Gasem says:

Great discussion. One thing I don’t understand is the assumption you will have an alternative in a severe downturn. Do you think side hustles expand or fail in the downturn? Do you think big box stores open or close? Do you think in competition for that Home Depot job your experience as a early retired software engineer 5 years out from employability is a better or worse when competing against a carpenter with 20 years experience for the same scarce big box job?

I’m also not impressed with how personal disaster was glossed over. 1/3 will get some kind of cancer. Of that 1, 20% will die and 80% will survive. Of the survivors 40% of the ones with bad disease go bankrupt in in 4 years. The incidence of Alzheimer’s at 65 is 1/10. This grows to 1/3 by age 85. Once diagnosed the length of living with the disease as it progresses is 12 years, each year requiring expanding need for support. A woman who lives to 65 has a 30% chance of living to 90 and a 1.5% chance of living to 100. A woman at 85 therefore has a pretty strong chance IF she acquires the disease of needing a decade of increasingly expensive memory generally uncovered by Medicare. People blow off the risk but the actual risk is likely if not inevitable. If there are 2 of you, you actually have need to cover the eventuality for both people at some level. That “wildly over saved” scenario is not so wild if viewed from this risk perspective.

The analysis of risk is not binary. It is not a risk aversion vs risk tolerant pick your level of risk out of thin air calculation. In fact failure rates can be calculated and the inputs adjusted to an optimized minimal risk. I found Karsten’s comment about real rate of return vs retirement length illuminating. In my own analysis I found the same thing, but I use a slightly different perspective. I start at the end and then work back ward reverse engineering the portfolio. If to live 50 years you will “need” 5M bucks how much do you have to start with to accomplish your goal and how likely is that to happen and what economic conditions will be necessary to get you there. If you start out with enough you will have no leverage on your future. In his example a return of 1.5% over 30 years was sufficient, so that’s your leverage on the future. In the 50 year case nearly 3x that amount of leverage was necessary to succeed. I think this is the real control variable to primarily focus on, not WR. A proper understanding of leverage on the future I think is a better measure of the probability of success, and some planned contingency to cover the risk of major disaster. The other problem I see in the average FIRE calculation is ignoring the vagaries of tax law. Tax law causes a dynamic load on your retirement. Just the death of a spouse can trigger an increase in taxes by 2 tax brackets. If your plan was tight to begin with 2 tax brackets is not trivial. Thanks for putting this topic together. These guys clearly know what they are talking about. It deals somewhat better with how things will be in retirement vs how things are in accumulation, though in my opinion it still misses some of the moving parts.