Hacker News new | past | comments | ask | show | jobs | submit login
CFIREsim 3.0 – Retirement Simulator (cfiresim.com)
73 points by LVB on Feb 13, 2022 | hide | past | favorite | 20 comments



I prefer Monte Carlo simulations. I think they do a better job of representing the possible futures based on the limited data we have from the past. But most of them are way slower than they ought to be. So I made a fast one. Just for my own use, so the UI is terrible. But it might be interesting to play around with anyway: https://james.darpinian.com/money/

It does a million simulations instantly as you drag the sliders around so that you can compare scenarios as fast as you can move your mouse. It also can optimize for any variable to answer questions like "how many years of work are needed to pay for each $10k increase in retirement budget". Someday when I have time I'll put a real UI on it.


Maybe add number inputs or change the minimum/step values in the ranges. Some of us don't even have $1000 to start with :)


FWIW I like the bare bones UX and instant feedback. One thing though: it seems to not work on Firefox (stuck on 0.0% success rate). Chrome is fine.


Thanks. Does WebGL 2 work in your Firefox? It's fast because computations are done on the GPU. It works in Firefox here on my phone.


Not sure, but I do get a WebGL warning: readPixels: Format and type RED/UNSIGNED_BYTE incompatible with this R8 attachment. (M1 Mac)


Interesting. Probably my fault, though it could be a platform specific browser bug. Hopefully one day I'll have time to look into it.


Same on Mobile Safari.


I was impressed by https://www.timelineapp.co/

I saw it on James Shack’s channel a few times, see here at 5:42 https://youtube.com/watch?v=ANjA7TEfezE&t=5m42s for an example in use.


Heh: "Retirement End Year".

Is that a synonym for "The Year You Die"? :)


Maybe we'll really luck out, and it'll be the year you get your rejuvenation treatment and go back to work.


It even says so:

"Retirement End Year - Time to get real grim here… this is when you’re expecting to not need your retirement fund anymore. Put in a different way, it is when you expect to die. There are a number of factors that play into this, but I would suggest looking at the Social Security Administrations chart here. As of the date that I wrote this post, I am a 39 year old female, and that chart tells me that on average I can expect to live 43.41 more years. So, I personally would set the Retirement End Year to 2072 (2021 + 43.41 = 2062.41 + 10 years of padding just to be sure)." -- https://www.cfiresim.com/tutorials/


Yes, but if a euphemism causes people to stay at the site and do some financial planning, what's the harm?

Ideally one would specify a probability distribution of death times.


Not necessarily. Some people only plan for a 20 year retirement, some plan for a 30 year retirement, etc.


And what do you think they're expecting to happen after the retirement?


I thought it was needless to say, but people don't immediately drop dead as soon as their planned retirement period expires. They either have to re-budget or go on things like a pension etc.


Inflation is challenging to model these days.


It's really not, but perhaps more importantly it shows why retirement income streams should respond to inflation and not be fixed.

As someone who has lived in a country with mild inflation all my life it's not as scary to me as it seems to be to US folk. It's just another variable in the formula.

(Americans seem to conflate any inflation with hyper inflation, which seems weird.)

I've come to understand though that most US retirement plans are fixed-income based, so _any_ inflation results in lower real income. Because inflation didn't exist for so long retirees were sold fixed income plans.

So, rather than ignore inflation, or hope it remains negligible, its better to plan with it in mind - to structure your retirement investments so they are effectively immune to inflation (ie their value gets higher as inflation exists.)

Incidentally inflation has equal benefits on the other side, loans are eroded in value as inflation happens, which is (to the borowwer) a good thing (and that ultimately applies to foreign debt as well).


How can you model something for which you have no real data ;) ? I guess multiplying the government provided numbers times 3 would give a somewhat accurate estimation.

Even if you trust the government numbers suddenly going from 2% inflation to 8% inflation ruins most FIRE plans which tend to have razor thin margins (so to say) on the returns they get on their investments.


Bear in mind that there is no "real data" of the future. There is only present and past data. And modeling based on that, well, let's say there are no guarantees in life.

Part of planning is understanding where the risks are, what the premises are that the model relies on and so on. Depending on inflation being 0-2% over a 30 to 40 year working-life seems optimistic.

So, in planning, some diversification is appropriate. Some part of your retirement earnings should be inflation-adjusted. For example say you get rent on a property. That rent will tend to rise following inflation. Equities, especially ones chosen for dividends will also tend to follow inflation (over the long term).

By contrast a fixed-yield item like say a bond returning 5% is not inflation linked, and leaves you vulnerable to inflation variances. If your model only relies on fixed-yield, and does not allow for variable amounts of inflation, then the model needs updating IMO.


Diversifying with some real-estate (maybe a REIT) sounds like a good approach to counter inflation.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: