Has anyone noticed changes in Firecalc?

[Among_Primates]

I'm a new member, so please forgive me if I commit any errors of custom.

I've used Firecalc pretty steadily for 3-4 years now. Until the last six months or so, my results were fairly consistent. Lately I've been getting two distinct scenarios from the same data (and occasionally a weird outlier). I'm at the point where I've been running the program a couple of times a day and tracking the results.

Result 1: the historically usual result, the one that I was used to over time, is what I'll call "the Kind Scenario." When it comes up these days, I find that my first-stage result (after entering savings, spending, retirement year, money to be added till then, pension and Social Security, and a 50-50 equity/bond split) is 81.9%. According to this scenario, I could retire now and life wouldn't be too bad -- especially when I add the "stage-two data" like portfolio changes (from the future sale of my house, etc.) Then it shoots up to 89.9%. Yay!

Result 2: More frequently of late, I face "the Harsh Scenario." Entering the exact same first-stage data gives me 47.9%. As soon as I see that figure, I know which scenario I'm facing. Adding portfolio changes to this one only gets me to 56.3%. According to this, I'd better continue working for a while.

Over the last few months, the Harsh Scenario's been coming up about 80 percent of the time.

I know you're thinking, "User error is the most likely explanation. No judgment! It's easy to make a typo, to add or lose a zero. Or maybe you did something like leaving the default on 'off chart spending' instead of changing it to 'pension income,' which would make it look as though you were losing money rather than adding. Or what if you put in the year of retirement wrong?"

So I check all those things, and more, religiously. I will add that I used to be a proofreader. And my computer pretty much has the exact same numbers stored and brings them up for me. (Which doesn't stop me from re-checking them obsessively.)

Just to add to the mix, twice now I've gotten what I call an "outlier scenario." Got one this morning, in fact, in which my usual first-stage data resulted in a ZERO success percentage. That was so crazy I figured it HAD to be some kind of typo!

Except... it doesn't seem to be. Bizarre. But at this point I'm no longer taking my results seriously.

Thoughts? Is anyone else seeing changes of any kind over the last several months? Has it become harder or easier to reach your goal? Or maybe people don't run Firecalc with the obsessive frequency that I do.

 

[USGrant1962]

HI, welcome to the forum!

I don't run it as obsessively as pre-RE days, but still fairly regular. I have not noticed this behavior, and I did my annual big model update today.

Do you save your model as a url? That makes sure your data input is exactly the same each time. In the results page at the top right it says "link to this set of data". You can copy/paste like the instructions say or, at least in Chrome, just drag the URL from the URL bar to your desktop to make a shortcut.

 

[Dtail]

I am retired, but still run it fairly frequently and have not noticed any outlier type results.

 

[Car-Guy]

Welcome to the forum....

Years ago, before retiring, I ran just about every retirement calculator that was out there "weekly", including FireCal. Since retiring I quit using all of them except maybe FireCal once or twice a year just for the heck of it and always found it to be consistent.... I haven't run it recently but I've never really noticed any changes over the years.... I always thought FireCal was okay, "but just okay"... For me I became much more comfortable with my own custom excel spreadsheet. I know how it works, I control all the input, formulas and I've tuned/tested/verified it for years.

Even though all the calculators told me I was good to go, I just never trusted any of them.

 

[Midpack]

I know you're thinking, "User error is the most likely explanation. No judgment! It's easy to make a typo, to add or lose a zero. Or maybe you did something like leaving the default on 'off chart spending' instead of changing it to 'pension income,' which would make it look as though you were losing money rather than adding. Or what if you put in the year of retirement wrong?"

I’ve never seen multiple FIRECALC results from the same data, and it seems unlikely. It just spits out results based on inputs, how could it spit out 3 different results using the exact same data? Especially a range including 89.9%, 56.3% and 0% - sorry, there’s just no way with the same inputs. Math calcs and computers don’t result in multiple outcomes - at least with these kinds of calcs, not rocket surgery, just tedious. You’re entering more inputs than many, a good use of FIRECALC, but it would be easy to make a mistake. Sorry I can’t be more helpful.

 

[Among_Primates]

Thank you! My first try at a reply, so hopefully it's going in the right place.

I just ran Firecalc in Firefox, so I could save the data, and got that oddball "zero success" scenario I got this morning. I was not successful at saving the data. So I opened Chrome, ran the whole thing again, and got the "Kind Scenario." Still not successful at saving the data, but the differing results make my head spin.

In trying to save the results, I note that right-clicking gives me the option of "save link as," "copy link address," and "copy." So far I haven't gotten very far with any of them. I did save something as a file at one point, but when I chose "open file" it told me I'd opened the Firecalc results screen without having entered anything to get results. I also tried dragging the URL in the actual URL box, but it didn't want to go to the dock at the bottom of my Apple screen; when I drag it to the horizontal bookmark row at the top it does appear -- but if I click on it, Firecalc comes up and says "you have called the results page directly" and tells me I need to enter information so results can be calculated.

If there is a way of saving my model, I fear you will have to tell me as simply as if you were speaking to an alien, or a small child. Or possibly an alien who is a small child.

 

[REWahoo]

OP, on the Your Portfolio tab, are you clicking the last option, "A portfolio with random
performance"? If so, you will get different results each time you run the calculator.

 

[Among_Primates]

Cancel that -- I just kept trying variations and I found one that works. Model saved!

I'm going to just keep experimenting with it and see where my runs take me. The thing is, if only one of my models is correct, which one is it? You could tell me to check my figures carefully, but I've done that with all of them, so my level of confidence is not high.

This last one gave me a promising future, so I'll keep running that and see what happens. I wish you guys were here standing over my desk! I swear these differing results are nothing like what I used to get.

 

[Among_Primates]

Interesting, Car-Guy. Any prepackaged retirement calculators you'd recommend I take a spin on? So far Firecalc has been the most precise I've found.

 

[SecondCor521]

Cancel that -- I just kept trying variations and I found one that works. Model saved!

I'm going to just keep experimenting with it and see where my runs take me. The thing is, if only one of my models is correct, which one is it? You could tell me to check my figures carefully, but I've done that with all of them, so my level of confidence is not high.

This last one gave me a promising future, so I'll keep running that and see what happens. I wish you guys were here standing over my desk! I swear these differing results are nothing like what I used to get.

You could describe your situation here and I bet most of the regulars could give you a pretty good gut-feel estimate of which is the "real" answer.

Alternatively, you can give us all your inputs and we can tell you what we get when we run FIREcalc. I've never seen the type of behavior you describe, by the way.

 

[compounder]

I've seen this something like this, not quite as drastic with the percentage difference. Only to then figure out that I had entered in Social Security as a monthly amount, opposed to the annual it requires.

 

[Among_Primates]

I'd have gotten back here sooner, but had some long days at work. I've found the answer! Yes, all my figures were correct, and yes, my computer had the numbers stored and filled them in. But I noticed that for one item -- Social Security -- there were two stored numbers to pick from. They were the same figure, but one had a comma and one didn't. I thought, "What if the program treats the comma as a decimal point, the way some European countries do, so instead of $46,000 it reads it as $46?"

When I choose the figure without the comma, I get the positive outcome, where the figure with the comma leads to sadness.

Am I right in my assumption that the no-comma version is the correct one?

As for the occasional weird outlier result... that I have no explanation for, but that doesn't really concern me as much.

 

[SecondCor521]

Yes, you should enter the numbers without commas, so "46000" in your example is the correct way to do it.

 

[Midpack]

I'd have gotten back here sooner, but had some long days at work. I've found the answer! Yes, all my figures were correct, and yes, my computer had the numbers stored and filled them in. But I noticed that for one item -- Social Security -- there were two stored numbers to pick from. They were the same figure, but one had a comma and one didn't. I thought, "What if the program treats the comma as a decimal point, the way some European countries do, so instead of $46,000 it reads it as $46?"

When I choose the figure without the comma, I get the positive outcome, where the figure with the comma leads to sadness.

Am I right in my assumption that the no-comma version is the correct one?

As for the occasional weird outlier result... that I have no explanation for, but that doesn't really concern me as much.

IOW YOU were entering two different inputs and getting two different results. Glad you found your mistake. Like almost any online calculator, manually entering a comma will make the input invalid or it’ll be ignored - taken as $0. It appears FIRECALC sees $0.

 

[mountainsoft]

Interesting, Car-Guy. Any prepackaged retirement calculators you'd recommend I take a spin on? So far Firecalc has been the most precise I've found.

My favorite calculator is the download version of The Flexible Retirement Planner. https://www.flexibleretirementplanner.com/wp/

It's easy to enter my values, easy to try out different scenarios via the "additional inputs" section, and best of all it saves my entries so I can play around with them at any time without having to reenter everything.

However, I occasionally run my numbers in Firecalc also just as a second opinion. I find FRP to be slightly more pessimistic than Firecalc, but don't know if that makes one more accurate than the other. I'd rather err on the pessimistic side and be surprised if things work out better than I planned.