Author Topic: FI Calc: a new retirement calculator  (Read 8960 times)

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #50 on: June 02, 2020, 07:37:46 PM »
First of all, ***thank you*** for making downloadable *.csv files that are import-ready for programs such as R and Matlab.  As a data-wrangler that's been my biggest beef with other retirement calculators (it's takes a lot of data-massaging to do any further analysis).

Oh, nice! Although I admit I didn't intentionally test supporting those programs, I'm glad that I didn't unintentionally break the CSV files for them. This is definitely a use case I want to support, though. Do let me know if there's anything I can do to make your life easier when it comes to exporting the data into these other apps that you use.

A few things I would like to see
  • Highlight/focus on years that resulted in failures in the "Results" section.  That's really what people care most about, and it can be enlightening to see under what conditions (e.g. market returns, inflation, SORR) these failures occurred
  • In the 'Results' section generate a single graph with the median return as a thick, red line and the CI from that return as a gray area, and the SD within as a blue area.  That will quickly show how fast a portfolio will grow/shrink with 'average' conditions, as well as the spread throughout the period.

Just my 2Ę. 

I love these suggestions, and both are things that I want to add. In fact, an earlier version of the app did indicate which years failed (you can see screenshots earlier on in this thread). I temporarily removed because some withdrawal strategies, such as VPW, intentionally bring you to $0 in the final year, so for those strategies it is a good thing when your portfolio hits $0 (because they are operating as intended). I am working on a way to differentiate how results are displayed based on the goals of each individual strategy, but there's still a way to go. Added it back for all strategies!

I also agree that there needs to be an overview graph. It's on the todo list.

BTW, as others have said "PTF" means "Posting to Follow".  Basically I did that because I wanted to come back and spend more time looking at your calculator and giving constructive feedback, but I didn't have the time to do a proper job when I first came across your thread.

Good to know! Thank you :D

Thanks for doing this!

Chuckling that I'm posting right after a discussion of PTF. I haven't had time to read all the details yet, am indeed posting so I can easily catch up later. :)

Fwiw, skimming, I was going to ask for downloadable results. It looks like you're doing that already; thumbs up.

If you want a light income stream to cover possible expenses, maybe you could set up a Patreon account for it?

Yo @BicycleB ! Thanks for stopping by. A Patreon is an interesting idea that I hadn't considered. I'll definitely keep it on the table as I think about the different options.

Atm I'm thinking I'll just setting aside enough of my personal money to pay the cost for a couple of years, just to give me time to build the best app possible w.o. thinking about the financial aspect.



Also, @themagicman , Monte Carlo is in beta. There is still a lot of work to do, and it's a secret option in the calculator atm, but you can see a verrrrrrry early version at this link:

https://calculator.ficalc.app/?calculationType=monteCarlo

There is obviously a ton of work that I need to do (way, way more simulation cycles, verifying the results, etc.), but the feature is well underway :)
« Last Edit: June 03, 2020, 04:25:46 PM by jamesplease »

themagicman

  • Bristles
  • ***
  • Posts: 394
  • Age: 30
  • Location: Atlanta, GA
Re: FI Calc: a new retirement calculator
« Reply #51 on: June 04, 2020, 09:44:33 AM »
Great this is going to be very beneficial! When I went to look at it, it looks like the the success percentage is gone (and the yearly) for all but the constant withdraw option. I was originally thinking this was because it is still being worked on but then I noticed that the success percentage is going away even in the historical results calculator. 

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #52 on: June 04, 2020, 10:30:30 AM »
Great this is going to be very beneficial!

:D

When I went to look at it, it looks like the the success percentage is gone (and the yearly) for all but the constant withdraw option. I was originally thinking this was because it is still being worked on but then I noticed that the success percentage is going away even in the historical results calculator.

It's been this way for awhile, and you are right actually that it's something that's still being worked on. This just the current state of where the Results section of the calculator is at right now.

I've been pretty focused on the "Configuration" side of the calculator the past few weeks, and the "Results" section hasn't been updated much at all. The Configuration section is maybe 75% where I want it to be, but the Results section is maybe 20% where I want it to be.

I'm realizing now to finish Monte Carlo I'll need to put some more thought into where I'm headed with the Results, so I'll be making some more updates to Results over the next few days/weeks as part of preparing for Monte Carlo. One of the planned updates will be making the results more consistent for all of the different strategies.

nereo

  • Senior Mustachian
  • ********
  • Posts: 13360
  • Location: Just south of Canada
    • Here's how you can support science today:
Re: FI Calc: a new retirement calculator
« Reply #53 on: June 04, 2020, 10:40:32 AM »
I like that you've added a color key to the years that failed.  Nice and effective visual. 
It might be even better if you expanded it to be a true 'heat-map', with Red (years that failed), Orange (years that ended with < 35% starting assets) and then a couple shades of green for portfolios that ended 100-200% of starting assets, 200-400% above starting assets and >400% above starting assets.

...or something like that.  It's great to see the failures and how they tend to cluster togerther.  It's also nice to see years that were runaway successes where you wound up with far more money than you ever needed.

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #54 on: June 06, 2020, 11:09:20 AM »
I like that you've added a color key to the years that failed.  Nice and effective visual. 
It might be even better if you expanded it to be a true 'heat-map', with Red (years that failed), Orange (years that ended with < 35% starting assets) and then a couple shades of green for portfolios that ended 100-200% of starting assets, 200-400% above starting assets and >400% above starting assets.

...or something like that.  It's great to see the failures and how they tend to cluster togerther.  It's also nice to see years that were runaway successes where you wound up with far more money than you ever needed.

My long-term goal is to allow a degree of customization with the squares and the analysis, so that each individual can specify the kinds of things that they are interested in. I think adding a green square by default might make sense, though. I'll try it out and see how it feels.



Today's update: charts on the Results overview page. They're not perfect, but I'm curious to get any feedback that folks have on the charts. Do these convey all of the information you'd like to see from a visual representation of the data? Is there something else you're interested in seeing that the charts don't show?

exit2019

  • 5 O'Clock Shadow
  • *
  • Posts: 51
Re: FI Calc: a new retirement calculator
« Reply #55 on: June 06, 2020, 12:05:25 PM »
I would appreciate it if a year-by-year withdrawal graph was presented so that one could verify that additional withdrawals are being accounted for (for example two lines showing baseline per the withdrawal strategy and the actual, modified withdrawal, above that).  Looking at the CSV, what I see is an interesting pattern where the "Additional Withdrawal Amount" drops to zero for long stints even when one configures additional expenses.

For example, I added two expenses representing potential excess medical expenses of 30k a year, one starting at +20Y and one starting at +30Y, each set to repeat for 100Y (that is, forever).  In the "Additional Withdrawal Amount" this seems not to be doing what it should be.
« Last Edit: June 06, 2020, 12:17:25 PM by exit2019 »

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #56 on: June 06, 2020, 12:19:27 PM »
I would appreciate it if a year-by-year withdrawal graph was presented so that one could verify that additional withdrawals are being accounted for (for example two lines showing baseline per the withdrawal strategy and the actual, modified withdrawal, above that).  Looking at the CSV, what I see is an interesting pattern where the "Additional Withdrawal Amount" drops to zero for long stints even when one configures additional expenses.

For example, I added two expenses representing potential excess medical expenses of 30k a year, one starting at +20Y and one starting at +30Y, each set to repeat for 100Y (that is, forever).  In the "Additional Withdrawal Amount" this seems not to be doing what it should be.

Withdrawals can drop to 0 if the portfolio runs out of money, which may explain what youíre seeing.

Can you share a link to this calculation? There is a ďShare CalculationĒ button on the overview page that will grab a URL you can share.



It's also nice to see years that were runaway successes where you wound up with far more money than you ever needed.

Agreed. I added a single color that represents > 300%. I strayed away from green, as I didn't want folks to think that these years are necessarily a good thing.

In the future, I'd like for folks to be able to customize this number as well as add additional ones.
« Last Edit: June 06, 2020, 10:09:58 PM by jamesplease »

themagicman

  • Bristles
  • ***
  • Posts: 394
  • Age: 30
  • Location: Atlanta, GA
Re: FI Calc: a new retirement calculator
« Reply #57 on: June 07, 2020, 06:19:45 AM »
I like that you've added a color key to the years that failed.  Nice and effective visual. 
It might be even better if you expanded it to be a true 'heat-map', with Red (years that failed), Orange (years that ended with < 35% starting assets) and then a couple shades of green for portfolios that ended 100-200% of starting assets, 200-400% above starting assets and >400% above starting assets.

...or something like that.  It's great to see the failures and how they tend to cluster togerther.  It's also nice to see years that were runaway successes where you wound up with far more money than you ever needed.

My long-term goal is to allow a degree of customization with the squares and the analysis, so that each individual can specify the kinds of things that they are interested in. I think adding a green square by default might make sense, though. I'll try it out and see how it feels.



Today's update: charts on the Results overview page. They're not perfect, but I'm curious to get any feedback that folks have on the charts. Do these convey all of the information you'd like to see from a visual representation of the data? Is there something else you're interested in seeing that the charts don't show?

Charts look great! I think for me the two big things you have covered with ending value and withdraw amount. I am not really sure on if there is any other chart I would want to see but I will be thinking on that. Two things of feedback on the chart (the application itself not the data)

1. It is not clear to me how I can can view sections on the chart. For example, if I want the chart to show me range and the percentage of 1 standard deviation (I am assuming those lines are standard deviation) from the mean and 2 staandard deviations from the mean, it is difficult to get the system to do this. It has done it for me before but I cannot figure it out. It would be nice to click in there and it select that. When I hover over it, it is showing me the whole 90% range. Does that make sense? If not, I can reword with screen shots

2. It would be nice to be able to change the calculating amounts on the charts. For example, I want to see ending values between X and Y and know what % that is.

themagicman

  • Bristles
  • ***
  • Posts: 394
  • Age: 30
  • Location: Atlanta, GA
Re: FI Calc: a new retirement calculator
« Reply #58 on: June 07, 2020, 06:22:22 AM »
Also on the Monte Carlo sim, does it take the data averages based on the history and use that for the sim?

I really hope that the monte carlo sim is in the early enough stages that it is just not calculating right and not the fact that my plan is not going to hold up. Getting a 99% chance of success of my plan using historical data and only 35% on monte carlo data
« Last Edit: June 07, 2020, 06:27:44 AM by themagicman »

themagicman

  • Bristles
  • ***
  • Posts: 394
  • Age: 30
  • Location: Atlanta, GA
Re: FI Calc: a new retirement calculator
« Reply #59 on: June 07, 2020, 06:38:39 AM »
Also, I can't remember if I had saw this before or not but love the start year,end year, current year columns on the CSV output!

Accountant

  • 5 O'Clock Shadow
  • *
  • Posts: 27
Re: FI Calc: a new retirement calculator
« Reply #60 on: June 07, 2020, 09:17:02 AM »
This calculator is great.  Thank you.

BikeFanatic

  • Bristles
  • ***
  • Posts: 494
Re: FI Calc: a new retirement calculator
« Reply #61 on: June 07, 2020, 09:55:30 AM »
Another feature that I would like is the ability to decrease my withdrawal rate once social security kicks in. I could not figure out how to calculate that in. Otherwise this is a great calculator.

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #62 on: June 07, 2020, 10:06:04 AM »
1. It is not clear to me how I can can view sections on the chart. For example, if I want the chart to show me range and the percentage of 1 standard deviation (I am assuming those lines are standard deviation) from the mean and 2 staandard deviations from the mean, it is difficult to get the system to do this. It has done it for me before but I cannot figure it out. It would be nice to click in there and it select that. When I hover over it, it is showing me the whole 90% range. Does that make sense? If not, I can reword with screen shots

The chart shows you the section that you're hovered over. So if you move your mouse (or finger) horizontally in the chart, you should see details of whatever specific item you're hovered on. It's the same interaction model used for all of the charts in the app. Let me know if that helps!

2. It would be nice to be able to change the calculating amounts on the charts. For example, I want to see ending values between X and Y and know what % that is.

Interesting idea. I've taken and a note and I'll think about how I can add something like this.

Also on the Monte Carlo sim, does it take the data averages based on the history and use that for the sim?

I really hope that the monte carlo sim is in the early enough stages that it is just not calculating right and not the fact that my plan is not going to hold up. Getting a 99% chance of success of my plan using historical data and only 35% on monte carlo data

I wouldn't recommend using Monte Carlo to test anything right now since there is a lot of work to do. For example, it only runs 0.2% of the target number of calculations that I want it to run, so the success rate metrics etc. will all be inaccurate for that reason. Also, I have no tests to verify that it's working as expected. I just sent over the link as a little demo, but it's very much a work in progress feature.

With that said, I'd say I'm 95% certain that the results that you're seeing are correct. Should that be cause for worry? Maybe, but probably not.

Here's what's going on: FI Calc uses data from 1871 to 2020 by default. This is OK for the backtesting calculations because it always goes chronologically like so: 1871...1872... ...2019...2020. Throughout this process, the market steadily grows, but never too much from year to year.

However, the total difference in stock price between years close to 1871 and years close to 2020 is enormous. Because the Monte Carlo calculation randomly samples data from the full data set, it's regularly flipping back and forth between these two extremes (as random chance would have it). Using the full data set for Monte Carlo (which again, is the default right now) is effectively like testing how your portfolio performs if the worst stock market crash that is mathematically possible occurred every couple of years. Very few portfolios will survive such havoc. The fact that you're getting 35% success rate could even be a cause for celebration.

Most retirement calculators restrict Monte Carlo sims to use data from about 1975 to 2020, which yields considerably better results because the year-to-year changes are relatively small. If you open up the "Historical Data" section of the Configuration and restrict your calculations to those years you should see an improvement in your success rates. But again, take the results with a grain of salt. There are only 500 cycles run with Monte Carlo right now, and it needs to be 100k - 250k.

When I get around to working on Monte Carlo again, the default historical range will likely be shrunken so that the results aren't so grim.

Also, I can't remember if I had saw this before or not but love the start year,end year, current year columns on the CSV output!

Thank you! I actually added some of those columns based on your feedback a couple of months back :)

This calculator is great.  Thank you.

🙏 I'm glad you find it useful!

Another feature that I would like is the ability to decrease my withdrawal rate once social security kicks in. I could not figure out how to calculate that in. Otherwise this is a great calculator.

The calculator actually behaves this way, although the charts are misleading right now. It's on my todo list to make this behavior more clear in the charts. You can read more about the behavior of Additional Income in the guides. Check out the section "Effect on Withdrawals".

Ideally, the guides should be able to answer any and all questions related to how the calculations work. If you have a question, I'd recommend checking out the guides to see if it's answered in there. If you can't find the answer in the guides, let me know and I'll add it!
« Last Edit: June 07, 2020, 10:28:00 AM by jamesplease »

themagicman

  • Bristles
  • ***
  • Posts: 394
  • Age: 30
  • Location: Atlanta, GA
Re: FI Calc: a new retirement calculator
« Reply #63 on: June 07, 2020, 01:45:05 PM »
1. It is not clear to me how I can can view sections on the chart. For example, if I want the chart to show me range and the percentage of 1 standard deviation (I am assuming those lines are standard deviation) from the mean and 2 staandard deviations from the mean, it is difficult to get the system to do this. It has done it for me before but I cannot figure it out. It would be nice to click in there and it select that. When I hover over it, it is showing me the whole 90% range. Does that make sense? If not, I can reword with screen shots

The chart shows you the section that you're hovered over. So if you move your mouse (or finger) horizontally in the chart, you should see details of whatever specific item you're hovered on. It's the same interaction model used for all of the charts in the app. Let me know if that helps!


See the attached. I am trying to find the numbers that are in between the red x and the percentage of withdraw that fall in that range. Does that make sense and is that possible?


Also changing the monte carlo do 1975 on changed it significantly. So is how the calculator is working it take a random year in the range and then chooses another random year in the range and so on? Is that why it was a low chance before? Because in reality poor years are generally followed by good years but in the monte carlo it is not? Does it take into account the mean returns or the standard deviation of the history returns or does it just choose a random actual year? Also I guess 1975 on is not as volatile and that is why it gives a higher percentage of success then the whole thing?

travel2020

  • 5 O'Clock Shadow
  • *
  • Posts: 29
Re: FI Calc: a new retirement calculator
« Reply #64 on: June 07, 2020, 03:05:55 PM »
@jamesplease thanks for putting this together. Just played around with it. Easy to use and had all the various variables I could think of factoring in.

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #65 on: June 07, 2020, 03:38:33 PM »
See the attached. I am trying to find the numbers that are in between the red x and the percentage of withdraw that fall in that range. Does that make sense and is that possible?

Ah okay, that makes sense! Thanks for taking the time to share that screenshot. I definitely misunderstood you before, but I think I get what youíre asking now.

Unfortunately thatís information is not currently possible from the graphs. I plan to add a tabular version of the data from the graphs, which would then allow you to see the all of the different quantile values numerically.

Also changing the monte carlo do 1975 on changed it significantly. So is how the calculator is working it take a random year in the range and then chooses another random year in the range and so on? Is that why it was a low chance before?

Yup, thatís right.

Because in reality poor years are generally followed by good years but in the monte carlo it is not?

Note: this explanation may actually be inaccurate. Leaving it here for posterity but the subsequent posts clarify it.

The difference is that Monte Carlo with the full historical time range can Ė and repeatedly does Ė produce a crash far, far worse than any crash in history.

Does it take into account the mean returns or the standard deviation of the history returns or does it just choose a random actual year?

It picks a random actual year. No means, standard deviations, or any other statistical variations are factored in. There are different kinds of Monte Carlo algorithms, and this is the one used by all of the popular Monte Carlo retirement calculators that I looked at, which is why I went with it.

Also I guess 1975 on is not as volatile and that is why it gives a higher percentage of success then the whole thing?

Update: it turns out this explanation I gave is probably wrong (see the following posts). Iím leaving it up here for posterityís sake but itís likely inaccurate!

The reason the success rate is higher is that the difference between the stock market price from 1975-2020 is a lot smaller than the difference between the stock market price from 1871-2020, so simulated crashes are less extreme, and, perhaps, more realistic.

To put some numbers on it:

Shillerís data has the S&P 500 valued at 4.44 in 1871. In 2020, it is 2930. So using the full data set can simulate situations where your portfolio drops to 0.1% of its value from one year to the next. Will any withdrawal strategy survive changes that are this extreme? No, probably not. Is it realistic? Again, probably not (in my opinion).

The value of the S&P 500 is around the $80ish range, which is ~3% of the value in 2020. This can still lead to some pretty serious drops, but comparatively itís less extreme, and probably more realistic, than using the full data set. There are also less likely to be fewer drops since the data is less spread out, so the standard deviation is smaller.

Let me know if that helps clarify whatís going on any better!

@jamesplease thanks for putting this together. Just played around with it. Easy to use and had all the various variables I could think of factoring in.

Hey, thanks! Iím glad you found it easy to use :D
« Last Edit: June 07, 2020, 09:39:55 PM by jamesplease »

secondcor521

  • Magnum Stache
  • ******
  • Posts: 3068
  • Age: 51
  • Location: Boise, Idaho
  • Big cattle, no hat.
    • Age of Eon - Overwatch player videos
Re: FI Calc: a new retirement calculator
« Reply #66 on: June 07, 2020, 05:04:01 PM »
Shillerís data has the S&P 500 valued at 4.44 in 1871. In 2020, it is 2930. So using the full data set can simulate situations where your portfolio drops to 0.1% of its value from one year to the next. Will any withdrawal strategy survive changes that are this extreme? No, probably not. Is it realistic? Again, probably not (in my opinion).

If I understand your description accurately, I would be very surprised if your Monte Carlo implementation is similar to others out there.

Let's take your example numbers from above, and further assume the market went up 5% in 1871 and 10% in 2020.

If your Monte Carlo simulation were to randomly pick 2020 and 1871 as the first and second years of the simulation, I think most others out there would take the initial portfolio value and increase it by 10% then by 5%.  What you seem to be saying is that you at some point would multiply by something like 4.44 / 2930.

To say the same thing a different way, a typical Monte Carlo simulation will have a table by year of annual investment returns, and it will repeatedly sample randomly from that yearly table.  You seem to be sampling from a table by year of S&P levels, which is, uh, waaaay wrong.

What I have seen typically is that Monte Carlo simulations tend to be about 25% to 50% (depending on the exact details of the implementation) more conservative than historical models.  So if a historical model says that a 4% WR is safe, then a Monte Carlo model with equivalent inputs will say that a 2% or 3% WR is safe.

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #67 on: June 07, 2020, 05:43:23 PM »
If I understand your description accurately, I would be very surprised if your Monte Carlo implementation is similar to others out there.

Let's take your example numbers from above, and further assume the market went up 5% in 1871 and 10% in 2020.

If your Monte Carlo simulation were to randomly pick 2020 and 1871 as the first and second years of the simulation, I think most others out there would take the initial portfolio value and increase it by 10% then by 5%.  What you seem to be saying is that you at some point would multiply by something like 4.44 / 2930.

To say the same thing a different way, a typical Monte Carlo simulation will have a table by year of annual investment returns, and it will repeatedly sample randomly from that yearly table.  You seem to be sampling from a table by year of S&P levels, which is, uh, waaaay wrong.

What I have seen typically is that Monte Carlo simulations tend to be about 25% to 50% (depending on the exact details of the implementation) more conservative than historical models.  So if a historical model says that a 4% WR is safe, then a Monte Carlo model with equivalent inputs will say that a 2% or 3% WR is safe.

Now that I think about it, the way youíre describing must be how FI Calc is doing the Monte Carlo calculation, as FI Calc only uses % changes, like you describe, and never uses absolute stock price values. So the explanation given in my last post must be wrong. Sorry about that!

It is odd, though, that the results would be so negative when including data back to 1871. Iíll need to dig more into why that is.

Whatever the reason, I know that thereís a lot of work to do with the Monte Carlo side of things. Itís nowhere near the point that it needs to be for folks to use it for evaluating portfolios, which is why it isnít a public feature. Itís probably best to just ignore it for now.

Itíll likely be a couple of weeks before I have time to really dig into it to make sure that itís behaving as expected. And once itís ready there will be an accompanying page in the guides that explain the algorithm in detail ✌️
« Last Edit: June 07, 2020, 05:55:57 PM by jamesplease »

highlandterrier

  • 5 O'Clock Shadow
  • *
  • Posts: 80
Re: FI Calc: a new retirement calculator
« Reply #68 on: July 03, 2020, 11:45:27 AM »
This is really exceptionally good, well done ! The options for dynamically changing allocation , the floor and ceiling withdrawal amounts and different withdrawal strategies are very useful, and the UI is spot on given it's something that could be over-complicated.

One observation that would make it even more useful for me, the 1/n and VPW withdrawal strategies have a target amount remaining. But the simulation counts a success as ending up with >0 as opposed to > target amount remaining. Using the target amount would make this simulation more meaningful for me as the target amount does not really matter.

Anyway, well done, and thank you, you should be proud of this.

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #69 on: July 06, 2020, 12:08:53 AM »
This is really exceptionally good, well done ! The options for dynamically changing allocation , the floor and ceiling withdrawal amounts and different withdrawal strategies are very useful, and the UI is spot on given it's something that could be over-complicated.

Thanks so much for the kind words! It was tough to strike a balance between making the app feel approachable while still supporting a lot of different options. I'm glad I was at least somewhat successful :)

One observation that would make it even more useful for me, the 1/n and VPW withdrawal strategies have a target amount remaining. But the simulation counts a success as ending up with >0 as opposed to > target amount remaining. Using the target amount would make this simulation more meaningful for me as the target amount does not really matter.

This is a great idea. One of the upcoming features will allow you to tweak those values in the analysis, so that you could specify exactly that. A little longer ways out is a more complete custom analysis system.

Anyway, well done, and thank you, you should be proud of this.

Thanks so much! I put a ton of effort into it, and I'm glad it's proving useful for some folks :)

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #70 on: July 22, 2020, 12:41:51 AM »
PTF

As a color confused person, a more distinctive difference between the not-so-good and the awful start years would be nice.

Maybe a brighter not-so-good notation.  This might also help in mobile usage in brighter ambient light.

@markbike528CBX , it's been some time, but I've been working on an update to FI Calc to help folks with color vision deficiencies. If you have time, I'd love to know if you find these colors to be an improvement.

To try it out:

1. Navigate to https://staging.ficalc.app (note: this is different from the usual FI Calc URL because this feature isn't deployed to production yet)
2. In the configuration panel, you'll see "User Preferences" at the bottom
3. Select the color deficiency that you have (based on your description I'd guess it's either protanopia or deuteranopia)
4. Click "OK"
5. View the new colors under "Simulations by Start Years"

I haven't updated all of the colors in the app, yet; only the ones under Simulations by Start Year will be different.

« Last Edit: July 22, 2020, 12:31:54 PM by jamesplease »

markbike528CBX

  • Handlebar Stache
  • *****
  • Posts: 1411
  • Location: the Everbrown part of the Evergreen State (WA)
Re: FI Calc: a new retirement calculator
« Reply #71 on: July 22, 2020, 03:43:49 PM »
PTF

As a color confused person, a more distinctive difference between the not-so-good and the awful start years would be nice.

Maybe a brighter not-so-good notation.  This might also help in mobile usage in brighter ambient light.

@markbike528CBX , it's been some time, but I've been working on an update to FI Calc to help folks with color vision deficiencies. If you have time, I'd love to know if you find these colors to be an improvement.

To try it out:

1. Navigate to https://staging.ficalc.app (note: this is different from the usual FI Calc URL because this feature isn't deployed to production yet)
2. In the configuration panel, you'll see "User Preferences" at the bottom
3. Select the color deficiency that you have (based on your description I'd guess it's either protanopia or deuteranopia)
4. Click "OK"
5. View the new colors under "Simulations by Start Years"

I haven't updated all of the colors in the app, yet; only the ones under Simulations by Start Year will be different.

Personally, I like the protanopia setting the most.

Setting -- comments
none- orange and reds don't distinguish well.    It may be the oranges are green, my color-visioned wife chimed in, but I forgot the exact colors she stated.
protanopia -- OK for me
deuteranopia -- pretty much the same as protanopia
tri-tanopia -- red and blacks too close for ease of distinguishing.

YMMV,   in fact Your Mileage Will Vary   :-)

Thanks for considering us color confused people!

test conditions to date:
         large 40" UHD monitor in dim room.
« Last Edit: July 22, 2020, 03:45:52 PM by markbike528CBX »

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #72 on: July 22, 2020, 05:01:50 PM »
Personally, I like the protanopia setting the most.

Setting -- comments
none- orange and reds don't distinguish well.    It may be the oranges are green, my color-visioned wife chimed in, but I forgot the exact colors she stated.
protanopia -- OK for me
deuteranopia -- pretty much the same as protanopia
tri-tanopia -- red and blacks too close for ease of distinguishing.

YMMV,   in fact Your Mileage Will Vary   :-)

Thanks for considering us color confused people!

test conditions to date:
         large 40" UHD monitor in dim room.

Thanks so much for trying this out, and I'm glad the protanopia/deuteranopia settings are an improvement for you.

I'm going to go through and update the rest of the colors in the app so that other things in the app should be more legible, too, and then I'll get this released. Thanks again @markbike528CBX !

markbike528CBX

  • Handlebar Stache
  • *****
  • Posts: 1411
  • Location: the Everbrown part of the Evergreen State (WA)
Re: FI Calc: a new retirement calculator
« Reply #73 on: July 22, 2020, 10:42:33 PM »
Personally, I like the protanopia setting the most.

Setting -- comments
none- orange and reds don't distinguish well.    It may be the oranges are green, my color-visioned wife chimed in, but I forgot the exact colors she stated.
protanopia -- OK for me
deuteranopia -- pretty much the same as protanopia
tri-tanopia -- red and blacks too close for ease of distinguishing.

YMMV,   in fact Your Mileage Will Vary   :-)

Thanks for considering us color confused people!

test conditions to date:
         large 40" UHD monitor in dim room.

Thanks so much for trying this out, and I'm glad the protanopia/deuteranopia settings are an improvement for you.

I'm going to go through and update the rest of the colors in the app so that other things in the app should be more legible, too, and then I'll get this released. Thanks again @markbike528CBX !

I would have liked to test it on my mobile devices, but my WiFi router died, so only the big screen got tested.   Maybe in a week or so when I get the new router.

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #74 on: July 22, 2020, 11:08:59 PM »
I would have liked to test it on my mobile devices, but my WiFi router died, so only the big screen got tested.   Maybe in a week or so when I get the new router.

I'd appreciate that if you have time. I just deployed this update so you can view these settings at the regular https://calculator.ficalc.app URL now.

markbike528CBX

  • Handlebar Stache
  • *****
  • Posts: 1411
  • Location: the Everbrown part of the Evergreen State (WA)
Re: FI Calc: a new retirement calculator
« Reply #75 on: July 23, 2020, 03:35:29 PM »
I would have liked to test it on my mobile devices, but my WiFi router died, so only the big screen got tested.   Maybe in a week or so when I get the new router.

I'd appreciate that if you have time. I just deployed this update so you can view these settings at the regular https://calculator.ficalc.app URL now.

Sure, I'll try to remember. I'm FIRed so usually nothing is pressing enough to be remembered.    It is now on my paper calendar.

I forgot how much we use the WiFi, so I just had standard (slow) shipping.  Dang

Simpli-Fi

  • Stubble
  • **
  • Posts: 188
Re: FI Calc: a new retirement calculator
« Reply #76 on: July 25, 2020, 07:16:49 AM »
one thing no calculator I've seen does is Alternative investment such as Real Estate.  It would be nice to see equity integrated into asset allocation somehow.

INFO: how RE market is trending vs equity growth...enables making informed decision about when/if to sell.

Dicey

  • Senior Mustachian
  • ********
  • Posts: 13758
  • Age: 62
  • Location: NorCal
Re: FI Calc: a new retirement calculator
« Reply #77 on: July 27, 2020, 10:24:59 AM »
Another PTF so I can find this later. Thanks!

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #78 on: July 27, 2020, 10:45:29 AM »
one thing no calculator I've seen does is Alternative investment such as Real Estate.  It would be nice to see equity integrated into asset allocation somehow.

INFO: how RE market is trending vs equity growth...enables making informed decision about when/if to sell.

Presently, real estate can be integrated using the Additional Income feature of FI Calc. There's more info about how to do that here.

If additional features other than the ones described there would be useful to you, let me know! I'd love to make the calculator have even more robust support for RE.

Sure, I'll try to remember. I'm FIRed so usually nothing is pressing enough to be remembered.    It is now on my paper calendar.

Ha, no problem at all if you don't remember, you've already been a huge help. Also, it's amazing that you're FIRE'd! It'll be awhile longer for me :)

Another PTF so I can find this later. Thanks!

✌️

DeskJockey2028

  • Pencil Stache
  • ****
  • Posts: 758
  • Aiming for 2028
Re: FI Calc: a new retirement calculator
« Reply #79 on: July 28, 2020, 08:39:00 AM »
This... this is amazing! Thanks! I'll be spending some time for sure in the next month or so exploring this calculator!

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #80 on: July 28, 2020, 02:41:43 PM »
This... this is amazing! Thanks! I'll be spending some time for sure in the next month or so exploring this calculator!

Nice! Feel free to post any suggestions / questions / feature requests in this thread if you have any :)

Mrs Brightside

  • 5 O'Clock Shadow
  • *
  • Posts: 43
  • Respect science Respect nature Respect each other
Re: FI Calc: a new retirement calculator
« Reply #81 on: August 01, 2020, 01:09:43 PM »
Really clean interface, I enjoyed using the calculator.

One suggestion so far that I didn't find: Ability to execute the allocation glidepath faster. All of the functions are pretty slow. Maybe a user input for # of years until the transition is complete? For example, ERN's blog suggests moving from 60/40 to 90/10 over 10 years is better than increasing over 20 or 30 years.

Also a question - I didn't follow carefully all of the monte carlo discussion above. Is that a beta feature? In other words, if I'm using the currently published calculator it should not be monte carlo. Rather it should be similar to cfiresim (will run your scenario for every 30yr period in history). Is that correct?

Keep up the good work!

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #82 on: August 01, 2020, 03:31:41 PM »
Really clean interface, I enjoyed using the calculator.

Thanks!

One suggestion so far that I didn't find: Ability to execute the allocation glidepath faster. All of the functions are pretty slow. Maybe a user input for # of years until the transition is complete? For example, ERN's blog suggests moving from 60/40 to 90/10 over 10 years is better than increasing over 20 or 30 years.

Thanks for the suggestion! I've taken a note to look more into this ✌️

Also a question - I didn't follow carefully all of the monte carlo discussion above. Is that a beta feature? In other words, if I'm using the currently published calculator it should not be monte carlo. Rather it should be similar to cfiresim (will run your scenario for every 30yr period in history). Is that correct?

Keep up the good work!

Your understanding is correct. Monte Carlo is in development, and is not available in the calculator. So the calculations currently use a similar approach to what you'd find in cFIREsim.

markbike528CBX

  • Handlebar Stache
  • *****
  • Posts: 1411
  • Location: the Everbrown part of the Evergreen State (WA)
Re: FI Calc: a new retirement calculator
« Reply #83 on: August 03, 2020, 10:34:33 AM »
Hi, I'm back and now have WiFi in the house


Setting -- large 40" UHD monitor in dim room. Mac Pro(2013, trashcan)
none- orange and reds don't distinguish well.    It may be the oranges are green, my color-visioned wife chimed in, but I forgot the exact colors she stated.
protanopia -- OK for me
deuteranopia -- pretty much the same as protanopia
tri-tanopia -- red and blacks too close for ease of distinguishing.


Setting -- iPhone 5s iOS 12.4.7 - dim screen
none- red and blacks  pretty close  but distinguishable, red and greens are actually better distinguished compared to bright screen
protanopia -- OK for me
deuteranopia -- pretty much the same as protanopia
tri-tanopia -- red and blacks  much too close


Setting -- iPhone 5s - pretty bright screen
none-  looks pretty good
protanopia -- OK for me
deuteranopia -- pretty much the same as protanopia
tri-tanopia -- red and blacks too close for _ease_ of distinguishing.


ipad2  iOS 9.3.5
Simply doesn't work
Black screen comes up, a small scroll indicator on right side, but that's it.


IPad Pro 13" (2017) bright screen
none-  looks pretty good
protanopia -- OK for me
deuteranopia -- pretty much the same as protanopia
tri-tanopia -- red and blacks too close for _ease_ of distinguishing.


IPad Pro 13" (2017) dim screen
none-  reds and blacks are really close
protanopia -- OK for me
deuteranopia -- pretty much the same as protanopia
tri-tanopia -- red and blacks too close for _ease_ of distinguishing.



This whole thing brings back a job I did, where I had to physically touch a screen and see what happened and see what colors came up.
I then had to write out in long hand, "white", "red", "blue" on a table in the worksheet paper what I had seen.
So I can confirm that the local system screens at Barakah Units 1,2,3,4 in the United Arab Emirates will be operable by color confused people. :-)
https://en.wikipedia.org/wiki/Barakah_nuclear_power_plant

mrmoonymartian

  • Bristles
  • ***
  • Posts: 287
  • Age: 38
  • Location: Brisbane
Re: FI Calc: a new retirement calculator
« Reply #84 on: August 04, 2020, 07:57:46 AM »
My formal diagnosis is red-green.  I can see, identify high saturation colors including fire engine red or Kelly green.  Pastels are a big source of confusion, although I can see the color (not just grey).
Mom found out when I was in kindergarten, and the teacher and I had a disagreement on the color of a man's overall I was drawing. I said blue, she said purple.

I still have my last school art from 9th grade. I explained my issues to that teacher, who seemed ok with it.

Since you are using a black background, just turning up the RGB or hex values on the "yellow" would be good enough for me.  YMMV

I usually only get to the first page of the Ishihara test. https://www.colour-blindness.com/colour-blindness-tests/ishihara-colour-test-plates/

On well kept, newer paper copies of the test, sometimes I can guess correctly on other pages.

Whoa, this is incredibly useful. Thank you so much for writing all of this up for me!

I'd love to get your thoughts on this idea I have: I'm not sure if one set of colors will provide the best experience for everybody, and I want this app to be the best experience for everybody. So rather than pick one set up of colors for everyone, what I can do is create different color schemes, so there will be an option that looks the best for, say, red-green. Once you set that once, the setting will be stored in your browser so you'll never need to set it again.

What do you think of that? If you agree that it could work, I can add it the app and then I'd love to get your feedback on the color choices.

Thanks again for sharing such useful feedback, @markbike528CBX !
If the colors are right next to each other, it is clearer.  The problem is if the are few and scattered various colors (success can be confusing?).
I tried grey scale and then the very dark and the OK (black) blend too much.
If you make the lightest color too light then you blend with the button color.
I checked out the colors
ca1010 (failed) and B36300 (warning/near fail) at https://htmlcolorcodes.com/color-picker/ and they have similar attributes (RGB, CYMK values)
mmmm...... not sure what I can suggest....

Not so relevant to the current discussion but a good thinking book, and a Great coffee table book for geeks is:
The Visual Display of Quantitative Information
https://www.edwardtufte.com/tufte/books_vdqi

Can confirm the colours look fine to the moderately protanomalous.


ETA: Oh, I didn't see the options...
 
Default - best for me. Orange and blue are are a bit drab, but can distinguish everything well enough.
Protanopia - mauve and pink ok to tell apart, but not as easy to use at a glance. Probably aimed at dichromats or stronger anomalous trichromats that struggle more with the default.
Deuteranopia - identical to protanopia?
Tritanopia - blue and purple too similar for me. Red also duller than default red. Not surprising; just saying for completeness.

« Last Edit: August 04, 2020, 08:22:48 AM by mrmoonymartian »

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #85 on: August 05, 2020, 11:51:44 AM »
Hi, I'm back and now have WiFi in the house

This is amazing! Thanks for the thoroughness.

I'm not surprised that the app doesn't work on the iPad 2. It takes more time and effort to support old devices/browsers, and I'm just the one person. I don't have a device with that operating system to even test on, for instance! I'm glad it worked on all of your other devices, though.

This whole thing brings back a job I did, where I had to physically touch a screen and see what happened and see what colors came up.
I then had to write out in long hand, "white", "red", "blue" on a table in the worksheet paper what I had seen.
So I can confirm that the local system screens at Barakah Units 1,2,3,4 in the United Arab Emirates will be operable by color confused people. :-)
https://en.wikipedia.org/wiki/Barakah_nuclear_power_plant

That is seriously cool. I'm going to add "FI Calc has been Quality Assured by a nuclear power plant Quality Assurance Engineer" to the homepage! (just kidding)

Can confirm the colours look fine to the moderately protanomalous.


ETA: Oh, I didn't see the options...
 
Default - best for me. Orange and blue are are a bit drab, but can distinguish everything well enough.
Protanopia - mauve and pink ok to tell apart, but not as easy to use at a glance. Probably aimed at dichromats or stronger anomalous trichromats that struggle more with the default.
Deuteranopia - identical to protanopia?
Tritanopia - blue and purple too similar for me. Red also duller than default red. Not surprising; just saying for completeness.

Thanks for the review!

I agree that the orange and blue are a bit drab. My primary goal is to ensure that the text remains legible for as many people as possible; aesthetics comes second. There's no way to make the orange brighter without sacrificing legibility, which then has a snowball effect on the rest of the color palette. I'd love for it to be brighter, though. Maybe one of these days I'll find some new colors that can brighten it up without sacrificing the color contrast.

markbike528CBX

  • Handlebar Stache
  • *****
  • Posts: 1411
  • Location: the Everbrown part of the Evergreen State (WA)
Re: FI Calc: a new retirement calculator
« Reply #86 on: August 06, 2020, 12:40:10 AM »
.......
This whole thing brings back a job I did, where I had to physically touch a screen and see what happened and see what colors came up.
I then had to write out in long hand, "white", "red", "blue" on a table in the worksheet paper what I had seen.
So I can confirm that the local system screens at Barakah Units 1,2,3,4 in the United Arab Emirates will be operable by color confused people. :-)
https://en.wikipedia.org/wiki/Barakah_nuclear_power_plant

That is seriously cool. I'm going to add "FI Calc has been Quality Assured by a nuclear power plant Quality Assurance Engineer" to the homepage! (just kidding)

My last actual title was "Principal Field Service Engineer" , but I'm a chemist, dammit.

That job was an off-season filler.  One other guy there had also been in fieldwork so we would trade true field stories that none of the computer testing people believed.

Simpli-Fi

  • Stubble
  • **
  • Posts: 188
Re: FI Calc: a new retirement calculator
« Reply #87 on: August 07, 2020, 02:06:37 PM »
one thing no calculator I've seen does is Alternative investment such as Real Estate.  It would be nice to see equity integrated into asset allocation somehow.

INFO: how RE market is trending vs equity growth...enables making informed decision about when/if to sell.

Presently, real estate can be integrated using the Additional Income feature of FI Calc. There's more info about how to do that here.



If additional features other than the ones described there would be useful to you, let me know! I'd love to make the calculator have even more robust support for RE.

Sure, I'll try to remember. I'm FIRed so usually nothing is pressing enough to be remembered.    It is now on my paper calendar.

Ha, no problem at all if you don't remember, you've already been a huge help. Also, it's amazing that you're FIRE'd! It'll be awhile longer for me :)

Another PTF so I can find this later. Thanks!

✌️

Additional income is never the problem...all calculators have that.  How do you show equity of real estate in your asset allocation (alternative investment) is the grand question?

make sense?  for instance say I have a $1M investment portfolio with $200k in REITs (20%) and then another $500k equity in a physical investment property (appraisal - mortgage); how to add the $500k into the alternatives asset in the allocation so it displays "full" portfolio not just traditional investment.  $1.5M with 47% in Alternatives

Now I can see I have too much in RE and need to rebalance.  Also nice to trend RE appreciation vs the other markets
« Last Edit: August 07, 2020, 02:10:43 PM by Simpli-Fi »

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #88 on: August 10, 2020, 01:04:03 PM »
Additional income is never the problem...all calculators have that.  How do you show equity of real estate in your asset allocation (alternative investment) is the grand question?

make sense?  for instance say I have a $1M investment portfolio with $200k in REITs (20%) and then another $500k equity in a physical investment property (appraisal - mortgage); how to add the $500k into the alternatives asset in the allocation so it displays "full" portfolio not just traditional investment.  $1.5M with 47% in Alternatives

Now I can see I have too much in RE and need to rebalance.  Also nice to trend RE appreciation vs the other markets

Ah okay, I understand what you're asking now. Thanks for the clarification.

This is something I've thought about, but so far I've decided against it because I'm not quite sure how it would work within the calculator. Physical real estate is quite different from the asset types listed. For one, it's not as straightforward to "rebalance" a portfolio that includes real estate as it is to rebalance, say, bonds and stocks. For example, there are additional fees like closing costs that are difficult to estimate.

One of the features of FI Calc is that the way that the numbers transform year over year are based on historical data. Historical data for homes in every market doesn't exist as far as I'm aware. Folks could enter their own values for growth and taxes, but that is at odds with the main value prop of a calculator like FI Calc. The cycles in the historical data are where the interesting results come from; analyses with static values are useful, but using historical data can be useful for other reasons. A goal of mine is to minimize the number of "static values" in the calculator, and rely on historical data as much as possible.

So yeah, I'd love to add physical real estate, but I first need to think of a way to include it in a way that makes sense. Do you have a more specific idea of how the algorithm for incorporating physical real estate might work?

Regarding REITs: I'm interested in adding it, and this is easier to add than physical property. I just need the historical data.

Xlar

  • Bristles
  • ***
  • Posts: 251
Re: FI Calc: a new retirement calculator
« Reply #89 on: August 16, 2020, 12:38:17 PM »
Is https://ficalc.app down for everyone or just for me?

Simpli-Fi

  • Stubble
  • **
  • Posts: 188
Re: FI Calc: a new retirement calculator
« Reply #90 on: August 16, 2020, 01:57:25 PM »
its down.



So yeah, I'd love to add physical real estate, but I first need to think of a way to include it in a way that makes sense. Do you have a more specific idea of how the algorithm for incorporating physical real estate might work?

Regarding REITs: I'm interested in adding it, and this is easier to add than physical property. I just need the historical data.
to answer the rebalance...if you are too heavy/light in "alternatives" you would sell/buy REITs or some other form of traded alternative; not necessarily physical RE.  The physical RE is just an anchor that appreciates and builds equity or falls the other way.

what bugs me with the mints or PC's out there...my equity adds to my networth but doesn't show a percentage of my total portfolio, so all the assumptions about future growth are skewed not accounting for investment properties.

Zillow seems to have an API that looks up Zestimate for any address, most property tax records are public...just a matter of time before someone finds a way to collect them in an automated fashion.

JMS

  • 5 O'Clock Shadow
  • *
  • Posts: 16
Re: FI Calc: a new retirement calculator
« Reply #91 on: August 17, 2020, 04:19:05 AM »
Iíve just tried to access the page using the link in the origina post and get Error 526.  Anyone else having the same problem?

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #92 on: August 17, 2020, 02:20:47 PM »
to answer the rebalance...if you are too heavy/light in "alternatives" you would sell/buy REITs or some other form of traded alternative; not necessarily physical RE.  The physical RE is just an anchor that appreciates and builds equity or falls the other way.

Makes sense. I've added REITs to my list of things to look into ✌️



Sorry about FI Calc being down! I'm working on getting it running again now. In the meantime, you can access the calculator directly through the URL https://calculator.ficalc.app. The issue is with the calculator guide website.

Update: this is a problem with GitHub (where the guide site is hosted). They're working on fixing it now.
« Last Edit: August 17, 2020, 02:26:37 PM by jamesplease »

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #93 on: August 18, 2020, 09:48:44 AM »
ficalc.app is fixed! Sorry for the downtime.

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #94 on: September 20, 2020, 01:47:02 PM »
A couple of updates over the past few days:

- Additional income/withdrawals can now be set as lasting indefinitely. This can be useful for, say, a pension or SS.
- The misleading "Target end portfolio value" has been removed from VPW (VPW does not allow you to set a target end portfolio value). It can still be found under CVPW, but it has been renamed to "PMT Future Value," which is a more accurate label.
- There's an update to the definition of success for Maximize Spend strategies like VPW: sims will count as failed if the final year withdrawal is smaller than your target minimum withdrawal. i.e.; if you want to spend $30k but your portfolio only has $20k in that final year, then that would count as a fail.

highlandterrier

  • 5 O'Clock Shadow
  • *
  • Posts: 80
Re: FI Calc: a new retirement calculator
« Reply #95 on: September 27, 2020, 02:22:18 AM »
Thanks for amending the definition of success, will be very useful !

Noticed a bug in this release with the Maximum Annual Withdrawal for 1/N and VPW scenarios. If you use the scenario linked below it has a 97.86% success rate, but if you tick the box to limit the maximum withdrawal per year the success rate reduces to 67.14%. Cannot be correct as limiting your withdrawals cannot decrease your success.

 https://calculator.ficalc.app?additionalIncome=%5B%5D&additionalWithdrawals=%5B%5D&bondsFees=0.05&bondsFinalRatio=15&bondsInitialRatio=15&cashFees=0&cashFinalRatio=5&cashGrowth=1.5&cashInitialRatio=5&changeAllocationsOverTime=false&equitiesFees=0.04&equitiesFinalRatio=80&equitiesInitialRatio=80&initialPortfolioValue=450000&maxWithdrawalLimit=31000&maxWithdrawalLimitEnabled=false&minWithdrawalLimit=30000&minWithdrawalLimitEnabled=true&numberOfYears=10&oneOverNTargetPortfolio=0&portfolioRebalanceEquation=linear&rebalance=true&rebalanceFrequency=1&retirementStartingAge=60&withdrawalStrategyName=oneOverN


TomTX

  • Magnum Stache
  • ******
  • Posts: 4232
  • Location: Texas
Re: FI Calc: a new retirement calculator
« Reply #96 on: September 27, 2020, 11:05:19 AM »
Thanks for setting this up! I'm enjoying it.

I do have a question on the additional income streams. Could you confirm my understanding of how the "no inflation" choice works?

Let's say I'm retiring today and in 10 years I'll start receiving a pension of $35k/year with no inflation adjustments either between now and when I can draw, or after I start drawing. Fixed dollar amount indefinitely. Is that how "no inflation" calculates things?

jamesplease

  • 5 O'Clock Shadow
  • *
  • Posts: 55
  • Age: 30
  • Location: California
Re: FI Calc: a new retirement calculator
« Reply #97 on: September 28, 2020, 11:37:47 PM »
Thanks for setting this up! I'm enjoying it.

I'm glad to hear it!

I do have a question on the additional income streams. Could you confirm my understanding of how the "no inflation" choice works?

Let's say I'm retiring today and in 10 years I'll start receiving a pension of $35k/year with no inflation adjustments either between now and when I can draw, or after I start drawing. Fixed dollar amount indefinitely. Is that how "no inflation" calculates things?

Your understanding is correct. That is how "no inflation" calculates things, and the pension that you're describing sounds to me like a "no inflation" additional income stream.

Feel free to ask any other Q's if something isn't clear. Cheers!