Supplement to my feasibility game - adding Player E

Loosen up Bro...right?

I've lost count over the years of the number of people that have chided me for my economic conservatism in retirement, from advisors to ex-girlfriends to family to just flat-out strangers (none of them are or were early retirees, btw). The comments can sometimes border on condescending but are probably just a reflection of their incomprehension or ignorance of the risk dynamics. I don't really blame them but I once did fire an advisor from a large national banking empire for being a little too glib and snooty on this kind of thing to my face.  It's my life and my risk and my family and I didn't appreciate a 40 year old advisor with no experience, no math, and no skin in my game looking down his nose and ripping on me in front of a room full of people.  "Live a little, bro." Buh-bye bro.


I got it again the other day on Twitter where one would and should expect this kind of thing.  An advisor chided me for comparing retirement finance and risk taking to Russian roulette, a floor-and-upside metaphor I picked up from Michael Zwecher's book on Retirement Income.  The advisor patted me on the head and reminded me in reassuring tones that retirement risk is not suddenly binary like Russian roulette. We have a chance to see it coming and adjust course.  He even advocates taking up to 30% fail risk (40% for himself) in order to spend more now. Obviously he may have a point worth examining here, but hey! doesn't he recognize the power and glory of RiversHedge? ;-)

But kidding aside, part of me had that visceral reaction again.  It's my risk and my family. I've always resented people that play word games with my money and my skin in my own game. It's like telling me to skate near the open water in an otherwise frozen pond. Better be a better skater than me and better follow me with a rope if you want any credibility.

Playing the Game with a new Player

I got off my high horse and decided to take him at face value.  Let's take his 40% fail rate advice and see what happens in the 1970s game I've been playing this week.  If 40% is good this year, it'll be good next year.  In fact, let's run a constant risk program where each year we spend at a level that produces a 60% sustainability evaluation.  Then along the way we'll check feasibility, utility, portfolio longevity and remaining capital at some threshold age.

The Process 

Mostly the process and assumptions are the same as the prior posts here and here.  We are still playing against the 1970s. The main difference here is that in each year past year-1 we are dynamically evaluating the LPR math we used last time to find, in a given year, the spend rate that is equivalent to a 40% probability of lifetime ruin, taking into account the full term structure of mortality at each age and the full constellation of asset exhaustion possibilities from evaluation age to infinity.  Spending at that level is implemented each year until we either last forever or something bad happens.

The Outcome - Real Spending

Here is the 40% constant fail spend path rendered in real terms. This is superimposed over the other players we saw in the first post.

I was surprised by this.  It looks like player C, just with a little bit of lifestyle borrowed on the credit card of the future self to fund a party in the first nine years.  I thought it would look radically different.  The Twitter Advisor may have a point after all but we'd have to be very careful about what we say.


The Outcome - Feasibility and Sustainability

Not surprisingly for a constant risk program the sustainability risk is constant. Feasibility, in the F/W terms we articulated before, is mostly underwater for the first 15 years but otherwise not too bad.  It hides some nasty lifestyle variability (and a crash at the end), though.


The Outcome - Game Metrics

Here is the table we used before but now updated with Player E.

Player E, against my wishes, plays pretty well.  He has a decent lifetime consumption utility and an ok portfolio longevity.  The Utility difference is purchased by way of the time preferences I think which is something to think about.  Also there is almost no margin for error at the end and a couple lifestyle crashes that need to be considered.  The presence of lifetime income, as I mentioned before, would reorder the results with Player A benefiting the most.

Some Other Thoughts

• This is an ant and grasshopper question: party now and pay later or pay now and relax later? I am constitutionally an ant. This sorting critiera sorts the world just like the question of whether one wants to get to the airport early or just on time.  Me? I get there early.  I guess this is why they invented risk aversion coefficients.
• The advisor touted his 40% thing as a way to spend more but he forgets that there are two sides to the coin.  To play the game in full one has to accept lower spend to maintain the threshold. The only way to spend more at some points is to break the threshold which makes his point of view potentially self-contradicting. I'm sure he neither knows nor cares about this.
• Portfolio longevity gets no big gains from playing it this way but the survival probabilities are so low at the point where it matters that it may not, in fact, matter.
• Me? I like to have a little bigger margin of error near the end. 
• I wonder if he knows the impact of lifetime income.  Its presence would make his high spend and deplete early very viable. 
• I'm not sure how I'd psychologically weather the long stretches of in-feasibility.
• This program is not friendly to early retirees.  If we started this game at 75 it'd be different. If we started at 50, my guess is that this approach would take a hit.  Pretty big gamble. Early retirement is sometimes like a grasshopper skating right along the edge of the open water.