Firecalc knowledge of the past?

[ERD50]

Okay I confess - don't know what the toes comment means.

He's ALWAYS "on his toes". A phrase meaning always sharp, always on top of things, hmmm, those are all metaphor too, I hope they help!

-ERD50

 

[Dtail]

He's ALWAYS "on his toes". A phrase meaning always sharp, always on top of things, hmmm, those are all metaphor too, I hope they help!

-ERD50

Yes, know that term. Very witty on your part...

 

[mrWinter]

Right. FireCalc does not "believe in reversion to the mean".

Suppose I'm ready to retire. FireCalc has input fields for my current assets, income, spending targets, etc.

It does not have an input for "recent market performance" or even "current market P/E ratio".

FireCalc is missing something in its calculations - recent market performance. It does not "believe in" reversion to the mean (at least not wrt averaging performance prior to the run start date with performance after).

Yeah, that's what I'm thinking. It sounds like you agree with me that it is 'missing' this, meaning it should have it and would be more accurate if it did, but maybe you aren't going that far.

I'm thinking the math/calculations wouldn't be that hard. Using past 20 or 10 year or however long total market CAGR, or perhaps using the CAPE, compare that to the CAGR or CAPE of the simulation starting year, and exclude simulations that start with a significantly different value. That might unfortunately trim the subset of potentially usable scenarios down to a smaller number of simulation periods, but those simulation periods would all be more meaningful, more applicable.

 

[mrWinter]

Kitces published a study of the SWR paradox and suggests that market valuation may be considered in picking a SWR. So in the case where there is a bear just before retirement, presumably valuations are lower and you can select a slightly higher SWR. This partially closes the gap in the paradox.

From https://www.kitces.com/wp-content/uploads/2014/11/Kitces-Report-May-2008.pdf

Thanks for sharing, this is just the kind of thing I'm thinking about. Reading now.

 

[samclem]

Like pb4uski said we can just use an off the cuff calculation to mentally trim the worst case runs if we know we just had a recession, just like we can mentally trim the best case runs if we are at the end of a big bull market. But can we do better than just mentally guessing about it? Can we plug actual data into the calculator?

I would think the most straightforward thing to do is to adjust your beginning portfolio value for valuation (maybe use CAPE10) so that it is "normalized," then run FIRECalc. This would tend to remove the irrational exuberance" or "irrational gloom" factor from the start point. Continuing to use the real-world historical returns from the FIRECalc database takes care of the rest.

Now, adjusting the WR in the same way (using PE10, etc) produces an effect that would be very hard to stomach. At the depths of a stock dip, you'd be artifically increasing the WR, which I can't believe would help survivability and which would be something I'd have a hard time doing in practice. So, I'd probably model withdrawal amounts based on the actual, unadjusted-for-PE10 portfolio value.


There's a good (long) thread here where some folks looked at ways to adjust WR for PE10, dis some FIRECalc runs, etc.

 

[pb4uski]

....Otherwise, you are supposedly "in trouble" if you have a bear market just before retirement or just into retirement, or will have issues retiring deep into a bull market.
So when can you retire safely?

Here's your answer:

https://youtu.be/q-JIfjNnnMA

 

[Dtail]

Thanks for sharing, this is just the kind of thing I'm thinking about. Reading now.

Just looked at the box info and at the highest quintile of Cape 10 ratio, still have a 3.9%-4.4% WR at a 40-60 equity exposure.
OTOH, the current ratio is higher than the max in the highest bucket presented.

 

[ERD50]

I'll preface this with "within the confines of the data FIRECalc has", and of course the future may vary from the past. And the finer we try to slice things, the closer we are getting to data-mining and/or curve-fitting. But that's OK for illustrative purposes, and "intellectual discussion".

...
What I'm wondering is, is there a solution to this? For example let's say it's 2009 and I see my less than 100% forecast. ERD50, you said that I need not adjust my plan, because based on the simulations I did in 2007, which took into account this market drop already, I should be good. Well what if I hadn't run that simulation previously in 2007? Maybe I just found out about firecalc in 2009. But perhaps I've kept a history of my balance over the years. I see no reason I couldn't go and run a simulation based on my data back in 2007 and use those results, they'd be just as real as if I had run the same simulation actually back in 2007. Not only could I do that, but the survival rate determined by that simulation set from 2007 would actually be more accurate than the one I was running based on 2009. Right?....

I'd say so, with the above caveats. And only if the 2007-2009 portfolio change was due to market changes and normal WR% spending. If the drop was due to 'investments' in fast cars, old whiskey and fast, young women, maybe not. Not that I'm against that approach.

Another thing you can try searching on the "retire again and again", or "retire again & again". That explains (caveats above again), why a retiree can recalculate a safe WR at any point, and increase their w/d amount, and inflation adjust that going forward, and they still survive in the historical data world.

Easy to visualize - look at all the upward swinging or even flat or shallow dip lines in a FIRECalc output. Clearly, all those people could have raised their spending along the way. The retire again & again approach is the (numerical) answer to not leaving too much on the table, and not spending too much and failing. Of course the future is not the same, but I'm comfortable taking an approach that considers modest increases in spending if things have gone well enough to do that and still have a good margin if (historical) safety.

edit: CAPE 10. Seems reasonable to consider it. I never looked at in detail. One concern I have is it is one number. Does it have the same meaning in a low or high interest rate environment? I dunno.

-ERD50

 

[USGrant1962]

Just looked at the box info and at the highest quintile of Cape 10 ratio, still have a 3.9%-4.4% WR at a 40-60 equity exposure.

Yes, well that is where the 4% rule comes from - Black Tuesday when CAPE was 30, and the mid-60s when CAPE was about 25 with the Great Inflation on the way.

BTW, be careful with CAPE. Due to GAAP accounting changes some people argue that current CAPE is not comparable to pre-1990 CAPE. Here's a discussion https://www.fa-mag.com/news/jeremy-...tios--fatal-flaw--the-bullish-case-24815.html

 

[Dtail]

Yes, well that is where the 4% rule comes from - Black Tuesday when CAPE was 30, and the mid-60s when CAPE was about 25 with the Great Inflation on the way.

BTW, be careful with CAPE. Due to GAAP accounting changes some people argue that current CAPE is not comparable to pre-1990 CAPE. Here's a discussion https://www.fa-mag.com/news/jeremy-...tios--fatal-flaw--the-bullish-case-24815.html

Yes understood. I guess what I was saying is that even with high Cape 10 ratios the WR is still around 4% not drastically reduced due to supposed high valuations.

 

[SecondCor521]

Phew great, you guys are getting what I'm getting at, thanks for sticking with me. And yes I feel like this is related to or the same as the paradox with two retirees one year apart, I'm remembering that other thread on Larry Moe and Curly now.

What I'm wondering is, is there a solution to this? For example let's say it's 2009 and I see my less than 100% forecast. ERD50, you said that I need not adjust my plan, because based on the simulations I did in 2007, which took into account this market drop already, I should be good. Well what if I hadn't run that simulation previously in 2007? Maybe I just found out about firecalc in 2009. But perhaps I've kept a history of my balance over the years. I see no reason I couldn't go and run a simulation based on my data back in 2007 and use those results, they'd be just as real as if I had run the same simulation actually back in 2007. Not only could I do that, but the survival rate determined by that simulation set from 2007 would actually be more accurate than the one I was running based on 2009. Right?

Like pb4uski said we can just use an off the cuff calculation to mentally trim the worst case runs if we know we just had a recession, just like we can mentally trim the best case runs if we are at the end of a big bull market. But can we do better than just mentally guessing about it? Can we plug actual data into the calculator?

What if we gave our retirement calculator our past balance history, or plugged in past market data, or provided current market CAPE, so it new that we were in the depths of a serious market drop already, and to simulate starting retirement going forward with another serious market drop would be worse than ever in past history. I'm not suggesting we can do this with firecalc, just wondering aloud. This would trim overly pessimistic scenarios when running the calculator at the end of a bear market, and trim overly optimistic scenarios when running it at the end of a long bull market.

It's not a paradox. It's pretty basic conditional (I think some may say Bayesian) probabilities. If I tell you I have a fair coin and offer to wager you on heads or tails, you may be willing to believe it. After it turns up tails nine times out of ten and you owe me money, you may suspect it is either not a fair coin or you have been extremely unlucky. The key point is that additional data can change your assessment of probabilities - in your posts it is the additional data of 2007 and 2008.

There are people out there who think like you do and are trying to include valuation as a thumb on the scale to adjust WR somehow. I believe someone here has adopted some form of this (audreyh1 or athena53 maybe?) by planning to vary both her investment mix and maybe her WR based on CAPE. There is also a blog series at ERN here that may have some interesting reading for you with quite a bit of analysis (or data mining, if you prefer):

https://earlyretirementnow.com/2016/12/07/the-ultimate-guide-to-safe-withdrawal-rates-part-1-intro/

For me I personally like the "retire again and again" strategy instead.