Loss modellers put insured losses from the November 12 Microsoft 365 disruption anywhere from under $100 million to about $1.4 billion. Our worked example shows why an 8 to 12 hour waiting period leaves a typical small business carrying 44% to 67% of its own outage loss.
The Microsoft 365 disruption that began shortly after noon Eastern time on November 12 cut businesses off from Outlook, Exchange Online and Teams, with knock-on problems in SharePoint, OneDrive, Copilot and Defender XDR. Severe disruption lasted roughly 18 hours, and the full event, including degraded performance, ran for about 38 hours. It was Microsoft's second major Microsoft 365 failure of 2026, after a January outage of almost 10 hours.
Within days, cyber risk modellers published their first insured loss estimates. They do not agree. This PolicyRix Team analysis lines the estimates up, compares them with earlier cloud events, and works through what a typical cyber policy would actually pay a small business.
The estimates side by side
CyberCube said insured losses were most likely below $100 million, with an upper bound of about $350 million, and put the impact on global cyber premiums at 0% to 2% of loss ratio. Morningstar DBRS estimated $300 million to $500 million. Kovrr published a range of roughly $300 million to $1.4 billion, alongside an estimate of about $16 billion in lost productivity worldwide. Euler ILS gave a range of $100 million to $1 billion. Marsh Re did not give a figure but said it expected the market impact to be minimal.
On our arithmetic, the widest published range runs from $100 million to $1.4 billion, a fourteen-fold spread. Even within a single modeller, Euler's top estimate is ten times its bottom one. That spread is itself the main finding: nobody yet knows how many affected firms bought cover that responds to this kind of event.
How it compares
For the CrowdStrike software failure in July 2024, CyberCube estimated $400 million to $1.5 billion of insured losses. For the October 2025 Amazon Web Services outage, it estimated $38 million to $581 million. CyberCube's upper bound for the Microsoft event, $350 million, is about 88% of the bottom of its CrowdStrike range, and its central view of under $100 million is about a quarter of it. So on CyberCube's numbers this is a smaller insured event than CrowdStrike, despite running longer, mainly because many machines kept working locally and staff could switch to other tools.
The waiting-period problem
Most cyber policies that cover business interruption from an outside provider's system failure, often called dependent or contingent business interruption, include a waiting period. Typical waiting periods are 8 to 12 hours. Losses inside that window are not paid. Many policies also cover only malicious events at a provider, or exclude non-malicious system failures altogether, and this event was a technical failure, not an attack.
Take an illustrative business that loses $1,000 of profit for every hour its email and collaboration tools are down, and assume the policy does cover non-malicious failures at a provider.
With an 8-hour waiting period, an 18-hour severe outage costs $18,000. The policy responds to 10 hours, or $10,000, so the business carries $8,000, or about 44%.
With a 12-hour waiting period, the policy responds to 6 hours, or $6,000, and the business carries $12,000, or about 67%.
If the policy excludes non-malicious failures, the business carries 100%. Deductibles and sub-limits reduce recoveries further. These figures are our illustration, not a quote from any insurer.
Parametric cover is the other route. Some products pay a set amount when a monitored cloud service is down for a set time. It is not clear this event triggers them, because it hit a software service, Microsoft 365, rather than the underlying Azure cloud capacity many triggers track.
What a business owner should check
Open the cyber policy and look for three things: whether dependent business interruption is covered for system failure as well as security failure; the waiting period in hours; and any sub-limit for dependent business interruption. Then estimate the hourly cost of losing email and file access. If an 18-hour outage would hurt and the policy would pay little or nothing, a broker can quote a shorter waiting period or broader system-failure wording. Keeping an offline contact list and a backup way to reach customers costs far less than any premium.
What the data cannot tell us
The estimates are early and built from models, not claims. They do not say how many businesses had cover that responds to non-malicious provider failures, and that share drives the final number. Productivity loss is not the same as insured loss: Kovrr's $16 billion figure is a measure of disruption, not of claims. Outage length also differed by region and by service, so the 18-hour figure is a broad guide, not every firm's experience. Our worked example uses round numbers to show the mechanics.
The short version: the outage was long enough to test cyber policies, but the wording and waiting period, more than the headline loss estimates, decide whether a business gets paid.
Key Points
- 1The November 12 Microsoft 365 disruption brought about 18 hours of severe outage and roughly 38 hours of total disruption.
- 2Insured loss estimates range from under $100 million (CyberCube's central view) to about $1.4 billion (Kovrr's top end).
- 3CyberCube's $350 million upper bound is about 88% of the bottom of its 2024 CrowdStrike range.
- 4On our example, an 8 to 12 hour waiting period leaves a business carrying 44% to 67% of an 18-hour outage loss, even when cover applies.
- 5Many cyber policies exclude non-malicious system failures at a provider, so check the wording, not just the limit.
Why This Matters
Almost every business relies on a handful of cloud providers. The gap between outage loss and insured loss shows owners where their own cover falls short before the next outage, not after it.
Understand the numbers
- GuidePercentage Deductibles Explained: Wind, Hail and Hurricane Cover
- GuideWhat Is a Combined Ratio? How to Read an Insurer's Earnings in Plain English
- GuideHow Much Home Insurance Do You Need? Replacement Cost vs Market Value
- CalculatorInsurance Deductible Break-Even Calculator · Compare a low and a high deductible using your own claim odds.
Related Stories
Who Should Refinance at 7.11%? Break-Even by Loan Vintage
November 18, 2026
Binance's $6.3bn Hack Shows How Little Crypto Theft Is Insured
November 18, 2026
Australian Home Cover Rises 6% as Rates Hit 4.85%: The Squeeze
November 18, 2026
Cowbell Launches Standalone Insurance for Damage Done by AI Agents
November 18, 2026
Daily Intelligence
The PolicyRix Daily Brief
Get the day's key insurance and finance stories in one email, each linked to its original source. No spam. Unsubscribe anytime.
Informational newsletter only. Not financial advice. Disclaimer