuming the next deal is just as good.¹ By year 10, most businesses have already had their re-rating, and growth has slowed, which creates the temptation to sell. The asset you hold can slow to 12.5% a year and still tie the seller who got impatient. As Munger said, “The first rule of compounding is to never interrupt it unnecessarily.”

Structural undervaluation. If an asset is undervalued for reasons that won’t correct within any normal holding period, you can’t sell into a re-rating. Think of Japanese companies that sat below book value for decades because of cross-holdings, or a conglomerate the market won’t price at the sum of its parts. You make your return by holding these to run-off and clipping cash along the way. When your own stock carries a large and durable discount, the best option can be large buybacks. Henry Singleton did this at Teledyne, where he bought back 90% of its shares in a ten-year period. Other investors have bought back stock all the way until they filed for restructuring. A permanent cart full of lemons only gets more sour with time.

Compounding advantages. Some assets grow more valuable the longer you hold them. This can look like the “Munger Trade,” where you put your ego aside so a generational steward can build their life’s work while you sidecar. It’s like holding Amazon from over $100 in late 1999 to under $6 in 2001 all the way to this decade’s highs, or wiring Buffett money in 1956.

Reputation. Firms with the partners’ names on the front door, passed down from generation to generation, doing everything from pre-seed investments to restructurings, have an advantage that’s hard to replicate with a revolving door of capital. Brown Brothers Harriman and some regional family offices fit this bill. BBH traces its partnership to 1818. When a family sells the business, BBH can get the first call because it banked the grandfather.

Patience. If you can wait a decade and then concentrate on a couple of opportunities, a permanent vehicle may fit. It could mean doing nothing for five years while your friends close a deal every quarter. Berkshire holds $365 billion in cash, and has avoided the temptation to spend it for the sake of activity. The rare creatures who can avoid action for years at a time retrofit nap rooms into their offices and know they’ll lose at sourcing as a volume sport.

A permanent vehicle isn’t a substitute for judgment on the asset. Plenty of families held Kodak forever too, and it didn’t end well. The McDevitts won because they picked IBM and nothing ever made them sell. They didn’t need a vehicle because their only job was to sit and wait. An operator who makes businesses better and recycles the cash has more jobs, and the right vehicle helps with every one of them.

Permanent structures should not be preconceived. They are the destination after discovering a personal investment edge, and live as an efficient way to exploit and amplify what you are excellent at.

Thank you to Abi, Harrison, and Jeff for contributing and reading drafts.