AOM, is it accurate in terms of forecasting?

My AOM is 270 days and I started to think about this a bit today in terms of accuracy and forecasting. What I found to be VERY interesting is that if I took my last 5 months of income (averaged out monthly) AND last 5 months of expenses (also averaged monthly), I came up with an average monthly deficit of X dollars which, when I divided by my current cash holdings, equaled to be exactly 9 months before I spent down to zero, presuming that my income remains as it was before. 

 

Is this a blip or a fortuitous event that I have just discovered?  

18replies Oldest first
  • Oldest first
  • Newest first
  • Active threads
  • Popular
  • AoM is not accurate in terms of forecasting, or budgeting health, or even financial health. 

    It simply tells you how long you have had the money you just spent.

    Does anyone know how to link to a specific reply on another post? 

    dakinemaui 's comparison of how AoM vs. CTB answer key questions could be of interest here. 

    Go to this link:

    https://support.youneedabudget.com/t/q6h3pmr/i-dont-budget-a-month-ahead-is-that-an-issue

    And scroll until you see the start of this reply: 

    • dakinemaui
    •  
    • 17 hrs ago
    •  
    Owlette said:
    What other tests tell if we've met the the spirit of the buffer/30-day-AOM guidance?

    I think you're mixing the two types of "buffers": 1) Jesse's hold money for 30+ days (Old Money) or 2) Live on Last Month's income (Classic Buffer). Furthermore, #2 increases the age of your money, but #1 does not necessarily yield #2.

    I do not know how to make those links pretty on my phone, sorry.

    Like 2
      • dakinemaui
      • dakinemaui
      • 3 mths ago
      • 1
      • Reported - view

      Move Light Sound Life  I don't know how to link to a specific post on the mobile site, either.

      Like 1
  • AoM is a backwards looking spending metric, and you could have a huge AoM and be flat broke. Under a VERY limited set of conditions -- consistent income and expenses -- there will be a weak correlation with cash on hand (and therefore a predicted point to run out of money). True Expenses (non-monthly outflows) will significantly obfuscate this, however.

    The fact it agrees now is pretty exceptional, since AoM is based on your last 10 cash transactions, which usually bears no resemblance to average spending.

    Like 2
  • Thanks all for the responses, and I do understand the stated metric behind AOM. I like to play with math and numbers and just found it very fascinating to see a (even if unintentional) correlation between AOM (270 days stated) and exactly how long my current cash holdings 'might' last based on averaging the gap between 5 months of outflow and 5 months of expenses. As an example, I have around $29K in current cash holdings (we recently sold a house), and with my current rolling monthly deficit of $3.2K (average expenses versus average income), my 'stash' came out to 9.01 months of 'coverage' which matches AOM very well. I'll keep playing with this on a monthly basis to see if my theory is accurate or not...

    Like
      • nolesrule
      • YNAB4 Evangelist
      • nolesrule
      • 3 mths ago
      • 2
      • Reported - view

      Patrick down in Texas You can't have a correlation based on one data point. 😉.

      I suggest you use the YNAB Toolkit and monitor their Days of Buffering calculation so you don't have to do the calculation manually in the future. It's essentially the same thing.
       

      I'm not sure why you would want to continue monitoring for a correlation. We've already told you there's no relationship and they don't actually correlate because you can have a big number and no money.

      You used the house as an example. What if you sold the house and bought another one and didn't have extra money sitting in your budget? But the money passed through your budget during the time between selling and buying? What would that do to your AOM, even though the money in your budget was neutral  before and after the pair of executed transactions?

      Note: I can answer this since I did exactly this last summer. If you understand the math, you can answer it too.

      Like 2
    • nolesrule , thanks for the reply and the toolkit suggestion. My Q was posed more as one of the curiosity of unintended consequences versus a built-in functionality of the process. I’ll likely test it at the end of each month just to see what’s what. The fact that my hypothesis matched the AOM exactly warrants a second look, but if no correlation exists I’ll set it aside and focus on the next thing to catch my interests...

      Like
      • nolesrule
      • YNAB4 Evangelist
      • nolesrule
      • 3 mths ago
      • Reported - view

      Patrick down in Texas If it never matched before, why do you think it would match next month?

      Here's a simple test. Create an outflow that's 50% of the money in your budget and then create an inflow for the same amount on the same date. Then rerun your analysis.

      Like
    • nolesrule I have no historical data on AOM (within my own budget) so it was posed as a Q of curiosity. With your example of 50% out and then in, what are you suggesting I will see as an impact to AOM?

      Like
      • dakinemaui
      • dakinemaui
      • 3 mths ago
      • 2
      • Reported - view

      Patrick down in Texas What do YOU think will happen? 🙂

      Like 2
      • nolesrule
      • YNAB4 Evangelist
      • nolesrule
      • 3 mths ago
      • 1
      • Reported - view

      Patrick down in Texas Here are a few other experiments. Do the scenarios below. For each scenario, run your forecasting analysis. Then enter an inflow that offsets the sum of the outflows. Do a forecasting analysis again. Then clear them out and go to the next scenario. Report back what you have learned.

      Scenario 1: Enter 10 transactions each worth about 9.9% of your budget, dated for today.

      Scenario 2: Enter 5 transactions  each worth about 18.8% of your budget, dated for today.

      Scenario 3: Enter 1 transaction for 99% of your budget, dated for today.

       

      Scenario 4: Enter 1 transaction for the sum of all your budget accounts so that you have no money.

      Like 1
    • dakinemaui likely (based on the answers I see here) a finding of no correlation. Given how spot-on the different data points were for today’s line of questioning, it’s worth a moment of my time to see if it repeats or not. Looking larger here, if there is a connection between these numbers and processes, it will be a helpful tool looking forward. I don’t suggest it is or isn’t accurate, just curious what the brain-trust had/has to say...

      Like
    • nolesrule Ok, I did it. Sort of. Not your suggestions, as you posted while I was playing.

      50% lost me 2 days

      75% lost me 2 more days

      97% lost me 2 more days

      Why did I think it was going to drastically drop again? Perhaps the drop will happen I have more transactions. Yep.

      +1 fake $1 transaction --> down 15

      Next $1 --> down 13

      3rd $1 - down 14

      4th $1 - down 14

      5th $1 - down 14

      6th $1 - down 15

      7th $1 - down 14

      8th $1 - down 15

      9th $1 - down 15

      10th $1 - down 5 (low point is 140 lower than normal with the wash at 97%)

      If the wash is back to 75%, the same 10 $1 transactions put my low AoM at 120 lower than normal.

      If the wash is at 50%, the same 10 $1 transactions put my low AoM at 63 lower than normal.

      With no wash, those 10 $1 transactions actually put my AoM at 14 higher than its value without them.... I wasn't expecting that one.

      Like 1
    • Move Light Sound Life And, just for fun, I "spent" more than 100% of my budget. AoM was just down 6 days from its normal today's value. If it meant what it's marketed to, I'd be able to live comfortably for another 5 months with negative bank accounts.

      Oh, and spending 200% of my budget supposedly means I can only live 4.5 months with as much money negative as I have positive in the real world.

      Like 1
    • Move Light Sound Life thanks for the number crunching. I suspected AOM would (reasonably) drop as well based on that type of data interaction.  At the end of the month I’ll update my own true numbers here on this post and I/we will see where the gap lies and if my theory is without merit...

      Like
    • Patrick down in Texas Of course, you are welcome to do that - the exercise is interesting.

      For other people reading, the flaw in the AoM logic is most obvious for me with the wash transactions (how could the same amount of money in the budget on the same day decrease the AoM?) and the 10 $1 transactions (how could a budget with $10 less cause a jump of half a month higher in my AoM?). I suppose operating with negative money while showing months of AoM is a big red flag, too.

      I mean, it shows what the calculation says... but why use that calculation? It's like saying multiplication and addition are the same because 2+2=4 and 2x2 also =4. 

      Like 2
      • nolesrule
      • YNAB4 Evangelist
      • nolesrule
      • 3 mths ago
      • 2
      • Reported - view
    • WordTenor
    • I have the honor to be your obedient servant
    • WordTenor
    • 3 mths ago
    • 2
    • Reported - view

    Nope. Carry on! Life is much simpler once you know. 

    Like 2
  • Move Light Sound Life said:
    Perhaps the drop will happen I have more transactions. Yep.

    AoM includes a 10-transaction averaging step, which introduces a lag to the response. Those $1 transactions simply clean out the window, effectively giving you the raw age (which could be higher or lower than the averaged value).

    Like 1
Like Follow
  • 3 mths agoLast active
  • 18Replies
  • 139Views
  • 5 Following