Building Eternal Legacies

The Native American practice of looking forward seven generations sounds noble but is essentially vague. Still, it seeds the idea for an investment math exercise that could provide not just for seven generations, but the generations beyond.

Can testaments be created that outrun heirs?

(Fund Manager Conference Call)

At the other end of the colonialism spectrum, noble houses that have stayed insanely wealthy across generations do exist. But setting aside those born into extreme good luck, can the portfolios of ordinary investors survive generations of inheritors?

Like Rabbits

The crudest model of population growth is the Fibonacci sequence. Although it does not take into account fertility windows, migration, and death, it apparently does not do a bad job of estimating actual population growth.

Derived from the Fibonacci sequence is the idea that the ratio of any generation's population to the generation before approaches the Golden Ratio (φ = 1.618...) as time passes.

Let's say you start with 100 family members. By the next round of births deaths and marriages you'll have roughly 162.

Breed's Need Speed

The compound annual growth rate (CAGR) of this model depends on the time generations take to procreate. Rabbits which mature and breed in a month would overwhelm any investment.

By comparison a human family which generates heirs every 18 years would have a CAGR of 2.71% p.a. Any investment with a higher CAGR would grow faster.

Thankfully, that is at first glance a low bar to hurdle.

Let's assume your returns match the S&P500.
Some historical returns:

• S&P500 real (inflation adjusted) returns 1871-2019: 2.313%
• Total real returns (dividends reinvested: 6.822%)

Dividends still need to be reinvested to grow the portfolio fast enough. However, a portfolio that simply grows without paying out is useless. A portfolio requiring continued contributions is worse, with progeny likely dismantling it for draining their incomes. The seven generation decision would not last two unless immediate benefits could be enjoyed. How much benefits?

S&P500 Dividend Yield 1900-2015 (Source: Robert Shiller)

• Average: 4.13%
• Minimum: 1.16%

So 1.16% of the portfolio every year is a 'very safe' withdrawal rate. It's almost certain that you would not be touching capital and that you would have surplus dividends to reinvest.

However, for that 1.16% to make payouts equivalent to the 2019 US median male wage of USD33,517, the portfolio would need USD2.89m in assets for each beneficiary.

Family Planning

Ideally, you would amass that USD2.89m in one lifetime, but in all likelihood building the portfolio will be a multi-generational project.

You can't really outperform the market and grow it faster than above without dialing up the risk and thereby jeopardising what should be a steadily growing income stream.

Instead of controlling the growth of the portfolio, can we control what it needs to beat?

Working backwards, a family that grows in size at a CAGR of 2.313% p.a. - matching the inflation-adjusted S&P500 - would produce offspring roughly every 21 years; still within the bounds of human fertility.

Assuming that the portfolio is placed in the hands of a third party, implementing this could be as basic as written guidance to custodians on the required age gap between eligible beneficiaries and their parents, or as strict as hard clauses in trust deeds.

Going further, requiring a 25 year age gap between parents and children would roughly grow the beneficiary population at 1.92% p.a., allowing the portfolio to distribute all dividends or diversify into safer (but slower) investments like treasuries or term deposits, while still keeping pace with their numbers.

So the good news to chiefs and elders out there is that the Eternal Legacy is possible, if you keep in mind ...

... what you can't control:

• The market
• Your family/tribe (😉amirite?)

... and what you can control:

• Asset allocation (likely overweight diversified equities)
• Who gets paid. 

Setting Limits

Continuous expansion and the Fibonacci sequence is only one model of population growth. Others tend to be defined along the lines of growth rates dependent upon finite carrying capacity. Rather than zoom up, they fluctuate or tail off. I'm not so concerned about using a model that assumes unfettered expansion because I'm not certain that carrying capacity and its return on investment is immutable in the face of productivity improvements. Even so, I acknowledge that my response to primitive notions of sustainability is itself also primitive.