The magic of compound interest?

Because thanks to an eccentric New York lawyer in the 1930s, this college [Hartwick] in a corner of the Catskills inherited a thousand-year trust that would not mature until the year 2936: a gift whose accumulated compound interest, the New York Times reported in 1961, “could ultimately shatter the nation’s financial structure.”

The real rate of compound interest may or may not in the long run exceed the rate of economic growth, so the size of the trust could be enormous in the 29th century but still small or mid-sized relative to the size of the economy.

The problem becomes especially acute when existential or catastrophic risk is significant.  Long-term savers then earn a higher risk premium and the compounding is more intense.  It is possible that the rate of return on these investments — if the world survives — is quite high relative to the rate of economic growth.  In effect, existential risk means the long-term savers will rule the world, if there is a world to rule of course.

The article is here — interesting throughout — and for the pointer I thank Umung Varma.

Addendum: Robin Hanson offers comment.


Small correction: 2936 would be the 30th century, not the 29th.

Given that there are no examples of financial systems that have survived a thousand years, I think the odd of this trust maturing are small. Even if the overall structure of the system survives and the trust doesn't succumb to mismanagement or theft-by-insiders, I think we can assume that someone in power would eventually tax or confiscate the trust once it became large enough.

300 years is typically max for any empire..

I guess you weren't aware of the 999 year long lease that was popular with European back before 1700s. (For various reason, sometimes it just doesn't make much sense to sell a property, as it will incurs taxes that neither parties are willing to pay). When it got to America, we're too impatient to wait for that long, and thus was born the 99 year lease. This practice have faded as real estate law professions and judges have been treating these 'lease' as sales.

Those 999 year leases are a perfect illustration of bbartlog's point.

I'm calling the recession of 2935.

It is also a thousand years for some clever financial person to find a way to siphon off the money in this trust. After all, funds available centuries after everyone involved are dead may not immediately reveal that they have been hollowed out.

After all, no investment vehicle lasts a thousand years (even cash would be eaten up by inflation) so some degree of active management seems inevitable.

I remember both Bill Gates and Warren Buffet saying that they wanted to give most of their money away before they died because multi-generational trusts are eventually consumed by ineffective and self-serving management.

Is everyone an idiot? They really think that some small sum left to compound for enough years would destroy civilization? That doesn't pass the smell test.

Agreed. Just who is obligated to pay the interest? I don't recall the US offering any 1000 year bonds, for example.

The easy idea here is that the money eventually vanishes to lawsuits, bad agency, government churn, black swan events. Even the Catholic church probably won't survive 1000 years as a recognizable institution.

The harder idea is that there no idea you can articulate that is recognizable or useful after 1000 years. In 1011AD, you might be able to do this (feed the poor, build big buildings?) Want can you possibly ask for in 2011 that will be interesting in 3011?

"Given that there are no examples of financial systems that have survived a thousand years,"

I volunteer a welfare organisation that survived for 500 years:
The "Fuggerei".

It is supported by a charitable trust that was founded by Jakob Fugger 1523 with the purpose of offering affordable housing for the poor.
It still exists today and still offers some small but nice appartments at an annual (!) rent of 0,88 Euro (plus 3 prayers a day for the soul of Jakob Fugger).

Oh, by the way: the Fugger trust has invested most of his money in forests, which turned out to be a sound, low profit investment, long term investment.

Clever (and interesting, I knew of Jakob Fugger but did not realize that he had a bequest that had survived the centuries). It helps that the mission is not to grow boundlessly, but just to provide an ongoing service; also investing in forests is genius in this context, since armies can't carry them off and even a rapacious manager can't render them entirely valueless by exploitation. Of course, it's still lucky that property rights have held for five hundred years. The Fuggerei would not have outlived Mao or Stalin.

Would the monies survive another 1000 years without danger of civil forfeitures etc? There's a hundred different reasons the rulers can use to justify appropriating your private property under exigent circumstances.

I imagine the endowment to be more of a contingency fund bequeathed to the future governments for posterity: "seize these funds and use them when things get so dire that you can justify seizing them"

I'm just going to put this here:

My understanding is that it couldn't destroy the economy because at some point, let's say in 2754, the account would have a bajillion dollars, but you'd still need to invest all of it in *something* to get interest on all of it for that year. More likely, as it grows to a certain size, it will be too large to be fully invested. So the rate of return will end up going down.

wellington burt had a 92 yr generation skipping trust (21 years after the death of his last living relative at the time of his death) that was worth ~100 million when it was opened up earlier this year. would have been interesting if he would have had it as a 1000 yr trust.


(The easy idea here is that the money eventually vanishes to lawsuits, bad agency, government churn, black swan events. Even the Catholic church probably won’t survive 1000 years as a recognizable institution.)

I think you mean the Catholic church will not survive another 1000 years. As an ex-Catholic I can assure you that it has already passed its 1000th birthday. Depending on how you count it, the Catholic Church is somewhere between 1800 and 2000 years old. I happen to agree that the Church will decline and not survive ANOTHER 1000 years. Heck, Islam, Buddhism, Hinduism, Taoism and Shinto are all well over 1000 years old. I have no idea which is older, but Hinduism, Taoism and Shinto all pre-date Catholicism.


I mean that the Catholic church of 1000 years ago would not recognize the current Catholic church as reflecting its goals and ideals.

Realistically, today, it's an organization at odds with much of science, perceived to be a haven for homosexuals (fine) and pedophiles (not so fine.) Major growth is in the third world, losing traction in the developed nations. Claims an infallible leader, still does some good works at the basic food, housing, and education level.

Pretty thin gruel, all in all.

Could you point to some science the Catholic church is at odds with?

Will trusts in England had a maximum life of 80 years until a recent change to 125 years. There's a nice, brief summary of the state of affairs in different jurisdictions in section 1.16 of

That Scottish document also contains a good history of the Thellusson case: 2.8-.

Beware of statements along the lines of "In the United Kingdom, dispositions of property subject to the rule before 14 July 1964 remain subject to the rule.[3] The Perpetuities and Accumulations Act 1964 provides for the effect of the rule of interests created thereafter." Most law is the UK is not uniform: you must expect the law to be different in Scotland, in Northern Ireland, and in England & Wales.

I have a similar plan to create "The Company Which Will Own The World". First buy some property - doesn't matter whether it's farmland, commercial, or residential property. Rent out the property (to a farmer, business, or individual). Use some of the rental income to maintain the building; and save the surplus income in short-term treasuries. When the cash pile is big enough, buy another piece of land. Repeat ad infinitum. Eventually the company will own every scrap of land in the world. Best of all, unlike shares or bonds or other financial instruments, land can never go bankrupt or default.

Unless someone else is doing that too, then eventually each will own just under half the world, and they will together bid up the price of the remaining one property to the point that neither can afford it.

I don't like partial equilibrium analysis.


This is exactly the issue I had in mind when I asked if the average human lifespan is related to the average rate of return on investments. If everyone could expect to live for, say, 10,000 years, wouldn't everyone eventually grow rich by just building up principle for the first few centuries? Or would the rates of return be naturally pushed downward in response to this? If the human lifespan was only 30 years, would returns be correspondingly higher?

Didn't Jean Luc Piccard say there was no money in the 24th century?

Sound's like this "trust" is just a Ponzi Scheme. Let's see them borrow against its value.

Even if there is money, it will probably take a $1 Googolplex dollars to go to a movie, and that's not even counting popcorn!

What about the egg management fees?

Sounds like those kooks that tried to cash in their collectible German Empire Bunds from pre-WWI.

Clever, but I think the fallacy is the assumption that the money supply will remain static. Even if the bank's obligation to the trust was payable in gold coin and all gold coins in the world eventually ended up in the vault then silver would be worth that much more. And after silver, copper. And after copper, tin. And of course with fiat money, somebody just makes another credit entry as the scarcity of funds becomes acute. Of course, practically speaking, nobody is going to honor a multi-generational obligation once the foregone opportunities rise to a certain value.

This sort of thinking actually reminds me of Walt Disney peddling the Scrooge McDuck meme: if we were on a gold standard, then greedy savers would hoard all the money in the world so they could dump it in their swimming pools and paddle around in it. Orphans would starve in the streets, desperately hawking newspapers for which nobody has ANY money to pay. At that point, we all lie down and starve to death and Scrooge McDuck inherits the planet. It's theoretically possible, the same way it's theoretically possible for Warren Buffett to acquire every bit of food in the world and charge people ten thousand gazillion dollars for a loaf of bread.

This is what keeps those sweaty, neck-bearded, bitcoin miners going... :)

The donor had the right idea by leaving money to a college. Universities are the country's oldest institutions, the large private ones virtually never fail, and they excel at long-term investment. Where the donor erred was in selecting a small college that was interested in spending the money. My own wealthy alma mater, like many, plows almost all of its endowment income right back into the endowment. So if I were to donate, virtually all of my donation would be allowed to accumulate income for decades if not centuries. I do not expect that situation to last a thousand years -- at some I expect expect governments to come after university endowments the way Henry VIII went after the monasteries -- but a wealthy university is the best place for money to compound over time.

That makes me wonder. Could you actually construct a portfolio that grows consistently and with low risk faster than NGDP? My intuition tells me no. If you could everyone would do it at which point, NGDP would start growing faster.

Wow! Just wow! All that talk about instability of financial systems is just obscuring the main issue.

For the sake of argument let assume a financial system which will stay stable for a million years. Should be enough. So how big can a thousand year compound interest fund grow? Pretty big, that's for sure, but... interest is a price, a price for loans. The bigger the fund grows the bigger the supply of loans, and lower the price/interest. There will be a time when no one will be willing to borrow at 5% per year. Then there will be a time when the fund will be happy to find someone to borrow at 0.005% per year. Where it would be a good or bad development is a rather different question.

Of course, if you start with the (absurd) assumption that the rate of interest is some sort fundamental constant, please don't act surprised when you end up with absurd conclusions.

Whether the accumulated interest "shatters the financial structure" depends completely on how well the money in the trust is invested. If it's invested poorly and earn less than the rate of inflation, then the only thing that gets shattered is the fund's real value. If it invests superbly and significantly surpasses the rate of inflation for 1000 years (a highly unlikely scenario), then indeed, you're talking about a nice chunk of change. If it makes investments earning a typical rate of return (like buying farm land, renting it out and then investing the rental income in gold -- which would probably be among the safer 1000 year investment strategies you could devise), nothing earth shattering is going to happen.

Well, the UK issued infinite life consols in the 19th century, although I think that they eventually paid them all off, or most of them anyway, in definitely finite time sometime last century.

There was a rather curious sci-fi short story published back in the late 1950s by Mack Reynolds that I think was called "Compound Interest." It focused on a board of directors of a trust that had been in existence since 1200 or thereabouts, having been initially started in Florence or Pisa or some such place. The deal was that on New Year's Day of each new century (well, to be precise, of the last day of the century, such as 1300), someone would appear at a particular location and time and tell the board what to invest in during the next century. This person would also tell the board where and when to meet a century later. However, in 1900, the person told them to meet in some location in mid-town Manhatten on New Year's Day, 1960, and this board was preparing for this meeting and discussing the matter. In particular over time the boards had become suspicious of what was going on, as these investments were always accurately in what would be the major rising events and locations of each succeeding century. In 1900, the board had secretly photographed the individual, and Wells's The Time Machine had been published, so they were hoping at the 1960 meeting that this photograph could be used to see if it was the same individual each century, perhaps using a time machine.

Indeed, when the person arrived, that was it. He had invented a time machine and used it to go back and in effect order history. When the board said "why?" (aside from making money), he gave some answer about, "Well, wasnàt it worth it to have the Renaissance and the Industrial Revolution and so on?

Blessed are the meek...

Comments for this post are closed