A different spin on SWR, question.

[smjsl]

@Gone4Good: I agree with your last post. Good point.

[REWahoo]

Quote:

Originally Posted by Cb You could also say that by holding less than ~80% equities you're giving up a lot of potential portfolio growth with little or no improvement in SWR.

Yes, but what isn't obvious from that chart is the greater volitility your portfolio is likely to experience with an allocation of 80% equities vs 40-50%. "Success" in FIRECalc is defined as not having your portfolio go to zero. Some of those near death dips could be very unpleasant to experience first hand.

Of course at 80% equities the roller coaster ride is likely to be so wild that you'd die of apprehension long before your ran out of money...

[Cb]

Quote:

Originally Posted by REWahoo Yes, but what isn't obvious from that chart is the greater volitility your portfolio is likely to experience with an allocation of 80% equities vs 40-50%. "Success" in FIRECalc is defined as not having your portfolio go to zero. Some of those near death dips could be very unpleasant to experience first hand. Of course at 80% equities the roller coaster ride is likely to be so wild that you'd die of apprehension long before your ran out of money...

If the 50% equity allocation has the same 3.8% SWR, would there not (by definition) be similar near misses that establish the 3.8% SWR?

[REWahoo]

Quote:

Originally Posted by Cb If the 50% equity allocation has the same 3.8% SWR, would there not (by definition) be similar near misses that establish the 3.8% SWR?

To continue the roller coaster analogy, the ride would follow the same pattern but the highs would be higher and the lows would be lower, which is exactly what you would expect from the higher volatility of an 80/20 AA.

Do a couple of FIRECalc runs on a $1M portfolio withdrawing $40k per year, using an 80/20 allocation on one and a 40/60 on the other. The success rate of the two won't vary by more than a couple of percent but if you do a close comparison of the "spaghetti graphs" on the results page, you'll see the grouping of lines on the 80/20 come much closer to the zero line during the 30 years than does the 40/60.

On first glance both charts appear very similar, but note that the values on the horizontal axis of each graph aren't the same. Take a look around the 20 year mark at how close many of the runs on the 80/20 are to the red "game over" line vs the 40/60.

For those of us more interested in keeping away from the red line than ending up close to the top line, an equity allocation between 40 and 50% can make a significant difference in how well we sleep at night.

[Cb]

Quote:

Originally Posted by REWahoo Do a couple of FIRECalc runs on a $1M portfolio withdrawing $40k per year, using an 80/20 allocation on one and a 40/60 on the other. The success rate of the two won't vary by more than a couple of percent but if you do a close comparison of the "spaghetti graphs" on the results page, you'll see the grouping of lines on the 80/20 come much closer to the zero line during the 30 years than does the 40/60. On first glance both charts appear very similar, but note that the values on the horizontal axis of each graph aren't the same. Take a look around the 20 year mark at how close many of the runs on the 80/20 are to the red "game over" line vs the 40/60.

The disparity in the ending values of the two graphs (~2.5x in my 40 year runs) is so large that I don't think that comparing the "spaghetti graphs" are all that informative.

I haven't really dug into it by pouring over the detailed results of dozens of runs, but my guess is that failures for the high FI retirees are more likely to caused by not keeping up with inflation over the long term - toward the end of the 40 year period things gradually tight up for the retiree.

For the high equity retiree, he probably has more problems if he runs into trouble in the first 5-10 years...if he clears that his portfolio has likely reached "escape velocity".

If that's the case, I'd argue that the guy 5-10 years from the cube (me) has a better shot at replenishing his portfolio or making other adjustments than the 80 year old does.

[REWahoo]

Quote:

Originally Posted by Cb The disparity in the ending values of the two graphs (~2.5x in my 40 year runs) is so large that I don't think that comparing the "spaghetti graphs" are all that informative. I haven't really dug into it by pouring over the detailed results of dozens of runs, but my guess is that failures for the high FI retirees are more likely to caused by not keeping up with inflation over the long term - toward the end of the 40 year period things gradually tight up for the retiree. For the high equity retiree, he probably has more problems if he runs into trouble in the first 5-10 years...if he clears that his portfolio has likely reached "escape velocity". If that's the case, I'd argue that the guy 5-10 years from the cube (me) has a better shot at replenishing his portfolio or making other adjustments than the 80 year old does.

If you are making the point that an 80% allocation to equities improves your chances of dying with a large portfolio, I completely agree. Just be prepared for what could be a wild ride to the cemetery.

[haha]

Something very easy to forget is that very many of us would bail out if our portfolios lost a large amount. Perhaps not if we are very over-reserved, but many here want to retire as soon as they get what they consider to be a moderate margin of safety. Back in '08 and early '09 retirees would write and ask advice givers , "should I sell out?" The answer often included, "if you could not take more loss, yes. Otherwise hold on". CYA maybe, but survival is what most retirees are thinking about thinking about under these circumstances.

We can look back on the recent problems and think-"That worked out fine", and we will feel emboldened- but that is far from the only way it might have worked out, and though we are out of that swamp, we are not exactly out of those woods.

We are all part of a big social experiment in self funded, self managed retirement, and the results are not yet in.

Ha

[Independent]

This has been mentioned before, but I think it's relevant to any bond/stock analysis with FireCalc. On the bond side, the system's annual returns seem to be new money interest rates, not the returns you would get on a mutual fund that holds bonds.

[Independent]

Quote:

Originally Posted by Cb The disparity in the ending values of the two graphs (~2.5x in my 40 year runs) is so large that I don't think that comparing the "spaghetti graphs" are all that informative. I haven't really dug into it by pouring over the detailed results of dozens of runs, but my guess is that failures for the high FI retirees are more likely to caused by not keeping up with inflation over the long term - toward the end of the 40 year period things gradually tight up for the retiree. For the high equity retiree, he probably has more problems if he runs into trouble in the first 5-10 years...if he clears that his portfolio has likely reached "escape velocity". If that's the case, I'd argue that the guy 5-10 years from the cube (me) has a better shot at replenishing his portfolio or making other adjustments than the 80 year old does.

Note that instead of looking at spaghetti graphs, FireCalc lets you download Excel files that show each run. I think it's informative to look at the lowest number for each start year. That's the "Is the market ever going to recover?" year.

[ERD50]

Quote:

Originally Posted by REWahoo Yes, but what isn't obvious from that chart is the greater volitility your portfolio is likely to experience with an allocation of 80% equities vs 40-50%. "Success" in FIRECalc is defined as not having your portfolio go to zero. Some of those near death dips could be very unpleasant to experience first hand. Of course at 80% equities the roller coaster ride is likely to be so wild that you'd die of apprehension long before your ran out of money...

Quote:

Originally Posted by Cb If the 50% equity allocation has the same 3.8% SWR, would there not (by definition) be similar near misses that establish the 3.8% SWR?

There seems to be a perception that a higher bond allocation has a large smoothing effect on the 'roller coaster ride'. Run these two scenarios:

FIRECalc: A different kind of retirement calculator

FIRECalc: A different kind of retirement calculator

I entered $1M start portfolio for easy math, 3.5% WR, and ten year period to see how we are doing at that point (apparently, FIRECALC reports the minimum as the minimum ending value, not intermediate values).

So at the 10 year mark, with 80% EQ, your roller coaster has dipped from $1M to $410K.

And the 40% EQ/60% Bonds has had this gentle lazy-river-ride, down to... $431K! Not much difference, really.

With those numbers the 40% EQ did slightly better at the 30 and 40 year marks, but the same number of failures at 40 years for each (4 failures). But I don't think the difference warrants "roller coaster" descriptions.

Seems the "Investigate" tab isn't working for me when I choose the option to display AA versus success rate. That would tell us more.

-ERD50

[REWahoo]

ERD, are you saying there is little difference in the volatility of a 40% vs an 80% allocation to equities over a 30 year period?

Quote:

Originally Posted by ERD50 Seems the "Investigate" tab isn't working for me when I choose the option to display AA versus success rate. That would tell us more.

It works fine for me - maybe you weren't logged in?

[ERD50]

Quote:

Originally Posted by REWahoo ERD, are you saying there is little difference in the volatility of a 40% vs an 80% allocation to equities over a 30 year period?

Not necessarily - those numbers were not looking at volatility specifically, they were looking at "how low can you go" - one side of the volatility picture.

Volatility would also be looking at how high things go, and the 80% allocation goes far higher than the 40% allocation. But for this discussion, we weren't concerning ourself with looking at how high we go, just if we run into trouble with a low scenario.

So, FIRECALC is saying that there is little difference in the lows of a 40% vs an 80% allocation to equities over a 30 year period.



RE: Investigate tab for AA versus success%:

Quote:

It works fine for me - maybe you weren't logged in?

It has worked for me before. I made a modest donation a few years back, and it was acknowledged, but I'm not sure I ever got the 'key' to any advanced features and I don't think I've ever logged in with a special PW or anything. I should also try another browser, I switched to Chromium a while back, that might be it.

-ERD50

[REWahoo]

Quote:

Originally Posted by ERD50 So, FIRECALC is saying that there is little difference in the lows of a 40% vs an 80% allocation to equities over a 30 year period.

Maybe I'm beyond the age of reason, but that would seem to defy logic.

The bottom line to me is this: Retired and at age 64, I don't have the cast iron huevos I think necessary to allocate 75-80% of my portfolio to equities.

[Cb]

Quote:

Originally Posted by REWahoo If you are making the point that an 80% allocation to equities improves your chances of dying with a large portfolio, I completely agree. Just be prepared for what could be a wild ride to the cemetery.

The point I was trying to make is a higher equity allocation (probably - I should research this some more) shifts more of the risk toward the earlier years, when an early retiree has more options. If the first decade doesn't go so well, full or part time employment is probably an option. It'd be the rare ER that went back to work at similar pay, but all he or she would need is some income to fill the gap. An 80 year being gradually squeezed by inflation would have fewer options.

As to the possibility of greater portfolio growth, I guess I wouldn't mind being wealthier at 75 than when I left work, so that does have some impact on my allocation.

[Cb]

For what it's worth, I left work at 47, and my wife is still waiting for that package at 50, so I spend very little time running 30 year scenarios. 40 year FIREcalc runs now include the late 60's failures, so I see little point in 30 year analysis for ER's.

[Cb]

Quote:

Originally Posted by REWahoo Maybe I'm beyond the age of reason, but that would seem to defy logic.

But that's what the chart you linked up-thread suggests...

[REWahoo]

Quote:

Originally Posted by Cb But that's what the chart you linked up-thread suggests...

Lets look at a real-life example:

On October 1, 2007 retiree A and B both had $1m portfolios. Retiree A had an AA of 80/20, retiree B had an AA of 40/60. The S&P 500 was roughly 1,500. If neither made any change to their AA, did A or B have the greatest dip in portfolio value when the S&P dropped under 700 in March of 2009?

I'm probably guiltier that most at looking at FIRECalc charts and attempting to interpret the details behind those squiggly little lines. While that might be a great academic exercise, I should heed what Dory said regarding his calculator:

Quote:

.. keep in mind that a tool like FIRECalc is intended to help you see if your plan pushes too close to the edge, decades from now. You can try 1001 different combinations of investments and timings to see what happens, but don't fall into the trap of measuring something with a micrometer when you'll be cutting it with an axe.

I'm trying to focus on the big picture as I see it - for me an AA on the lower end of the acceptable equity scale fits my comfort level. That decision should be different for each individual based on their own personal circumstances and risk tolerance.

I've said my piece on this subject and I'm off to read Alan and T-Al's latest jokes...

[ERD50]

Quote:

Originally Posted by REWahoo Maybe I'm beyond the age of reason, but that would seem to defy logic. The bottom line to me is this: Retired and at age 64, I don't have the cast iron huevos I think necessary to allocate 75-80% of my portfolio to equities.

I don't think it defies logic, unless there is a logic flaw in the FIRECALC calculations.

But I understand that it defies what our guts are telling most of us, including me. I'm pretty surprised that bonds don't cushion more than they seem to.

I just think people should be aware that a high bond allocation isn't going to take them from 'roller coaster' to 'smooth sailing'. A 40% EQ allocation might be a fine choice for many, but it doesn't appear you can count on it to do much in the way of avoiding dips. It looks like it does some, and that might be enough for some people, but it doesn't seem to be much.

edit - cross posted with your last post there -

-ERD50

[Cb]

Quote:

Originally Posted by REWahoo On October 1, 2007 retiree A and B both had $1m portfolios. Retiree A had an AA of 80/20, retiree B had an AA of 40/60. The S&P 500 was roughly 1,500. If neither made any change to their AA, did A or B have the greatest dip in portfolio value when the S&P dropped under 700 in March of 2009? .

Portfolio A!

Let's say 50/50 and 80/20 portfolios both have 2-3% 40 year failure rates at, I dunno, a 3.7% SWR. Seem reasonable?

Ninety some odd times out of a hundred, the heavier equity allocation yields a better ending value. When they don't they both fail.

Better sleep during bear markets certainly has some value for the low equity ER.

So does the possibility (probability) of greater wealth later in retirement for the high equity ER.

The "micrometer" of historical SWR analysis suggests that for longer retirements that ER's contemplate (40 years and more), equity allocations >75% are optimal. I happen to think that in terms of ER failure, high equity allocation (especially when diversified beyond the S&P 500) IS the low risk approach.

[REWahoo]

Quote:

Originally Posted by Cb I happen to think that in terms of ER failure, high equity allocation (especially when diversified beyond the S&P 500) IS the low risk approach.

Yep. Ultimately it comes down to what each of us thinks.