I can’t get Portfolio Visualizer and FireCalc to agree.

Non-investing personal finance issues including insurance, credit, real estate, taxes, employment and legal issues such as trusts and wills
Post Reply
User avatar
Posts: 16
Joined: Sat Dec 20, 2014 4:05 pm

I can’t get Portfolio Visualizer and FireCalc to agree.

Post by Emiliania » Sun Jul 22, 2018 3:37 pm

I can’t get PortfolioVis and FireCalc to agree.

In some investigations to see if my personal backtest sheet has flaws, I found it agrees with Portfolio Visualizer, but not FireCalc. I’m trying to determine why the difference, and narrowed it down to a simple comparison. I get Portfolio Vis and FireCalc to agree only if there are no withdrawals, but if I add withdrawals, FireCalc gives a lot smaller final porfolio value.

To make a simple comparison, I get them to show me one backtest starting in 1972 with 100% US stock market, withdrawals at 4% starting value with inflation adjustments.

Here are the settings:
(All settings are default except those given.)

Portfolio Vizualizer: Backtest Portfolio Asset Class Allocation
Start Year 1972
Initial Amount 1,000,000
Spending 40,000
Inflation adjusted yes
Asset Allocation: 100% US Stock Market
For the Plot, be sure to click “Inflation Adjusted”

PV shows about 1.3M$ at the start of 2018

FireCalc -------------------------------------
Main Screen:
Spending 40,000
Porfolio 1,000,000
Years 45
Your Portfolio Tab
fees 0
Total Market: performance since 1972
Total Market stocks 100%
With these settings, FireCalc shows two traces, I think they are ‘72 and ‘73. The final value with the ‘72 run gives something around -0.3M$ and ‘73 is lower. (FireCalc always shows at least two traces, even if I increase the # of years beyond the years we have data for.) So that’s about 2.6M$ different than PV! :confused

I’ve been beating my head on my computer for awhile with this one. Can anybody enlighten me on why I’m getting such different results with these two models? Portfolio Visualizer will show you a table of the returns and numbers in an individual run, so I can see exactly what’s happening, and it does agree with my own calcs. But I can’t see what FireCalc is doing. I know it’s been widely used and checked out, so I think the difference may be some real difference in the model. Maybe? :?: For example, one on-line retirement model (I forget which one) adds a 2% annual spending increase due to research that retiree expenses have more housing and medical and therefore have a higher inflation rate. But I don’t believe FireCalc does this.

For those Bogleheads out there that have done a deep-dive into FireCalc, can you enlighten me? Or give me a lead on what to check? (I’ve tried checking past threads on FireCalc, but couldn’t find enlightenment there.)

Thanks much.

Post Reply