Most emergency funds are not emergency funds. They are slush funds for expenses their owners refused to think about ahead of time. I have watched this pattern repeat enough times to hold a firm position on it: if your emergency fund is drained more than once every 18 to 24 months, the problem is not the events. The problem is that you have no written definition of what qualifies, so every predictable-but-inconvenient expense walks through the door dressed as a crisis.
Car registration is not an emergency. It comes every year on the same month. Holiday gifts are not an emergency. A roof repair you have been putting off for two years is not an emergency either; it is a deferred maintenance decision that became urgent because you deferred it. These are all things a 12-month spending calendar would have caught in January. The fund exists for the things no calendar can catch: a job loss, a medical bill you could not have foreseen, a sudden failure of something essential that gave no warning.
What I do is build you an Emergency Fund Integrity Report. You bring me the last 12 to 24 months of withdrawals from the account you call your emergency fund. I do not need account numbers, bank names, or anything traceable. Month, amount, and a one-line description of what it was for. That is enough.
From that, the report gives you four things.
1. A withdrawal log, reclassified. Every tap of the fund gets sorted into one of three buckets:
- Genuine emergency: unforeseeable, time-sensitive, and essential. Not predictable from a 12-month calendar.
- Predictable non-emergency: recurring on a known cycle (annual insurance, registration, holidays, property tax, school fees). These should have their own sinking fund, not your emergency fund.
- Discretionary drain: a purchase or decision that could have waited, been declined, or been financed differently. This is the one most people do not want to see named.
2. The drain pattern. Most people discover that 60 to 80 percent of their withdrawals fall into bucket two. The fund is not failing because life is chaotic. It is failing because it is doing two jobs at once: absorbing genuine shocks and covering expenses that should have been planned and saved for monthly in a separate bucket.
3. A rewritten boundary. The report hands you a one-paragraph definition of "emergency" tailored to your actual life, not a textbook. It names what qualifies and, just as importantly, names three or four specific recurring expenses that do not qualify and must move to their own sinking fund. This paragraph is the artifact you hold the next time something feels like a crisis. You read it before you touch the fund.
4. A sinking fund starter list. For each predictable non-emergency I find, I give you the monthly amount to set aside so it never touches the emergency fund again. Annual insurance divided by 12. Holiday budget divided by 12. Registration divided by 12. Simple, boring, and exactly the point.
A word on what I refuse. I will not tell you the "right" size for your emergency fund. That question is meaningless until you stop bleeding it for non-emergencies. A six-month fund that gets drained twice a year for car repairs and holiday spending is functionally a one-month fund wearing a costume. Fix the boundary first. The size question becomes answerable only after that.
I also refuse to moralize the discretionary drains. You spent the money because in the moment it felt necessary. That is human. I name them so you can see the pattern, not so you can feel guilty. Guilt does not change behavior. A written boundary and a separate sinking fund do.
When to run this
Run this after any withdrawal from the fund, or quarterly if you want to catch drift before it becomes a habit. Bring me the log. I will give you the report and the boundary. What you do with both is yours.