This help text appears at the top of the settings page:

Rates of return can be highly volatile. Please be very cautious setting your long-term average rates of return below. Being very cautious with this and all other MaxiFi inputs will lead you to spend less and save more, helping protect you against adverse outcomes.

But why not enter what I think I'll get on average? The 1.25% seems very conservative, and when I use 7% I see I have a much higher annual discretionary spending allowance. 

You can read this page to learn more about deterministic and stochastic planning approaches and how they work in MaxiFi Planner and how they influence our decisions about rate of return assumptions. 

Using a deterministic model like the Base Plan, the assumption is that you are using what you view as a "safe" or "conservative" rate of return, a rate that will reflect your spending "as if" you were getting that rate (despite the fact that you might be getting a better return rate than that on average). Statistically, per your asset allocation, you may *average* a better rate of return, say 7% average, but the idea is that you set up the model to spend cautiously, say 0% or 1% real return, and recalibrate the model annually. From the point of view of the present year looking forward at an uncertain future, you would be spending cautiously (I think I'll get 7% on average but just in case I'll spend as if I'm getting 2%), and retrospectively, looking back at the model after a life lived (hopefully) you will have under spent your model annually thus (again retrospectively) see not a flat spending level line chart, but a spending level chart that tips upward over time, increasing incrementally each year. 

And it is very important to note that when we observe something like "the market gets on average 7% each year" that doesn't mean that you will get that average. People make this comment about "average returns over the long term" as if, naturally, they are going to get the average! But as common sense observes, and the Monte Carlo analysis shows us, an "average" is average precisely because some people don't get anything close to the average! And other people do much better! Which are you? You don't get to choose and a for many, even being in the bottom 25th percentile (a real possibility) is not good news. As the old saw goes: "you can drown crossing a river that is, on average, just two feet deep." In other words, even if you do get the average, you still have to be concerned about sequence of return risk. To extend the metaphor a bit, we can cross the river in say 500 different places, and we don't know in advance where the high and low spots are in advance and, indeed, given the path we take, we don't know what *our average* will be compared to a different path we might have been on. 

Thus, assumed rate of return MaxiFi is asking for here is also bundling an assumption about spending behavior, say, cautious, moderate, or aggressive. By using a safe rate of return that spends cautiously, you are able to view your model, perhaps not as "for sure" or "worst case" but certainly as safe and probable or cautious. This serves well for the Base Profile run in non-Monte Carlo mode. Of course you can set up and run an alternative profile with higher rates of return to see what that compares to your base profile. Indeed, you could even adopt the higher rates of return as your new base profile, but it's probably wise to first run a risk analysis, Monte Carlo, using the asset allocation specifics that will generate this higher rate of return along with the spending behavior before establishing these assumptions in your base profile. The downside risk you observe may (or may not) give you reservations about locking in a long-term model at these higher assumed rates of return/spending behavior.