Self-Insure a $123,000 Disaster Gap or Pay $2,560/Year? The 1.9-Point Break-Even on a $410,000 Home With Mortgage Rates Above 7%
It's September 30, 2026. Say you own a home that would cost $410,000 to rebuild, and you're holding a $2,560/year quote for earthquake, flood, and hail add-ons. Both numbers are an example I built for this post, so swap in yours. The other option is to park $123,000 in a reserve and self-insure. Which is cheaper?
The market backdrop just moved the answer. NerdWallet's Mortgage Rates Today, Wednesday, September 30 says rates are "steadily above 7%" and inflation is "still running hot." The Bureau of Labor Statistics lists August CPI at +0.4% for the month. In the example below, the two strategies tie at a 1.9-point spread between what your reserve earns and what that money could earn elsewhere. Above that spread the policy wins, and below it self-insuring wins. Where you land depends on inputs only you have.
What the September 2026 numbers change
Here are the readings from the sources I pulled, and what each does to a gap analysis:
| Indicator (source) | Latest reading | What it changes |
|---|---|---|
| CPI, Aug 2026 (BLS) | +0.4% in one month | Repeated for 12 months, about 4.9% (1.004¹² − 1). Limits and deductibles drift. |
| Unemployment rate (BLS) | 4.1% | Your reserve may also be your job-loss cushion. |
| Payroll employment (BLS) | +162,000 (preliminary) | Hiring continues, but a regional disaster hits local employers too. |
| Average hourly earnings (BLS) | +$0.10 (preliminary) | Building a reserve from income is slow. |
| Mortgage rates (NerdWallet) | Steadily above 7% | Sets the return on paying down debt instead of holding cash. |
| Stock market (Mr. Money Mustache) | Record levels, AI-bubble worries | A reserve held in equities carries market risk. |
I wouldn't forecast anything from one month of CPI. Annualizing it is an illustration, not a prediction. On rebuild cost, a repeat of +0.4% for a year would add roughly $20,100 to a $410,000 rebuild. CPI isn't a construction-cost index, though, so check what your policy's inflation guard actually tracks.
On wages, the BLS release has the exact level. At any hourly wage in the mid-$30s, a $0.10 gain is under 0.3%, which trails 0.4% CPI. That matters because the reserve is funded from savings rate, not from hoping the gap shrinks.
Step 1: What a standard policy leaves on you
A standard homeowner policy (HO-3) excludes earthquake and flood. Wind and hail are covered, but often with a percentage deductible and a depreciated payout on older roofs. The National Flood Insurance Program (NFIP) caps single-family building coverage at $250,000, which doesn't bind in this example.
Example losses, with placeholder annual probabilities (not forecasts):
| Peril | Example loss | Annual odds (placeholder) | You keep, standard policy only | You keep, with add-ons |
|---|---|---|---|---|
| Earthquake | 25% damage = $102,500 | 0.4% | $102,500 (excluded) | $61,500 (15% deductible) |
| Flood | 30% damage = $123,000 | 0.5% | $123,000 (excluded) | $2,000 (deductible) |
| Hail, roof | $24,000 replacement, 12-year-old roof | 4% | $20,200 | $4,100 |
The hail line is where hidden costs show up. With a 2% wind/hail deductible ($8,200) and the roof paid at 50% depreciated value ($12,000), the insurer pays $12,000 − $8,200 = $3,800. You cover the other $20,200. A $260/year endorsement in this example buys replacement-cost roof coverage and a 1% deductible ($4,100), which leaves you with $4,100. Hail-heavy regions can look very different. I covered that in the Midwest hail math.
Notice that the largest single loss without add-ons is the $123,000 flood. That is where the reserve number comes from. You size a reserve to the biggest event you can't absorb, not to the sum of every peril. The exception is perils that arrive together, like wind and flood from one hurricane. In that case, add them up.
Step 2: Price both strategies per year
Add-on premiums in this example are $1,280 for earthquake, $1,020 for flood, and $260 for the roof endorsement, for $2,560 total. Expected retained loss is probability times what you keep:
- Strategy A (self-insure): (0.004 × $102,500) + (0.005 × $123,000) + (0.04 × $20,200) = $410 + $615 + $808 = $1,833/year
- Strategy B (add-ons): (0.004 × $61,500) + (0.005 × $2,000) + (0.04 × $4,100) = $246 + $10 + $164 = $420/year
| Line item | A: Self-insure | B: Add-ons + deductible reserve |
|---|---|---|
| Premiums | $0 | $2,560 |
| Expected retained losses | $1,833 | $420 |
| Expected cost before reserve | $1,833 | $2,980 |
| Reserve you must hold | $123,000 | $61,500 |
On expected value alone, self-insuring wins by $1,147 a year. But A needs a reserve twice the size of B's, and money sitting in a reserve has a cost.
This is the kind of side-by-side Vorilanex runs for you, so you don't have to build the spreadsheet yourself.
Step 3: The carrying-cost break-even
Define spread as what the reserve money could earn or save elsewhere, minus what it earns as a reserve. "Elsewhere" could be paying down a mortgage or retiring a card balance. Then:
- Net expected cost of the add-ons = $2,560 − ($1,833 − $420) = $1,147/year
- Reserve freed up by buying them = $123,000 − $61,500 = $61,500
- Break-even spread = $1,147 ÷ $61,500 = 1.87%
| Spread | A: loss + carry | B: premium + loss + carry | Cheaper |
|---|---|---|---|
| 0% | $1,833 | $2,980 | A by $1,147 |
| 1% | $3,063 | $3,595 | A by $532 |
| 1.87% | ≈$4,133 | ≈$4,130 | Tie |
| 3% | $5,523 | $4,825 | B by $698 |
| 4% | $6,753 | $5,440 | B by $1,313 |
Here is how that plays out for two different owners:
- A new buyer at 7%+. If cash earns 4% (my assumption, so use your own yield) and extra mortgage principal "earns" 7%, the spread is 3 points. That's the 3% row, where B wins by $698/year. My earlier post on self-insuring a coverage gap with mortgage rates topping 7% walks through this same tension.
- An owner with a 3% mortgage. Cash earning 4% means the spread is zero or negative, so A wins by $1,147 a year, provided you actually have the $123,000.
Most households don't have it. If you'd borrow to cover the loss instead, $123,000 at 7% is about $8,610 of first-year interest, and the self-insurance math stops working.
What changes the break-even
Flood odds move it most. Hold everything else constant and change only the annual flood probability:
| Flood odds/year | Net cost of add-ons | Break-even spread |
|---|---|---|
| 0.1% | $1,631 | 2.65% |
| 0.5% | $1,147 | 1.87% |
| 1.0% (typical high-risk-zone threshold) | $542 | 0.88% |
At 1% flood odds, B wins at almost any spread. At 0.1%, you'd need a wide spread to justify the flood line, and dropping that line and re-running the numbers is probably the better move.
Inflation moves the reserve. If the dwelling limit indexes up by the $20,100 illustrated earlier, the 15% earthquake deductible goes from $61,500 to about $64,500. A percentage deductible scales with the limit, so the reserve you need to hold grows with it.
Equity reserves shrink exactly when you might need them. Mr. Money Mustache's Will the AI Bubble Destroy our Retirement? opens on how the market keeps surprising us, both on crashes and on records. I'm not forecasting either. But if your $61,500 deductible reserve sits in stocks, a 30% drop leaves $43,050, which is $18,450 short. I ran that stress test in this post on a 30% portfolio drop and a disaster reserve. Count a reserve at its stressed value, not its record-high value.
What the averages hide
Expected value averages over outcomes you'll never actually experience. In this example's placeholder odds, over 10 years:
- Chance of at least one hail claim: 33.5% (1 − 0.96¹⁰)
- Chance of an earthquake loss: 3.9%
- Chance of a flood loss: 4.9%
- Chance of at least one earthquake or flood loss: about 8.6%
So roughly 91% of the time, Strategy B costs you about $25,600 in premiums over 10 years (at flat $2,560, and premiums reprice) and mostly pays out on hail. The other ~9% of the time it saves you a five- or six-figure hit. Buying insurance means paying a smaller, certain cost to avoid that tail. Self-insuring accepts the tail and keeps the premiums. Neither is wrong. They are different bets on your own tolerance and liquidity.
Hidden costs outside the model
- Displacement. A standard policy pays temporary housing only for covered perils. After an excluded earthquake or flood, you're paying. At $3,100/month for 6 months (an example figure), that's $18,600 on top of either reserve unless your add-on includes loss-of-use. Credit-card perks like the fourth-night-free benefit in NerdWallet's sponsored IHG Premier piece are nice, but they don't cover months of rent.
- Waiting periods. Most NFIP policies take 30 days to start. You can't buy flood coverage when the storm is already on the radar.
- Premium drift. The $2,560 is a first-year quote. Insurers reprice every year and can non-renew in high-risk areas.
- The double-counted reserve. If the same $61,500 is also your job-loss fund, at 4.1% unemployment it isn't available twice.
When each strategy tends to win
Self-insuring tends to win when:
- Your reserve would be a small fraction of liquid net worth.
- You have a low-rate mortgage, so the spread is small.
- Your flood and earthquake odds are low.
- You could survive six months of displacement without touching the reserve.
Supplemental coverage tends to win when:
- The reserve would drain most of your liquid assets.
- You're in a high-risk flood zone.
- Your alternative use for cash is debt at 7%+.
- Your reserve sits in equities.
Most people fall between these lists, which is why the decision needs real numbers rather than a rule of thumb.
Run it for your home this week
Your numbers will differ from my example. Five inputs move the result most:
- Your actual rebuild cost, not the assessed value
- Each peril's deductible and exclusions, read off your declarations page
- Your real quotes for earthquake, flood, and wind/hail
- The spread between your reserve's yield and your next-best use of the cash
- Your honest annual odds per peril, by address
If you want a walkthrough of step one, how to calculate your coverage gap in 5 steps shows the method on a different home. To get your own break-even spread rather than my 1.9-point example, you can model it for your situation at Vorilanex. With rates above 7% and inflation running hot, the answer you got a year ago may no longer hold.
Sources
- Major Economic Indicators Latest Numbers — Bureau of Labor Statistics
- Mortgage Rates Today, Wednesday, September 30: Steadily Above 7% — NerdWallet
- Will the AI Bubble Destroy our Retirement? — Mr. Money Mustache
- 3 Reasons This Hotel Credit Card Deserves a Spot in My Wallet — NerdWallet
- Why Bilt’s New Launch Could Be the Most Rewarding Card to Rule Them All — NerdWallet