I am running a scenario where a cash shortage is automatically covered by Pralana's Unscheduled Withdrawals from a taxable account (before 2035) and then a tax-deferred account (after 2035).
My analysis consistently shows a one-year lag in the taxation of these withdrawals: the income/gains are generated in Year T, but they are only reflected in the Year T+1 AGI/tax calculation.
This lag creates a self-perpetuating issue: the tax for the withdrawal in Year T is paid in Year T+1, which then creates a new, smaller cash shortage in Year T+1, requiring another unscheduled withdrawal.
I have attempted to solve this by manually entering a multi-year Scheduled Withdrawal series equal to the calculated cash shortage for each year.
My key observation from this attempt is:
-
The Initial Scheduled Withdrawal successfully forces the capital gains/ordinary income to be calculated in the correct year (Year T).
-
However, this initial withdrawal only covers the spending shortage. The newly calculated tax on the withdrawal itself creates a residual cash shortage in that same year, requiring the user to manually increase (or "gross-up") the scheduled withdrawal amount to cover both the spending and the tax.
My questions for the Pralana team are:
-
Is this one-year tax lag the intended and expected behavior for Pralana's Unscheduled Withdrawals?
-
Can you officially confirm that the manual, grossed-up Scheduled Withdrawal (iterated until the cash shortage is zero) is the intended and recommended best-practice solution to eliminate this tax timing lag and ensure the tax on the withdrawal is paid in the same year the withdrawal is made?
Thank you for your guidance on this core mechanism.
@tdejene1068 The one-year lag is intended and will be necessary until we solve the problem of performing iterative tax calculations to gross up the taxes on the unscheduled withdrawals. In the meantime, we do not necessarily consider it best practice to manually set up scheduled withdrawals in an attempt to force same-year tax calculations. We do have a truth model to compare with and the long-term error associated with the one-year lag is small. Since we consider Pralana tools to be long-term forecasting models we don't believe this is a major issue. Nevertheless, we have been pursuing the iterative calculation enhancement for a long time and are hopeful of fielding it sometime in the next several months. This difficult problem has finally been solved in our Excel-based model, Pralana Gold, and will released in the 2026 model in January.
Stuart
