Varying Inflation Rates for Expenses

Post questions about how to use the planner, user inputs, how the planner works, and comments and suggestions.
Tex

Varying Inflation Rates for Expenses

Post by Tex »

Your free online tool is quite amazing. I compared your results to the software I currently use. Doing my best to input the same assumptions, the results were only off slightly.

One thing that I couldn't find a way to input is differing rates of inflation for certain expenses.

i.e. college cost inflating at 7% per year,
or holding lifestyle expenses flat after age 80.

Is there a way to do this in the online version? What about in the upcoming release of your desktop version.

jimr
Posts: 634
Joined: Thu Feb 28, 2008 6:48 pm

Re: Varying Inflation Rates for Expenses

Post by jimr »

Sorry, but while you can change the inflation rate by year, you can't change it by expense item.

Jim

siamon
Posts: 1
Joined: Fri Jul 19, 2013 3:23 am

Re: Varying Inflation Rates for Expenses

Post by siamon »

When you click the additional inputs button on the upper right side of the planner window, a popup window should appear that allows you to be very specific with planner inputs. For each type of cash flow (eg taxable savings, tax deferred savings, retirement expenses, etc) you can enter and annual amount and a start and end year for that cash flow. For one-time items you can enter the amount with the start and end years set to the same year. Using this approach, I think you can model everything you describe.

mallen531
Posts: 16
Joined: Sat Sep 06, 2014 2:04 pm

Re: Varying Inflation Rates for Expenses

Post by mallen531 »

Jim,

This is in regards to your reply below changing the inflation rate per year. Over the weekend, I have had the change to get to run a number of scenarios with your program. This expense inflation question is exactly one that I wanted to ask you based upon what I learned using your program.

I tried using the Monte Carlo section, but when I looked at the detailed results the the expenses were generally twice the annual amounts...I then set up the scenario so that I had a fixed annual expense, which did not use Monte Carlo simulation, the results were untrue, since NO inflation was used.

I then created a detailed Excel spreadsheet to simulate financials using a 3% annual inflation rate with my expenses...which basically cased them to tripled from 2015 to 2040.

So...my question is how does one change the inflation rate from year to year?

I thought about taking a ten year average from my spreadsheet and then creating three different periods of fixed expenses to simulate the inflation rate averages for those years. This would eliminate the Monte Carlo simulation and give me a better idea of how inflation effects my retirement.

Mike

jimr
Posts: 634
Joined: Thu Feb 28, 2008 6:48 pm

Re: Varying Inflation Rates for Expenses

Post by jimr »

mallen531 wrote:So...my question is how does one change the inflation rate from year to year?
The upper table in the Additional Inputs window lets you enter an inflation rate, along with a start and end year. Entries in the upper 'Rate Table' will override the rate information that you enter on the main planner inputs page, at least for the years specified using additional inputs Inflation entries.

Jim

mallen531
Posts: 16
Joined: Sat Sep 06, 2014 2:04 pm

Re: Varying Inflation Rates for Expenses

Post by mallen531 »

Thanks Jim.

I appreciate prompt feedback.

Mike

mallen531
Posts: 16
Joined: Sat Sep 06, 2014 2:04 pm

Re: Varying Inflation Rates for Expenses

Post by mallen531 »

Jim,

I set up an annual 3% inflation rate.

My settings are as follows:
Investing Style-Below Average Risk
Annual Retirement Spending-$100K
Inflation-Average 3%
Inflation Deviation 1%

Portfolio return-8%
Inflation rate-3% 2% SD

I unchecked all expense inputs except SS numbers and ran scenario.

Yearly simulation details show $100K as planned expense for for each year between 2015-2014, with medium withdraws for each year between $46K to $92K with 85% of withdraws in $40K range. This scenario shows no problems having funds in the account in 2040.

I must still not be understanding what is happening in the planned expense numbers. I would have thought that the 2040 planned expense number would have been close to $215K after 25 years of 3% inflation based upon $100K in 2015?

If I do NOT uncheck the expense inputs and run, the planned expense numbers average X2 times the $100K and thus shows the account empty within 8 years.

It is probably something basic that I don't understand about the program. Mike

jimr
Posts: 634
Joined: Thu Feb 28, 2008 6:48 pm

Re: Varying Inflation Rates for Expenses

Post by jimr »

You don't mention what the Spending Policy on the main input page is set to. When you're troubleshooting unexpected things happening it can help to set that to stable, which turns off the dynamic spending adjustments based on portfolio performance.


Also, I'd suggest keeping the standard deviation of inflation at 0. I put that feature in to enable modeling inflation as a stochastic variable, however, in the end it doesn't offer much insight and only adds confusion. Instead, it's better to use sensitivity analysis to test how your plan does under varying inflation assumptions.

Despite those suggestions, I suspect you're getting confused by something that trips up a lot of folks. The planner always shows dollar amounts in "today's value" dollars. This is important to keep things consistent when comparing amounts from year to year, but it also can be a source of confusion.

To understand how this works, I'd suggest running a simple experiment using a new file. Leave everything at the default setting, but put $800,000 in the tax deferred portfolio, set the spending policy to stable, and run the planner.

The detailed view, should show the annual spending staying exactly flat at $50000. This is because values in that table are shown in today's value dollars. The steady Planned Expenses amount from year to year indicates that planned expenses (and also the withdrawal) are exactly keeping up with inflation.

Next, zero out the $50,000 annual retirement spending on the main inputs page. In its place, create an "Other Expenses" cash flow using additional inputs. For this cash flow, start year is retirement start, end year is end of plan. However, unlike the main input spending amount which automatically gets adjusted for inflation, for this cash flow set the COLA Type to "No Cola" and click the cash flow.

Now when you run the planner and look at the detailed view, you'll see that the $50,000 decreases from year to year. The nominal amount of spending will still be $50,000 each year, but as the table shows, that $50,000 will be losing purchasing power each year.

mallen531
Posts: 16
Joined: Sat Sep 06, 2014 2:04 pm

Re: Varying Inflation Rates for Expenses

Post by mallen531 »

Jim,

Thank you for the feedback.

The spending style was: Conservative.

I am following you on "today's" dollars.

I think that I am getting confused around what you said in your last paragraph, "...you'll see that the $50,000 decreases from year to year. The nominal amount of spending will still be $50,000 each year, but as the table shows, that $50,000 will be losing purchasing power each year".

It's the thought of your expense amounts diminishing (I understand that this is done due to inflation showing today's numbers), but it seems as if these amounts should be increasing (still in today's dollars) to compensate for inflation...not decreasing...since this would only increase the length of time that you have funds not decrease the time that you have funds...which doesn't seem correct? This is where I am getting confused.

I will run the scenarios and keep working on learning this your program. Gosh, I do hope that these questions are helping others! ;) It is a great tool...but, one definitely needs to know how to operate it or otherwise is it "garbage in...garbage out". Thanks again for your PATIENCE! :) Mike

jimr
Posts: 634
Joined: Thu Feb 28, 2008 6:48 pm

Re: Varying Inflation Rates for Expenses

Post by jimr »

mallen531 wrote:...but it seems as if these amounts should be increasing (still in today's dollars) to compensate for inflation...not decreasing...since this would only increase the length of time that you have funds not decrease the time that you have funds...which doesn't seem correct?
Mike,

These are excellent questions and I'm sure this thread will help others. This is a constant area of confusion and you're not alone.

The "today's dollars" amount shown for expenses is decreasing because we explicitly told the tool not to adjust the spending for inflation (No COLA or cost of living adjustment). We also told the tool to use the stable spending policy, so there's no adjustment in expenses due to portfolio performance. The tool is doing exactly what we told it to.

However, you are correct that all things equal, taking a non-inflation adjusted withdrawal each year, compared to adjusting the withdrawal for inflation, should result in a higher chance of the plan working and lower chance of running out of money. This shows up in the results in two ways. First, you'll notice that in our simple example, the probability of success increases from 96% when taking inflation adjusted withdrawals, to 99% when we choose the No Cola option. In addition, the median ending portfolio value is around $4 million in the inflation adjusted withdrawal case, while it's almost $5 million in the No Cola case. In other words, although this hypothetical retiree will have to cut their real (inflation adjusted) spending each year by the inflation rate, by doing so they improve their plan's chances of success and the expected ending portfolio value that could be used for a bequeath.

Of course, the goal for most people is to create a plan where they don't have to make such drastic spending cuts during periods of normal inflation. As you say, garbage in yields garbage out. The key is that the tool can help show what might happen under many different sets of assumptions.

Post Reply

Who is online

Users browsing this forum: No registered users and 7 guests