I'm using the Monte Carlo simulation with Conservative Spending to assess the risk that my plan won't meet basic living requirements far into the future. To do that, I'm looking in the Monte Carlo results at the Percentile Distributions of both Living Standard and Income. Specifically, I'm looking at the 5% column to get an idea of the worst case.

The worst case doesn't look very realistic.

What I see is that the program is projecting a 5% probability of Income, and hence Living Standard, far below the "floor" that Social Security provides. In other words, Social Security income in the late years of life well exceeds the values in the 5% column.

How can this be?

In my case, the 5th percentile income in the last 10 or so years was around $15,000, while Social Security benefits would come to about $43,600. Am I interpreting the Percentile Distribution incorrectly?

Note: before I noticed this anomaly, I attempted to address the risk by modelling a QLAC to kick in at my wife's age 85. The program still showed incomes lower than Social Security by itself, let alone the bump that should have been provided by the QLAC.