Author Topic: Did the Great Resignation class of 21-22 just pick the worst time to retire?  (Read 110391 times)

maizefolk

  • Walrus Stache
  • *******
  • Posts: 7400
While that big blue blob isn't all that informative, believe it or not those lines were already at an alpha of 0.3. Should have been reasonably transparent with any reasonably number of overlaps. But there's just an awful lot of possible 30 year intervals if you look at every possible start month from January of 1871 onward and those settings clearly weren't cutting it.

I played around a bit with both line width (wider) and transparency (much more transparent) and I think this version (while looking a bit blurry) does a better job of representing where portfolios tend to more commonly go.



Some sort of kernel density style transformation -- or a heatmap? -- might work better, but I don't know how to do that in a way that would look correct for time series data.

dividendman

  • Handlebar Stache
  • *****
  • Posts: 1900
Let's go green line! I need to know when to buy my lambo!

Tass

  • Magnum Stache
  • ******
  • Posts: 3188
  • Age: 30
  • Location: Crossing some mountains
One thing I've always found interesting about @maizefolk's graph is that the deepest dips in the first 1-2 years are all blue lines. The red lines don't start showing up at the bottom of the distribution until years 2-5. Is a steep drop in year 4 really more dangerous than in year 1, or is there a weird artifact going on there?

AlanStache

  • Magnum Stache
  • ******
  • Posts: 3175
  • Age: 44
  • Location: South East Virginia
Nice, yeah I think keeping the path dependent information is good.  Just showing an end point removes the "OMG - look how low/high it went part way along".

maizefolk

  • Walrus Stache
  • *******
  • Posts: 7400
One thing I've always found interesting about @maizefolk's graph is that the deepest dips in the first 1-2 years are all blue lines. The red lines don't start showing up at the bottom of the distribution until years 2-5. Is a steep drop in year 4 really more dangerous than in year 1, or is there a weird artifact going on there?

My suspicion is it's a sample size issue. Blue represents a lot more of total possible stock market trajectories than red.

There are 1,447 possible start months that ended above zero after 30 years (the blue lines). And 31 start months that end up at or below 0 after 30 years (about 2% of all possible start months). And even those 31 start months aren't really independent but are clustered in really only two time periods in the historical record which are going to have similar trajectories.

Failure months:
1929.01
1929.02
1929.03
1929.04
1929.05
1929.06
1929.07
1929.08
1929.09
1929.1
1930.03
1930.04
1930.05
1965.05
1965.09
1965.1
1965.11
1965.12
1966.01
1966.02
1966.04
1967.09
1967.1
1968.06
1968.07
1968.09
1968.1
1968.11
1968.12
1969.01
1969.05

Tyson

  • Magnum Stache
  • ******
  • Posts: 3025
  • Age: 52
  • Location: Denver, Colorado
One thing I've always found interesting about @maizefolk's graph is that the deepest dips in the first 1-2 years are all blue lines. The red lines don't start showing up at the bottom of the distribution until years 2-5. Is a steep drop in year 4 really more dangerous than in year 1, or is there a weird artifact going on there?

My suspicion is it's a sample size issue. Blue represents a lot more of total possible stock market trajectories than red.

There are 1,447 possible start months that ended above zero after 30 years (the blue lines). And 31 start months that end up at or below 0 after 30 years (about 2% of all possible start months). And even those 31 start months aren't really independent but are clustered in really only two time periods in the historical record which are going to have similar trajectories.

Failure months:
1929.01
1929.02
1929.03
1929.04
1929.05
1929.06
1929.07
1929.08
1929.09
1929.1
1930.03
1930.04
1930.05
1965.05
1965.09
1965.1
1965.11
1965.12
1966.01
1966.02
1966.04
1967.09
1967.1
1968.06
1968.07
1968.09
1968.1
1968.11
1968.12
1969.01
1969.05

So the 4% approach works for 1,416 times out of 1,447 months.

maizefolk

  • Walrus Stache
  • *******
  • Posts: 7400
It worked -- for a 100% stock portfolio -- in 1,447 of 1,478 start months (1447+31), which is ever so slightly better.

Tyson

  • Magnum Stache
  • ******
  • Posts: 3025
  • Age: 52
  • Location: Denver, Colorado
It worked -- for a 100% stock portfolio -- in 1,447 of 1,478 start months (1447+31), which is ever so slightly better.

That's insanely robust.

Tass

  • Magnum Stache
  • ******
  • Posts: 3188
  • Age: 30
  • Location: Crossing some mountains
My suspicion is it's a sample size issue. Blue represents a lot more of total possible stock market trajectories than red.

There are 1,447 possible start months that ended above zero after 30 years (the blue lines). And 31 start months that end up at or below 0 after 30 years (about 2% of all possible start months). And even those 31 start months aren't really independent but are clustered in really only two time periods in the historical record which are going to have similar trajectories.

This is what I was missing (and seems obvious in retrospect), thanks! Those steep blue dips at the beginning are also probably a cluster from just a few time periods.

SuperNintendo Chalmers

  • Stubble
  • **
  • Posts: 103
  • Location: Colorado, USA
With March 1st falling on Friday I have three more months worth of data to plug into tracking how the hypothetical '21-22 retiree's trajectory compares to every starting month whether the 4% rule has failed with a 100% stock portfolio.



Taking into account withdrawals the '21-22 retiree is sitting on a portfolio worth about 23 years of their current (inflation adjusted) expenses. That doesn't feel all that bad for being a bit more than two years into FIREing at the peak right before a bear market. It still wouldn't be unprecedented for this start month to end in a failure. But it's looking less and less likely. 

Interestingly, the recent run up has pushed the hypothetical 2000 tech bubble retire high enough to push them from the least-bad four percent failure scenarios to the least-food four percent success* scenarios.

*Where "success" means having a net worth >$0 thirty years after retirement.

The 2000 vs. the 2007 retiree lines show how SORR plays out.  With that said, it's interesting that 2007 look more like a failure in those crucial early years only to go on to complete and utter success (so far).  2000 looked like they were hanging in there only to get jumped in year 7 and now even all those good returns from this historic US bull market (thank you Fed + National debt + above average Shiller-CAPE!) are barely enough...

that is really a tough line they are travelling. 2000 didn't recover and then 2007 kicked them when they were down. not sure if a 3% wr or 2 years of cash would have been enough to keep them in the ballpark. Maybe both??

even if they do make, I'd have preferred not crapping my pants for 30 years.....you know before i get to the nursing home for pant crapping...

My risk-averse self would have been greeting Walmart shoppers by year 2

NorthernIkigai

  • Bristles
  • ***
  • Posts: 330
  • Connoisseur of Leisure
In year 2 you could probably still get back into your old career, depending on what it was of course. Which might be more stressful than a taking a barista job, but financially more satisfying. 5 years in, it might be much harder.

PhilB

  • Walrus Stache
  • *******
  • Posts: 5703
It worked -- for a 100% stock portfolio -- in 1,447 of 1,478 start months (1447+31), which is ever so slightly better.

Another way of looking at it is to say that there have been only two 'events' (with the 2000 tech bubble possibly being a third) that have been bad enough to break a 4% US stocks portfolio.  The Wall Street Crash / Depression and the stagflation of the 1970s.  Presumably a lot of the close shave portfolios are also heavily influenced by these events.

The good news is that economists have, hopefully, learned from those events making a repeat less likely.  The bad news is that there have been lots of other events hitting other economies over that period which could potentially happen to the US at some point.  And a big enough bubble can always cause a wipe out whatever actions the politicians / economists take to aid recovery - Japan anyone?

 

Wow, a phone plan for fifteen bucks!