If you retired in 1988 and died in 2018 This would have been your S&P 500 SOR. Bold is negative returns. Only 6 years in the last 30 were down years! In a 1988 retirement you didn’t experience a bad SOR till 12 years into retirement, just one small loss near the beginning. The average 30 yr reinvested return was 10.012% Not bad!

Here is a retirement in 1998 now 20 years old. It’s average return is 5.86% and 5 down years. Notice the negative SOR is bunched up early.

Here is a 19 year retirement beginning Dec 1999. it’s average return is 5.09% and 5 down years.

And here is a 18 year old retirement with a 5.735 return and 5 down years.

Notice how the worst return starts the year BEFORE for bad SORR. 1999 the year before the dot com bust was the worst year in the past 30 years to retire. The next year (2000) the year of the actual bust was better.

Get ready FIRE crowd!

This is a 2008 retirement!

This 10 years yielded a whopping 13.654% annual return and 2 down years.

One year later only 11.967% return, and only 1 down year.

The thing to note NONE of these includes withdrawal of any money, It’s just what would have happened to the portfolio had you filled up the portfolio on the so called year of retirement and let it ride adding no more money, nor subtracting money. Look at the variability WRT to when retirement started and when the bad SOR started in the sequence. It makes a huge difference in income and portfolio viability. Also it points out the red herring of the FIRE movement which came of age during good SOR. It’s hard not to become a Johnny come lately millionaire if you’re making 13%.

I read a recent article on Bucket portfolios by Swedroe, extremely interesting. The paper contains a discussion of techniques to prolong portfolio longevity based on AA optimization in the face of SOR. You have to drill down to the PDF files cited to get the full discussion. The upshot is the bucket portfolio destroys end of life portfolio value based on it’s constant withdrawal of higher risk assets into a risk free asset which earns virtually nothing. A risk free asset is also a return free asset or may even loose some money due to inflation. What the study found on the average the optimum portfolio is a 60/40 portfolio. It beats the bucket because it is properly risked. It also shows people running 80/20 portfolio are suboptimal. It discusses the creation of a metric that looks at half the SD, the downward pushing half, which is the half that drives you into failure. It’s more complicated than I want to write about but I agree with the concept. I’ve been brewing a scheme for a couple years that combats SOR early in retirement. It allows re-sequencing the higher risk portfolio (say the 60/40 portfolio which seems to be optimal) to a more favorable sequence by providing a respite to 60/40 withdrawal from a fund different fund that is the efficient frontier tangent fund. The EF tangent is the balance point of risk and reward, the point where your reward costs the least risk. It has a bit more risk that risk free but also has the propensity to grow. The tangent portfolio of a S&P 500/Short term treasury fund is 20/80 with an expected 6% return. So if you put $500K in a tangent $400K is risk free and 100K is risked optimally. If the market drops 50% the next year (which we saw was worst case). Your tangent would drop $50K in the stocks leaving $450K available, only a 10% dent. You would start living off tangent money and leave the 60/40 closed to withdrawal except for re-balancing. If you go 2 years with no bad SOR the tangent will grow to 561K. Extract 61K and use it to add to the 60/40 by withdrawing less and re-balance the tangent. In other words for the third year of retirement at 100K/yr withdrawal pull 61K out of the tangent and 39K from the 60/40. This effectively donates 61K to the portfolio and you are 3 years into retirement. Do the same at 5 years except donate 100K to the portfolio by removing no money from the 60/40 that year and 100K from the tangent. Re-balance everything. You are now 6 years into retirement. If bad times come close the 60/40 and live off the tangent. If not continue to dollar cost average into the 60/40 by modulating the wealth transfer from the tangent into the 60/40. Since the hit from a SOR is most deadly early you need SOR re-sequence insurance early. As time goes on the portfolio becomes less susceptible to SOR. Eventually the tangent WILL run out of money which means it did its job. There is no need to refill the tangent since the tangent’s job is to refuel your portfolio not act as a drag as in the bucket method. In this case the tangent acts as a supercharger by limiting SORR.

I haven’t worked out the details so the money injections into the 60/40 are my guesstimates and not quantitative. I did do some preliminary work on AA and portfolio SOR risk and it turns out There is a definite danger in excess risk if you draw a bad SOR. On about the 20% SOR line a 80/20 portfolio does much worse and fails much more often than a 50/50 or a 20/80 for the same WR using a total stock/total bond portfolio.

Lotta stuff to think about in this post. In the end low WR big, nest egg and a couple pool’s of variable risk and a plan is a good thing. Early retirement in the context of unusually high rate of return (good SOR) regression to the mean plenty of threads to pull on