# Predicting future expenses

I'm trying to develop a tool that predicts my future expenses based on my past expenses.

Example:

I've spent \$150, \$200, and \$180 on fuel during each of the last three months. Accounting for the rise in price of fuel and my recent expenditures, what will be my expected fuel costs during each of the next three months?

Is there a formula I can use to make this prediction?

• If you spent \$480 on fuel over the last three months, it is reasonable to assume that you will spend approximately \$480 over the next three months. It will be close. Is there a reason that you need to be more precise than that? Feb 12, 2016 at 15:39
• Yes there is! This was just an example. I pretend to use the formula for any expense type (food, bank, etc) in any period (months, years). Feb 12, 2016 at 15:46
• @BenMiller I guess OP wants to factor in the inflation of fuel and budget accordingly. :) Feb 12, 2016 at 15:47
• That's the idea. Feb 12, 2016 at 15:52
• You are looking apply a level of precision that's far finer than the data you are analyzing. By this, I mean your 3 data points have a range of 50, and standard deviation of 25. If gas rises, say 5%, and you add \$9 to your average cost, your variation is still nearly 3X that \$9. Spending is far less tied to inflation than it is to personal habits, choices, and weather. In my opinion. Feb 12, 2016 at 16:02

Here's a formula:

Average of past 3 months * expected increase factor = next month's predicted expense.

Repeat this for each expense, using actual past 3 months for each, and your guesstimate for the expected increase for each.

For example, if you spent \$150, \$200, and \$180 and your guesstimated increase factor is 1.05, then your model predicts that your next month's expenses will be about \$185.

Try to get really good at guesstimating, because the formula isn't very useful without that.

• Fair enough, but 1.05 offers a huge annual rate of increase. Right? The normal inflation is eclipsed by his own variation. So if the equation were just the 3 mo average it would be no less accurate. iMHO Feb 14, 2016 at 18:31
• 1.05 was just an example. It must include inflation, changes in personal use, and changes in the underlying expense itself. It could be negative, it could be positive, it could be 0. Gasoline recently has had declining prices, while healthcare has had increases. Guesstimating is hard stuff... Feb 14, 2016 at 18:36
• Exactly! Now, Imagine having to add this factor for each and every item of regular spending. Will he add a factor for weather on his heating or ac bills? Or would an average monthly number based on last 12 months suffice? Feb 14, 2016 at 21:49
• @JoeTaxpayer - as always, it depends. If you want precision, you spend time slogging through the details. If you don't care about precision, then you just average all your expenses together and take a guess. (There is no magic formula. As you said, spending is far less tied to inflation than it is to personal habits, choices, and weather. ) Feb 15, 2016 at 1:21
• @JoeTaxpayer - I agree. I was not as clear as I wanted to be. My point is that it isn't possible to come up with a simple over-arching formula for predicting future expenses this way. As you point out there are far too many unknowns that you must guesstimate, making the overall results less than meaningful. Feb 19, 2016 at 12:11