"Low for long" interest rates and retirement planning
Loose monetary policy after the global financial crisis has resulted in low interest rates that are also expected to remain low for a long time. A prudent person beginning a career today must therefore assume that over his or her life time, the average global real interest rate would probably be negative. By taking some risk, the return could be increased by harvesting the equity risk premium. Accounting for taxation and asset management costs, it would still be reasonable to assume that the realized post-tax weighted average rate of return would be close to zero. (A prudent person would not plan for the expected outcome but for an outcome in say the 10-25% lower tail.)
Rising life expectancy would imply that a person must plan for a post retirement life span roughly equal to the working life. For example, a person might work for 35 years from the age of 25 to 60, and then live for another 35 years till the age of 95. (Again, life expectancy does not have to be 95 because a prudent person would plan for the 10-25% tail outcome).
Some expenses may be lower post retirement, but medical and related expenses would be much higher (and these are rising much faster than inflation). The simplest assumption would be that average monthly expenses post retirement would be roughly equal to that before retirement.
The above assumptions lead to a very simple savings rule: you must save 50% of your income during your working life. The logic is straightforward:
The assumption that the realized post-tax weighted average rate of return would be close to zero means that the discount rate is zero. In the absence of discounting, one can simply add incomes and expenses over one’s entire life span (time value of money – which is probably half of modern finance – becomes superfluous).
The assumption that post retirement life span is roughly equal to the working life together with the assumption that average monthly expenses post retirement would be roughly equal to that before retirement means that total expenses during retirement are roughly equal to that during the working life.
I assume that total lifetime income equals total lifetime expenses (no net bequests). Therefore working life expenses (being half of lifetime expenses) must be also half of lifetime income.
Assuming that there is no non-investment income post retirement implies a savings rate of 50%.
If in addition, the prudent person is worried about a rising profit share in GDP and a possible secular decline in real wages as artificial intelligence destroys well paying jobs, then the savings rate in the early part of one’s career would have to be even higher.
Casual empiricism suggests that a 50% savings rate is far beyond what anybody plans for today. Several hypotheses suggest themselves:
People are not actually risk averse and are planning for median outcomes and not for the 10-25% tail. This implies a significant minority of people will live out the last decades of their lives in poverty.
People have adaptive expectations and because of high realized rates of return in the 20th century, they are excessively optimistic about rates of return going forward. If this is the case, even the median person could experience poverty in old age.
The worry about “lower for longer”, secular stagnation, and the threat of artificial intelligence is just irrational pessimism, and, in fact, all is well with the world.
Ultimately, the state will be forced to step in to provide support in retirement because in an ageing society, senior citizens have too many votes (and they do come out to vote). Since state support will probably be means-tested (and financed by heavy taxation), it is stupid to save too much.
Climate change will end civilization as we know it long before our savings run out, and so retirement savings are the wrong thing to worry about.
Society will decide that retirement is a luxury that we cannot afford, and we will all be working even with one foot in the grave.
Posted at 4:27 pm IST on Mon, 20 Jan 2020 permanent link
Comments