@mgarbm I'm sure a lot of folks would be interested in reviews!
For other new folks like me, all these tools can be set up to have identical "errors" re: Roth conversions. Pralana addresses this in the manual, while it is the default (and only?) option for other tools. I urge everyone to read the Roth Conversion section of the manual. ????
I noticed that Boldin just did an update to their monte carlo. Just now, Boldin finally switched from using CAGR (geometric average) return estimates to Arithmetic Average return estimates for purposes of the monte carlo. As they correctly note, the cagr already includes deviation drag, so using it in monte carlo "double counts" variance; whereas an accurate monte carlo simulates a normal curve with variation equal to the standard variation around the arithmetic average. Using a CAGR in monte carlo underestimates returns, making your plan look more dangerous than it is.
I note that Pralana has done that part of monte carlo correctly since the beginning, thanks to @smatthews51 's thoughtful attention to detail. Users enter CAGR for their expected returns of asset classes (since that is how we're used to thinking of long term returns). Behind the scenes, Pralana estimates the arithmetic average from that geometric average, and uses that in the monte carlo draws.
Hats off, gentlemen.
I noticed that Boldin just did an update to their monte carlo. Just now, Boldin finally switched from using CAGR (geometric average) return estimates to Arithmetic Average return estimates for purposes of the monte carlo. As they correctly note, the cagr already includes deviation drag, so using it in monte carlo "double counts" variance; whereas an accurate monte carlo simulates a normal curve with variation equal to the standard variation around the arithmetic average. Using a CAGR in monte carlo underestimates returns, making your plan look more dangerous than it is.
I note that Pralana has done that part of monte carlo correctly since the beginning, thanks to @smatthews51 's thoughtful attention to detail. Users enter CAGR for their expected returns of asset classes (since that is how we're used to thinking of long term returns). Behind the scenes, Pralana estimates the arithmetic average from that geometric average, and uses that in the monte carlo draws.
Hats off, gentlemen.
With respect to success rates- after the shift from CAGR to AAR- in the Boldin world, it has been revealed that those scenarios with fewer accounts being analyzed have a better chance of seeing their success rates improve (and, obviously, those with relatively numerous accounts are seeing a dip- sometimes significantly- in MC success rates). Thought I'd note this in the context of the comment around variable(s) that make plans look more dangerous than they actually are.
With respect to success rates- after the shift from CAGR to AAR- in the Boldin world, it has been revealed that those scenarios with fewer accounts being analyzed have a better chance of seeing their success rates improve (and, obviously, those with relatively numerous accounts are seeing a dip- sometimes significantly- in MC success rates). Thought I'd note this in the context of the comment around variable(s) that make plans look more dangerous than they actually are.
This has to do with the fact that Boldin also changed things so that all accounts are now correlated in their monte carlo. But it sounds like a bug: it should not impact a monte carlo how many accounts you have! It's the asset mix (and taxes) that should affect the monte carlo.
Pralana has a different way of doing things, where you either check a box so that differing assets are 100% correlated and rise and fall together by different amounts; or leave it unchecked (as I recommend) and all assets are uncorrelated (0% correlation: no linear relationship with each other). FWIW, I recommend leaving unchecked since the historical correlation of stocks and bonds is 0.23, which closer to 0% than 100%, and Bills and Cash are even closer to zero.
So I don't think Pralana has that same "bug" or someone would have posted about it!
Again: Boldin often not "ready for Primetime"-- why do folks bow down before it?
