I’m always glad to see more attention on Roth conversion analysis, but I think Vanguard’s BETR framework has a meaningful flaw.
On the surface, the idea sounds smart: calculate a “break-even tax rate” to decide whether converting today makes sense, especially based on where the tax payment comes from. Fair enough. That does matter.
But in my view, it’s being given too much weight, while a much more important issue gets glossed over:
your marginal tax rate today is not the same thing as your effective tax rate in retirement.
That distinction is where a lot of Roth conversion math goes sideways.
If I’m converting dollars today in the 32% or 35% bracket, I’m paying tax at my highest marginal rate. But if those same dollars are withdrawn gradually in retirement, they usually won’t all be taxed at one flat future bracket. Some may be covered by the standard deduction. Some may fall into the 10% bracket, then 12%, then 22%, and only a portion may reach 24% or above.
That means the “future tax rate” on traditional IRA withdrawals is often much lower in practice than people assume.
And that’s my main issue with the BETR approach. If the framework treats retirement taxation too simplistically, it can make a Roth conversion look better than it really is.
I also think the taxable account comparison can be misleading if it assumes an unrealistically tax-inefficient portfolio. In real life, many investors hold broadly diversified ETFs, index funds, and long-term positions that are fairly tax-efficient. If that’s the case, the “cost” of using outside assets to pay conversion tax may not be nearly as punitive as the model suggests.
In other words, I don’t think the question is simply:
Where do I get the money to pay the tax?
The bigger question is:
What tax am I locking in today, and what tax would I likely pay over time if I leave the money alone?
That’s the heart of good Roth conversion planning.
Now, to be clear, I’m not anti-conversion. I recommend Roth conversions often when they fit the broader plan. They can be extremely valuable for managing future RMDs, creating tax flexibility, reducing survivor tax risk, and sometimes avoiding or minimizing IRMAA later on.
But I don’t believe these decisions should be driven by a one-year formula or a neat break-even shortcut.
They should be built around a long-term retirement tax map that looks at:
- future income needs
- Social Security timing
- pension income
- RMD exposure
- taxable account efficiency
- Medicare premiums
- filing status changes later in life
- and the likely spread between today’s marginal rate and retirement’s effective rate
That’s the fiduciary way to look at it.
A Roth conversion is not automatically good just because you can pay the tax from a taxable account. And it’s not automatically bad because your future bracket might be lower. The analysis has to be deeper than that.
If there’s one takeaway I’d leave investors with, it’s this:
Don’t let elegant tax formulas replace actual planning.
The best Roth conversion decisions usually come from thoughtful multi-year tax modeling, not from a single break-even calculation.
Sometimes the right answer is convert.
Sometimes the right answer is wait.
Sometimes the right answer is partial conversions over a number of years.
But I would be very cautious about any framework that underestimates the difference between marginal tax rates now and effective tax rates later. That’s not a small detail. It’s often the whole ballgame.