Can you use FIRECalc to forecast your portfolio?

[iamjameswalters]

Hi there,

My name is James, I'm brand new here, and I wasn't sure which forum this would best fit under (I first posted it in Other...sorry). So, I reposted it over here, which seems to be the best fit.

I'm 22, and my wife and I are looking into retirement planning. I found FIRECalc, and it's a tremendous resource for forecasting how long my nest egg could last me with different withdrawals and portfolio scenarios. The historical data behind it is impressive. But, it made me wonder--can this same historical data be employed to give a more realistic investment forecast? I know FIRECalc is useful for lots of calculations, but it can't do everything. Is this within the parameters of the calculator to do? To try this, I set the first page parameters to 0 spending, 0 portfolio and 30 years. Then, on the Not Retired page, I set the retirement date to 2046 (30 years out), and set my yearly contribution to $15,000 (see below).

So here's my test case. I tried to use some simple numbers. Let's assume a $100,000 annual income. We'll contribute 15% (considered an aggressive savings rate by most) or $15,000 per year for thirty years, and we'll invest it according to FIRECalc's default Couch Potato portfolio (75% stocks/equities, 25% bonds/fixed income). I'm going to assume an average return of 8% off of that. Running that through a compound interest calculator, we end up with $1,835,188. Running it through FIRECalc, the average ending balance is $1,226,254. Following the 4% rule, the first calculation would yield a $73,407 annual withdrawal, the FIRECalc number $49,050. Those dollars are inflated by thirty years; guestimating back to today's dollars from a 2.75% inflation rate, I'd estimate them to be about $32,500 and $21,800, respectively. Both of these pale in comparison to the $100,000 income we're replacing (we're looking at a 22-33% replacement factor, compared to the 60-80% replacement factor most retirement gurus suggest).

So, help me out here: I'm not a strong math guy. Have I used some bad data? Have I messed up some numbers somewhere? Did I miscalculate? Or does FIRECalc not have the data to produce this kind of a prediction? Did I try to get FIRECalc to do something it really can't do? If that's the case, then if it can't forecast how retirement investments will grow*, then how can it forecast how retirement investments will be depleted?

If this is right, then have we severely underestimated our retirement savings rates?

This has been an interesting study, I'm looking forward to any insight you all can bring. I haven't been studying this for long, and I know that around here there are folks who are already living the early retirement dream. Your perspectives will be grounded in some real world experience, which I would welcome.

Thanks guys!



* - According to historical data sets. Obviously, FIRECalc can't predict the future, but assuming all the different investment cycles that it tests through.

 

[galeno]

I use fire-calc all the time. Vanguard's retirement calculator is simple and excellent also.

 

[NW-Bound]

...I set the first page parameters to 0 spending, 0 portfolio and 30 years. Then, on the Not Retired page, I set the retirement date to 2046 (30 years out), and set my yearly contribution to $15,000 (see below).

So here's my test case. I tried to use some simple numbers. Let's assume a $100,000 annual income. We'll contribute 15% (considered an aggressive savings rate by most) or $15,000 per year for thirty years, and we'll invest it according to FIRECalc's default Couch Potato portfolio (75% stocks/equities, 25% bonds/fixed income). I'm going to assume an average return of 8% off of that. Running that through a compound interest calculator, we end up with $1,835,188. Running it through FIRECalc, the average ending balance is $1,226,254...

Yes, it says that "The lowest and highest portfolio balance at the end of your retirement was $0 to $2,317,860, with an average at the end of $1,226,254. (Note: this is looking at all the possible periods; values are in terms of the dollars as of the beginning of the retirement period for each cycle.)"

The $0 above is wrong (this bug in FIRECalc is still not fixed). The chart in the result page shows the worst case ending value of about $470,000. These are in 2016 dollars. With inflation, the number can be 2X that or higher.

... Following the 4% rule, the first calculation would yield a $73,407 annual withdrawal, the FIRECalc number $49,050. Those dollars are inflated by thirty years; guestimating back to today's dollars from a 2.75% inflation rate, I'd estimate them to be about $32,500 and $21,800, respectively. Both of these pale in comparison to the $100,000 income we're replacing (we're looking at a 22-33% replacement factor, compared to the 60-80% replacement factor most retirement gurus suggest)...

FIRECalc does everything in today's dollars. You do not have to compensate for inflation, as it uses actual historical inflation in conjunction with actual investment returns.

If at the end of 30 years, you end up with $1.2M in today's dollars, then another FIRECalc run will tell that you can spend $49,000 for 30 years of retirement. Again, everything is in today's dollars. The problem is what if you end up with the worst case of $470,000?

PS. Note that the worst case of $470,000 divided by $15,000/year is 31 years. This means that in the worst case, your savings just barely keep up with inflation, and there's no real investment return.

 

[mrftexas]

To compare with your compounding example, in FIRECalc, you should change your portfolio to one that has a constant 8% growth rate (the third choice on the page) for both fixed and market growth. Set inflation = 0 and set your investment fees to zero. If you then rerun it, you'll get a future value of $1,835,188, matching your value.

 

[iamjameswalters]

FIRECalc does everything in today's dollars. You do not have to compensate for inflation, as it uses actual historical inflation in conjunction with actual investment returns.

If at the end of 30 years, you end up with $1.2M in today's dollars, then another FIRECalc run will tell that you can spend $49,000 for 30 years of retirement. Again, everything is in today's dollars. The problem is what if you end up with the worst case of $470,000?

PS. Note that the worst case of $470,000 divided by $15,000/year is 31 years. This means that in the worst case, your savings just barely keep up with inflation, and there's no real investment return.

Ah, gotcha. So, if I'm understanding you correctly, FIRECalc does math including inflation, but it spits out the result in today's dollars?

Also, with 1.2M yielding a $49,000 withdrawal, does that mean that the replacement rate for the $100,000 income is only 49%? That seems pretty lousy for an accelerated savings rate (15%).

Thanks for that really helpful response, I'm learning a lot here.

 

[iamjameswalters]

To compare with your compounding example, in FIRECalc, you should change your portfolio to one that has a constant 8% growth rate (the third choice on the page) for both fixed and market growth. Set inflation = 0 and set your investment fees to zero. If you then rerun it, you'll get a future value of $1,835,188, matching your value.

I tried this, and you're right! Thanks for showing me that. Let me pose this question, however: Running that calculation again with a 3.0% inflation rate yields a result of $1,049,777. So is that what the $1,835,188 will be worth in today's dollars? The 4% rule gives us a withdrawal of about $42,000. This would be a 42% replacement rate of the $100,000 income. Still shy of the 60-80% replacement many are shooting for. Did I run those numbers right?

 

[iamjameswalters]

Alright, so I tried reverse engineering this a little bit. Let's begin with that $100,000 income, and the assumption that I'm shooting for an 80% replacement rate, which would be $80,000. The 4% rule tells me I'll need a nest egg of $2,000,000 to make that happen. In order to wind up with an average of just over $2,000,000, investing across 30 years with the same default 75/25 stocks/bonds portfolio, you would need to invest $24,500 per year, for an aggressive 24.5% savings rate. The annual contributions drop to about $20,500 if you invest in 100% stocks/equities, still over 20% savings rate.

So, if this model is true to scale, then in order to secure an 80% replacement rate, you'd need an aggressive 20-25% savings rate across 30 years. I tested this with a $60,000 income, saving $15,000 a year (25%), which gave us the same $1,226,254 number from above, but the $49,050 withdrawal is about 81% of the $60,000 income.

So, it seems to be the case when you're comparing to the historical data, that if you start saving 30 years from retirement and invest in a portfolio of 75% stocks/25% bonds, if you want to secure an 80% replacement rate, you'll need to contribute at a 25% savings rate. Is that a sensible conclusion to draw?

 

[NW-Bound]

Ah, gotcha. So, if I'm understanding you correctly, FIRECalc does math including inflation, but it spits out the result in today's dollars?

That's correct.

This way makes a lot of sense, because you do not have to figure out what $50K in inflated dollars will buy you 15 years, 20 years, or 30 years from now. When it says you can spend $50K in today's dollars, you can be assured of maintaining the living standard.

Also, with 1.2M yielding a $49,000 withdrawal, does that mean that the replacement rate for the $100,000 income is only 49%? That seems pretty lousy for an accelerated savings rate (15%).

Of course it does not matter that you make $100K or $200K. What matters is the stash you have when you start your retirement. If you hang around here longer, you will see people mentioning the Trinity study (1998) that showed that one could safely draw 4% (and adjust for inflation every year) from his stash, and expect it to last 30 years. That is the worst case, where you go exactly broke in 30 years. More often, you would end up with some money, occasionally even more than you started out with.

The $49K from $1.2M is about 4% each year, right? This is called the SWR or Safe Withdrawal Rate. Search the Web or this forum to learn more.

By the way, I discovered this forum when I surfed the Web searching for info regarding what I could expect to get from my stash. I learned a lot hanging around here.

By the way, you are welcome.

 

[iamjameswalters]

Of course it does not matter that you make $100K or $200K. What matters is the stash you have when you start your retirement. If you hang around here longer, you will see people mentioning the Trinity study (1998) that showed that one could safely draw 4% (and adjust for inflation every year) from his stash, and expect it to last 30 years. That is the worst case, where you go exactly broke in 30 years. More often, you would end up with some money, occasionally even more than you started out with.

Right, and I was trying to plan on a 60-80% replacement income being the SWR. It sounds like you'd be saying that's unnecessary, because the SWR is the worst case scenario, but there's a decent chance you're not going to get the worst case scenario, so you can plan on starting out withdrawing more and adjust down if you have to. Am I following you here?

Is it unrealistic to try and plan for something like an 80% replacement rate to be your SWR? Should you seek to make the SWR a minimum that can take care of your needs, but withdraw more so long as the portfolio's faring well?

 

[NW-Bound]

Oh man, we have been debating this SWR topic forever. In fact, the most recent threads on this subject came about just a week or two ago.

Basically, the problem is this. Many people fear that the worst case in the past may be exceeded, or at least repeated by what is to come. The US has been through its industrialization of the 20th century, and if we look at European countries and Japan, their economies do not grow like what the US has seen. So, without portfolio growth, your WR will have to be lower.

So, how much lower? If you can just keep up with inflation, then at 3.33%WR your stash will last 30 years, and that's many posters here are using instead of the 4%.

If your portfolio growth exceeds expectation, nobody will have a problem "ratcheting" up (that is indeed a word used in a recent thread ). The harder problem is scaling down in the opposite scenario, when one is used to a plush life. Again, we have been talking about these problems forever.

 

[NW-Bound]

By the way, you do not have to do the silly 80% or 50% replacement method. Here's why.

When you make $100K, how much do you spend? What about taxes, savings, or all other expenses that you will not have in retirement? Out of $100K gross, how much do you actually spend? And then, your expenses in retirement will be even lower than that. You should no longer have a mortgage, child expenses, college tuition, etc...

I absolutely do not expect to spend even 50% of our past gross incomes in the long run, although I have exceeded that often due to "one-time" events.

PS. The healthcare cost is the big unknown and most worrisome. This wildcard is the topic of so many active threads among early retirees here. But you are young, and when you get there, who knows what it will be like.