Well, I'm frustrated again, hehe. I must be "a tough room", when it comes to software.
I downloaded ESPlanner and fiddled around with it for many hours trying to get it to do what i-orp already does for me (for free). But ESPlanner never got even close. I spent a couple more hours reading FAQ and forum posts and concluded that it has less capability than I imagined when it came to what I was looking for the program to do for me, which is tell me the optimal way to pull from my retirement accounts. I guess that's the problem with modeling software, if it's not set-up to model what you need modeled, it's a non-starter.
On the inputs side, I was surprised to see that the user could select the sort order for which types of accounts to pull from first. That was the job of the optimizer, I thought. But I then I figured that was for 'what if' analysis, and let that go.
The bigger problem was that the results always included two flat income streams, one from my retirement accounts and one from my spouses retirement accounts. And since I'm modeling retiring early, it was not leveling the spending. In fact the spending was jumping all over the place. The term that ESPlanner folks use is "borrowing constrained", but there's really no need for me to borrow since I can access my various retirement accounts with the "substantially equal payments" rule and leaving employer in the year you turn 55 rule, etc.
Here's the output from i-orp showing pulling every which way to get level and minimize taxes:
That's a download of the i-orp output, pasted into a spreadsheet, and then I levelled it out to today's dollars (that's why it's a little jagged...rounding errors).
Anyway, THAT is what an optimizer should do, IMHO. Too bad the inputs on i-orp are a bit on the simplistic/non-detailed side.
--Dale--
If you'll notice, the word "optimizer" appears neither within Esplanner nor on the esplanner.com website. The essential difference between Esplanner and an optimizer is that an optimizer is making decisions for you. Esplanner, on the other hand, is just a calculator that is calculating out iteratively the consequences of the choices and assumptions you have made. (With the exception that Esplanner does propose a level of life insurance to buy.) So, it's more like Excel, which is not an optimizer, but which can be used to do optimizations by considering scenarios posed by the user, such as alternative business plans for a startup.
The advantage of Esplanner is that, after calculating out the consequences of your choices and assumptions, it reports to you the level of disposable income available to you. (Esplanner's default assumption is that you want to keep this level of disposable income constant, but you can easily override that.) The value of planning according to disposable income is that is what most of us have done all of our working lives: look at the paycheck and then decide on our household spending. This is the opposite of most of the free planning tools available that insist that you start with your level of spending and then adjust the level of risk that you take (aka "the level of risk you need to take") in the hopes that market returns will supply the missing financing for your lifestyle without wiping you out.
So, Esplanner is intended to be used to let you setup varying scenarios and compare the results, where it matters, in disposable income available to you. For one example of the value of this approach, consider the folks who try to approach the delay-SS-or-not question without a tool like Esplanner. They know that spending down savings now to get a better monthly benefit in the future will lower their net worth immediately and for some time. They don't attempt to calculate the lifetime effect on spendable income and instead do a break-even analysis, a calculation which is available, but irrelevant. More importantly, they don't consider the alternative to delaying SS is to save your own assets against the possibility of your living longer that you expect. Those take-SS-now folks never mention increasing savings against the prospect of a longer old age. Esplanner does exactly this. If you put in your expected final age of 100, it proposes spending now and until 100 at a constant rate. If you compare that rate of disposable income against the rate it proposes if you delay SS until age 70 you will notice that your disposable income goes up immediately and forever in the latter case because you are freed from the need to finance your own extreme old age from savings.
Without a comprehensive calculator like Esplanner, people who are planning their futures have to fall back either on heuristics ("4% SWR and cross your fingers that the future turns out to be like the past") or optimizing subgoals that are more readily calculated ("minimize taxes" or "break-even on delaying SS") but of lesser importance, if any. Neither approach is satisfactory, in my opinion.
All that said, however, Esplanner does not let you test an infinite set of variables. You have noticed that you can't specify which retirement accounts to drain and when. I would have liked Esplanner to allow me to specify levels and times of Roth conversions. Esplanner doesn't do these things. There are limits to the granularity of a computer program can have and the implemented options depend on the developer's sense of what the market wants to prioritize and is willing to pay for. My own priorities on the granularity in Esplanner would have been a little different, but that's hardly surprising and does not encourage me to overlook the unique value of Esplanner's approach.
