Ray Kurzweil is a brilliant inventor and author and is currently Chief Scientist at Google.  He is known for many things, but perhaps most notably for the idea of the “Singularity.”  This is a time in the fairly near future, according to him anyway, when processing power of computers will have developed to the point that machines are smarter than people.  When that happens, Ray wants to become a machine.  Separately, he notes that life expectancy is growing rapidly.  It is currently increasing at about one year every three or four years.  Ray expects medical science to soon reach the point that life expectancy increases by one year every year.  At that point, Ray wants to become a machine and live forever.  To accomplish this, all he needs to do, according to his calculations, is to survive another 30 years or so.

Toward that end, Ray takes about 150 pills a day and conducts a never ending stream of blood tests designed to provide an early warning system on any approaching disease.  In the event he spots a problematic condition, he is prepared to drop all other projects and focus on developing a cure.  Once he accomplishes that, he expects to return to his other projects.

While I don’t have any insights about the singularity, or living forever, the idea does raise an interesting investment problem.  How should you invest and how much can you spend if you plan to live forever?

Obviously, you would have serious longevity risk (that is, the risk of outliving your assets).  It would seem that there might be great opportunities in life annuities offered by insurance companies, at least until the insurance companies catch on to the fact that no one is dying.  On the other hand, there is the problem of potential failure of insurance companies that don’t catch on.  Maybe life annuities wouldn’t be the solution.

In a sense, you would be like a University endowment fund, which also has a perpetual horizon.  Most endowment funds spend annually about 5% of the fund to supporting operating expenses of the University.  That large payout is feasible in part because endowment funds enjoy tax free status.  As a taxpayer, Ray would be well-advised to spend less than five percent of assets each year.  Assuming a 40% tax rate on investment income, maybe it is reasonable that Ray target spending 3% of his wealth each year.  But I don’t think so.

The university has a couple additional advantages over Ray.  First, there is the likelihood of a continuing stream of financial gifts by alumni and others.  This will tend to keep the endowment fund growing even if annual spending exceeds portfolio earnings.  Second, a steady five percent of the current value of the portfolio means that spending will decline should the market value of the fund decline.  In principle, Ray could adjust in this fashion as well, but I think for most people it is quite painful to reduce your spending level once you get used to it.

This is why I think my preferred spending rule – the “Gamma Percent Ratchet Rule” – is appropriate.  This rule says you can spend each year gamma percent of the “high water mark” (the maximum level) of wealth.  It is called a ratchet because spending should never decline, it moves up when wealth reaches new highs, but it does not move down when wealth falls (unless wealth falls to zero).  This rule will not work if you set gamma too high.  If the spending percentage is set too high (greater than the average after-tax rate of return over an extended period) then the rule will eventually fail dramatically because each year you dip into a declining wealth, and therefore you will eventually run wealth down to zero.  The largest “safe” gamma depends on the outlook for investment returns (and, of course, on your investment strategy).  Based on what I consider to be reasonably conservative return assumptions, and based on extensive simulations, I find the largest “safe” gamma is 1.0.  To state the conclusion more positively, you (or your heirs) can safely spend 1% of the high water mark of your wealth each year forever, without incurring substantial chance of wealth dissipation.

But, maybe the problem is moot.  As a machine, perhaps Ray’s day-to-day spending needs will be modest, just power and periodic tune-ups.  Also, Ray may be able to generate income forever as an inventor and thinker.  With low spending needs and an ongoing income stream, it is likely that financial insolvency will not be a major problem for Ray.