Retention vs Premium: A Worked Example of How Your Cyber Deductible Changes the Price
A hypothetical $60M manufacturer prices $50k, $100k and $250k retentions. Once waiting periods and coinsurance are counted, the cheapest premium is the most expensive option.
Your renewal question: how much of the first loss should you keep?
At renewal, your broker puts three or four retention options in front of you. Higher retentions carry lower premiums, so the obvious move is to take the biggest discount your balance sheet can stand. That logic misses two things. Your cyber insurance deductible and retention rarely changes alone: underwriters often lengthen the business interruption waiting period at the same time. And if your policy has a coinsurance clause, it applies on top of whatever retention you choose.
This article prices one decision from start to finish. A hypothetical manufacturer with $60M in revenue compares $50k, $100k and $250k retentions and tests each against plausible claims over five years. The goal is to find the point where a higher retention stops paying for itself. You can then repeat the method with your own quotes. All premiums below are illustrative, typical 2026 market ranges. Actual pricing varies by carrier, country, industry and contract, so confirm every figure with your broker.
The hypothetical manufacturer and its three quotes
Hypothetical example. Say a precision-components manufacturer has $60M in annual revenue, about 250 employees and two plants. Its operational technology (OT) network is only partly segmented from corporate IT. It buys a $5M aggregate cyber limit. Controls include MFA everywhere, EDR with a managed detection service, and immutable backups for IT systems. OT recovery is slower. The underwriter quotes the same limit with three structures:
| Option | Retention (per claim) | Annual premium | BI waiting period | Ransomware coinsurance | Premium saved vs A |
|---|---|---|---|---|---|
| A | $50,000 | $68,000 | 8 hours | 10% | n/a |
| B | $100,000 | $61,000 | 12 hours | 10% | $7,000 |
| C | $250,000 | $54,500 | 24 hours | 10% | $13,500 |
Look at the shape of those numbers. Option C quintuples the retention but cuts the premium by only about 20%. Moving from A to B saves $7,000. Moving from B to C saves another $6,500, even though that step adds three times as much retention. The pattern is normal, and it has a structural cause.
Why each extra dollar of retention buys less premium
Carriers start from a base rate set by revenue band and industry, which is why buyers so often benchmark cyber insurance premium by revenue. They then adjust for controls, loss history and program structure. For a manufacturer of this size buying a $5M limit, illustrative 2026 quotes might fall somewhere between roughly $40k and $120k a year, with OT exposure and control maturity deciding where. Premium taxes come on top and differ sharply by country: the UK levies Insurance Premium Tax, India applies GST to general insurance, and various EU and Gulf markets have their own charges. Compare quotes on a like-for-like, tax-inclusive basis.
Retention credits shrink as retentions rise because of how claims are distributed. Most cyber claims are small: a contained malware incident, a misdirected payment, a privacy notification. A retention in the $50k to $100k range removes a large share of claim count, and the carrier credits you for it. Above that, you're mostly absorbing a slice of the larger, rarer claims that the insurer still pays the bulk of, so the credit is thin. Meanwhile the carrier's fixed costs for underwriting, reinsurance and breach-response panel access don't fall at all.
The practical upshot: the first step up in retention usually earns its keep. Every step after that has to prove it.
The business interruption waiting period is a second deductible measured in hours
A business interruption waiting period is the number of hours an outage must last before lost income is covered. Everything inside that window is uninsured, and for a manufacturer those hours are expensive.
Our hypothetical company estimates covered BI loss at about $4,500 an hour during a full plant stoppage: lost net profit plus continuing fixed costs such as payroll, or roughly $108k a day. On that basis, each waiting period costs the following before the retention even applies:
- Option A, 8 hours: about $36,000 uninsured
- Option B, 12 hours: about $54,000 uninsured
- Option C, 24 hours: about $108,000 uninsured
Moving from A to C therefore adds $200k of retention and another $72k of uninsured downtime, to save $13,500 a year. Partial outages cost less per hour, but ransomware that reaches OT tends to stop lines completely.
Check one wording detail before you model anything. Some policies apply the waiting period and the monetary retention together, which is the assumption used here. Others apply whichever is greater. On a single claim, that difference can be worth tens of thousands of dollars. Ask your broker to show you the clause.
Coinsurance: the percentage that follows you past the retention
Many carriers attach coinsurance to ransomware or widespread-event cover, particularly where OT segmentation or backup controls look weak. Percentages vary widely, from around 10% to as much as half of the loss in harder cases. Under coinsurance you pay that share of every covered dollar above the retention, up to the sublimit. Our manufacturer carries 10% on ransomware under all three options.
Buyers often miss how coinsurance and retention interact. A higher retention slightly reduces your coinsurance bill, because less loss sits above it to share. In the ransomware example below, coinsurance falls by $27,200 between Option A and Option C. That sounds useful until you set it against the extra $272k of retention and waiting-period cost the company absorbs to get it.
Coinsurance is also usually the more negotiable term. Evidence of OT segmentation, offline backups and tested recovery times can persuade an underwriter to reduce or remove it, and that often saves more than any retention credit.
Three claims, three options: what the company actually pays
Now run three plausible claims through each structure: a minor malware cleanup, a funds transfer fraud, and a ransomware event that shuts a plant for 72 hours. The ransomware claim includes $420k of incident response, legal, forensics and restoration costs plus $324k of BI loss, for $744k in total.
| Claim (hypothetical) | Option A ($50k) | Option B ($100k) | Option C ($250k) |
|---|---|---|---|
| Malware cleanup, $40k, no downtime | $40,000 | $40,000 | $40,000 |
| Funds transfer fraud, $180k, no coinsurance | $50,000 | $100,000 | $180,000 |
| Ransomware, 72-hour shutdown, $744k total | $151,800 | $213,000 | $396,600 |
Here's how Option A's ransomware figure is built. The first $36k of BI falls inside the 8-hour waiting period. Of the remaining $708k, the company pays $50k as retention. The carrier pays 90% of the remaining $658k and the company pays 10%, or $65,800. Total: $36,000 + $50,000 + $65,800 = $151,800. Option B works out to $54,000 + $100,000 + $59,000 = $213,000. Option C: $108,000 + $250,000 + $38,600 = $396,600.
The small claim costs the same under every option, so retention is irrelevant there. Fraud and ransomware are where the choice bites.
Five-year total cost and the break-even point
Premium savings accrue every year; retentions hit only when there's a claim. A five-year horizon puts both on the same scale. Each row below is a separate scenario, not a running total:
| Five-year scenario | Option A | Option B | Option C |
|---|---|---|---|
| Premium only, no claims | $340,000 | $305,000 | $272,500 |
| Premium plus one funds transfer fraud | $390,000 | $405,000 | $452,500 |
| Premium plus one ransomware shutdown | $491,800 | $518,000 | $669,100 |
| Premium plus both claims | $541,800 | $618,000 | $849,100 |
Break-even is easiest to read as claim frequency. B saves $7,000 a year over A but costs $50,000 more per fraud claim and $61,200 more per ransomware event. So B comes out ahead only if you expect a qualifying claim less often than roughly once every 7 to 9 years. C saves $6,500 a year over B but costs $80,000 more per fraud claim and $183,600 more per ransomware event. To win, C needs fraud less than about once in 12 years and a ransomware shutdown less than about once in 28.
For this manufacturer, Option B is the right call. The step from $50k to $100k is close to break-even on fraud and pays off on ransomware unless claims come often, and a $213k worst case is a hit the business can absorb. The step to $250k fails the test: it saves $6,500 a year while putting an extra $184k at risk on the event a manufacturer fears most. If the company runs at, say, $6M of EBITDA, that's a difference the board would notice.
Retention wording red flags to check before you sign
Two policies can show the same retention figure and still pay out very differently. Raise each of these with your broker before you compare prices:
- Self-insured retention vs deductible. Under a true self-insured retention, you fund defense and response costs until the retention is exhausted, and the carrier may be only lightly involved until then. Under a deductible, the insurer usually pays from the first dollar and bills you for your share. Cyber wordings use both terms loosely, so read the mechanics, not the label.
- Separate retentions by coverage part. Ransomware, BI, funds transfer fraud and regulatory sections may each carry their own, higher retention. Confirm that one incident triggers only one retention.
- Panel vendor conditions. Some policies reduce or waive the retention if you use the carrier's breach panel; some increase it if you don't.
- Dependent business interruption. Outages at suppliers or cloud providers often carry longer waiting periods than outages on your own systems.
- Currency and local rules. Multinationals buying a USD master policy may find retentions that don't line up with local policies issued in EUR, GBP, INR or other currencies. In some countries, admitted-insurance rules also restrict how a global program can respond locally.
A decision rule to apply at your next renewal
If the goal is to lower cyber insurance premium spend without quietly shifting the cost onto your balance sheet, work through these steps with your broker:
- Ask for at least three retention options, each stating the waiting period, coinsurance and any sublimits it changes.
- Estimate covered BI loss per hour for a full outage and convert each waiting period into dollars.
- Model three claims: a small one, a fraud loss and a serious ransomware event using your realistic downtime.
- Divide each option's extra out-of-pocket cost per claim by its annual premium saving. The result is the number of claim-free years the higher retention needs to break even.
- Rule out any option whose worst-case out-of-pocket cost exceeds what you'd fund without board approval.
- Before raising the retention, ask which control evidence would remove coinsurance or shorten the waiting period. That's often the bigger saving.
Our rule: take the highest retention whose break-even interval is at least seven years and whose worst-case out-of-pocket cost stays within your finance team's sign-off authority. Coverage terms and their tax treatment vary by jurisdiction, so have a licensed broker and your finance and legal advisers confirm the final structure.
Quick answers
Is a cyber retention the same as a deductible?
Not always. A deductible usually means the insurer pays the claim and then recovers your share. A self-insured retention usually means you pay costs yourself until the threshold is reached. Cyber policies use the terms loosely, so check how payment actually works in your wording.
How much does raising the retention typically lower a cyber premium?
The first step up, for example from $50k to $100k, tends to earn the most meaningful credit, often somewhere in the mid-single to low-double digits in percentage terms, though it varies widely by carrier and risk. Bigger jumps usually save less per dollar of extra retention, because large claims still fall mostly on the insurer. Confirm actual credits with your broker.
Does the business interruption waiting period count toward the retention?
It depends on the wording. Some policies apply the waiting period and the monetary retention together; others apply whichever produces the larger amount. For a manufacturer losing thousands of dollars an hour, the difference is material, so ask your broker to confirm the mechanics in writing.
Can we negotiate away ransomware coinsurance instead of raising the retention?
Often, yes, particularly if you can document network segmentation, offline or immutable backups, tested restore times and MFA on remote and privileged access. On a serious claim, removing coinsurance can be worth more than any retention credit. Underwriters in most major markets will weigh this evidence at renewal.
What retention is appropriate for our company size?
There's no market standard. The practical ceiling is the most you could pay out of pocket for a serious ransomware event, including the waiting period and coinsurance, without disrupting operations or needing board approval. Model that claim against each option and choose the highest retention your finance team could fund comfortably.
Should a multinational use one global retention or local ones?
Many multinationals set one retention on the master policy and accept different, usually lower, retentions on local policies required by admitted-insurance rules. Check how the two interact on a cross-border incident and which currency the retention is measured in. A broker with a global network can map this country by country.
Written by the Auditinsure Direct editorial team. Prices, features and rules change often - confirm details with the vendor or a licensed professional before you decide.