Now that we’re back in Japan, we’re pretty much in what I consider as the last phase of our “pre FI” journey. Some days I feel we’re safe with a significant safety margin, others I’ll go back to tools such as cFIREsim to crunch the numbers.
Over the years, I feel my numbers have become more precise. One of the main reasons, I have to admit, is that I want to pull the plug as fast as possible from the corporate job: my initial numbers were overly conservative and had lots of buffer. But, as I’m sure many FI/RE bloggers have experienced too, this would mean working longer, much longer.
At some point my numbers kind of went to the other extreme, making dangerous assumptions on the safe withdrawal rate, the USD/JPY exchange rate, and our family’s expenses.
Among other things, I made the assumption that my side income would be here “forever”, and increase with inflation. I now believe this is a dangerous guess to make: there are lots of factors that could disrupt this business, in particular my own interest for it, which I’ve seen dramatically fading over the past 12 months.
There’s also the growing voice in the financial independence community that the 4% Safe Withdrawal Rate is not so safe after all. There’s a great study on a blog out there (of course I can’t find the link anymore), as well as additional literature, that have now convinced me that 3.5% is probably a better bet for people with a long retirement time such as early retirees. 3.5% is also often referred to as the “perpetual withdrawal rate”. The rate at which independently of the time horizon for retirement, one’s wealth is not depleted in historical data.
Furthermore, this interesting post on the MMM Forum confirms some of my gut feelings that we all need an additional buffer: tools such as cFireSim make assumptions on the “initial state” of one’s retirement journey, that are probably not true for most of us. Namely, CFireSIM assumes you’ll start your ER on a “random” year, when in practice it is much more likely that people will reach their target on a “peak” year for the market. Which, statistically, increases the risk that the following years will be negative or “stagnating”, increasing the chances of failures. I’m not explaining it right, but the post does. It’s an important read, as the OP calculated that this decreases chances of success from 98% to 40%. The author concludes that one way to get back to a 90+ percent of success probability is by being above one’s target for 5 years in a row (yes, that means working 5 more years than planned. Ugh.)
Then again, there’s still the contagious optimism of people like MMM or The Mad Fientist, telling us that nah, 4% is fine. I’ve decided to take a bit of both, and go with 3.75%. More specifically, 3.5% after fees.
Right now, cFireSIM and my gut feeling tend to agree: I need one more year, even after including things I didn’t want to rely on before, such as social security.
My new target is much more precise than before. But the more precise I get, the less wiggle room I feel I have. It seems I have exchanged one kind of stress for another.