Rob Berger Interview on 10/14/2025

A big part of Berger's focus is (rightfully) about the different monte carlo results with Pralana and projection lab and with boldin. Boldin I think has its exact ror and sd numbers hidden from users; and they also completely revamped their monte carlo since his video. With Projection lab, at time 6'56" I noticed he's using PL's 8% return 16% sd for stocks, and 3% returns 2% sd for the bonds for the monte carlo. Was Pralana set up with these same figures in his monte carlo comparison (or did he rely on Pralana's default)? As we know in pralana you have to set SD separately in the tabs if you want to use your own, and it's not something many people do. I also see in his video that Projection Lab incorporates separate monte carlo for inflation (in PL: 3%, sd 8%) and dividend yields (which Berger doesn't use), neither of which Pralana does. The SD especially might account for the difference in success he's pondering.

My concern with Pralana's competitors are their inability to keep asset allocation constant in a tax efficient portfolio where you preferentially put bonds in tax deferred. A lot of people practice this tax efficient asset location and as far as I know, Pralana is unique among commercial consumer tools in its ability hold overall asset allocation constant which is the only way to do proper comparisons.

Because of this shortcoming, Pralana's competitors will give you wrong answers for things like Roth Conversions as they see the low returns in tax deferred (due to lots of bonds) and high returns in Roth (due to stocks) and so recommend overly large Roth Conversions that end up really just chasing the higher average return of stocks because those programs failed to hold the overall asset allocation constant. The only way those programs work is to have the same asset allocation in all accounts, which is not a very tax friendly thing to do and not useful to people that actually hold a tax efficient portfolio.

I don't think Ron Berger understood this in his review and would like to see a deeper dive.

@ricke Having watched that segment of the interview, I still don't think he understands it (despite Bill's valiant efforts to explain it). He seems to think it is invalid to evaluate Roth conversions if the expected return of the Roth account is higher than that of the tIRA (e.g., if the Roth is all stocks and the tIRA is mostly bonds), even if the overall portfolio asset allocation is maintained by converting stocks to bonds elsewhere (as in Advanced Mode 2). Seems to me that this change in asset location is a legitimate advantageous consequence of the conversion (in addition to the tax rate arbitrage effects).

@plaut

Obviously if you wanted a higher asset allocation, you could just do that, it is a completely separate decision from Roth Conversions. Berger sounds like an idiot, it's a shame he has a big audience.

@ricke Yeah, welcome to the internet 🙂 Lots of folks with big audiences and often you wonder why and actually worry. But Rob has a ton of really good content that helps folks, his Medicare stuff is good. He admits he's been more focused on ProjectionLab and Boldin, but said a few times during the interview he really liked what he was seeing in Pralana and needed to get back to it. I was pretty shocked when he said Boldin doesn't let you set your asset allocation?

@hines202

Glad you tried to explain things to him. Maybe he'll learn a bit more.

I'm always amazed with the gross simplifications that many of these tools force on you. The cottage industry of websites and YouTubers that do reviews never seem to actually dive into the numbers to see what's going on and explain the limitations of the tools and when it might matter - I guess it's not how you get clicks.

I didn't realize there's another edition of the book on the horizon.

Bill/Stuart:

I watched the interview with Rob. One topic I think would be good to cover, that was briefly touched on... the idea of leaving a certain $ amount in pretax to cover long term care expense. This was a message board comment a few weeks back the feature of leaving a $ amount in that pretax account(s) for covering Long Term Care.. that sounds like a good idea. In addition to the $ amount, does Pralana currently know, when people input the years of long term care expense into Pralana, does Pralana know during those years requiring long term care, to take the funds from the pretax reserve left in there (or currently just the amount manually left in there because there is no $ limit to set in there for long term care)?

In other words as Pralana goes about shuffling money among the different accounts in the Advance modeling mode 2 does it know... "Oh during these years that I have long term care if the bill > 7.5% of AGI, I am to take the expense for that from pretax?". It seems like the two would go hand in hand,... have a $ amount reserved in Pretax, and the program pulls from the money if > 7.5% of AGI, in the years the user has defined as requiring long term care amount in the Healthcare section.

LOL. Rob is far from an idiot, though he would be the first to tell you you're welcome to that opinion. He asks a question, around 26:00, that, respectfully, I don't think Bill answered adequately. Rob was just too polite to keep badgering him for an answer. There's a nuance here that obviously is not apparent to the majority of folks, and difficult to convey. Even Stuart had to outsource the explanation to Rick in the manual. I've read Rick's wall of text in the User Manual more than once, and it's taken me a while to wrap my head around it. Still not sure if I understand it 100%.

If the answer is, "Of course Roth conversion results are skewed by asset allocation. That's inescapable, but it's also what happens in the real world. The trick is to ensure the tool you're using takes your overall portfolio allocations into account, and doesn't just shift all of your assets into the Roth without re-balancing the rest of your portfolio," then Bill should have used that opportunity to explain that. I don't think he did. (And if that's not the answer, then I'm an idiot right there with Rob.)

I think Rob recognizes that the complexity of long-term planning makes many of the assumptions about returns kind of suspect, and so he's focusing on the tax implications of conversions as the primary driver. But at the same time, he's looking for that tool that DOES take all of that other complexity into account and actually gives you some useful results. Maybe Pralana does that, but the case was not made (to the layman target user) in that interview. Personally, I'd like to see some way to know whether the change in portfolio balance is due to taxes saved, higher return, or some combination. Show me Active vs Baseline of overall taxes paid in the Roth results right there with the Portfolio balance, along with total lifetime taxes paid before and after.

To pretend that the user manual answers all of the question is disingenuous, too. It's written by engineers who know what they're talking about, but who still struggle to bring it to a non-engineer level of speech and understanding. I've got a EET degree, by some measures an IQ north of 125, and let me tell you, it's a slog. I'm still looking for the two examples of Roth conversions that Bill said make everything crystal clear.

Overall, Bill and Grady did a good job. I think many folks are intimidated by Pralana's user interface and depth of understanding needed to really reap its benefits. But maybe this interview reaches some people who really want a more powerful tool.

@tcbarney

I shouldn't have called him an idiot, but I do think that since Mr. Berger has a large audience doing retirement planning videos (so is presumably making substantial money from this), he really ought to be informed. Normal folks with normal lives aren't supposed to know the ins and outs it any more than they know heart surgery, but the modern way we learn is from folks like Mr. Berger, so his knowledge gap is pretty disappointing.

It seems like Mr. Berger should grasp that you have to control asset allocation in order to make a comparison on the effect of Roth conversions. If you don't, the differing asset allocations create much different outcomes and that overwhelms the tax efficiency of Roth conversions that you are looking to study. In the real world, you can and should choose your asset allocation first and then do Roth conversions to optimize the outcome within your risk profile, but the conversions should not be changing your allocation and therefore your risk.

I get animated on the subject because, as far as I know, the only other tool besides Pralana that controls your asset allocation when using tax efficient asset location is the free RPM spreadsheet at the bogleheads.org wiki. But RPM is so limited and manual in nature that it's not in the same league as Pralana. Pralana's competitors can cost you big time if you follow their path. I've read posts in retirement forums where folks converted everything all at once because some misbegotten program had this problem and the user didn't understand that the program couldn't handle Roth Conversion math when they preferentially held bonds in tax deferred.

P.. S. Sorry for the wall of text in the user manual, I could probably write it more simply now than I did back then.

It was frustrating that he didn't get the point about convertions. Rob Berger's suggestion that programs assume an equal AA 'behind the scenes' in doing a Roth conversion is not a bad idea compared to what the other programs do now. But Pralana's method is way better, since it allows control of the actual AA and also allows you to have tailored caps with IRMAA, tax brackets, Capital gains etc. I agree with @tcbarney that Pralana's method was not adequately explained either.

You could tell Rob wasn't into Pralana, he lacked the passion he does when using his favorite programs.

I am awaiting the Maxifi guy (Kotlikof) in a couple weeks, since I use an actuarial method myself.

Watched Rob Berger's youtube show - Bill & Grady great job!!

A few comments:

At about 32 min into the youtube, Bill said that most people don't go into the weeds when it comes to Roth conversions, they use non-mathematical reasons, do not use optimization, just use marginal tax brackets. I like that. I think we make Roth conversion decisions too complicated. Long term - do you expect to be in a lower tax bracket when say, you turn 70 years old, then don't. Same or higher tax bracket, then convert. Remember at age 70 you will be getting Social Security (SS), maybe pension, and RMDs. Add those up. Using Pralana, see if your RMDs are greater then your needs, if so, convert. Note on SS, waiting from age 67 to 70 you will get a 8% increase per year, PLUS COLA. I think the one think most people are not talking about and definitely do not plan for in Pralana, is very likely future increases in tax brackets, that could be a BIG determinator in the Roth conversion decisions (I don't know if you can change future tax brackets in Pralana on your own).

Brad @bfisher101 in addition to setting aside money in pre-tax accounts for long-term-care, set aside money to cover standard deductions. Standard deduction for a married couple in 2025 is $31,500, times 30 years (not counting inflation and old age deductions) is $945,000.

Posted by: @pizzaman

↑

Brad @bfisher101 in addition to setting aside money in pre-tax accounts for long-term-care, set aside money to cover standard deductions. Standard deduction for a married couple in 2025 is $31,500, times 30 years (not counting inflation and old age deductions) is $945,000.

What is the logic in setting aside pre-tax money to "cover standard deductions"?

@jkandell

I would guess he is saying that if you have no other income, RMDs/withdrawals from tax deferred up to the standard deduction would have a 0% tax rate. But of course it's not that simple back here on planet earth. You have to also subtract any non-qualified dividends earned in taxable, look at the impact of SS benefits and take into account any other income. Plus, we hope that investments grow faster than inflation, so you still have to do math to estimate how much you set aside now to fill up whatever gap you are trying to fill later. Oh, and if a spouse passes early, then the standard deduction is cut in half.

Even the point about reserving money for LTC is only partially correct. You do want a way to get enough income to use the entire medical deduction, but again, you may have other income streams besides the RMD that count.

I see it as something he heard that he thought sounded cool. Using Pralana, all these factors are already in the math, no need to try to work it out by hand.