Captives by the numbers 2026

For the past few years, I have written about the inadequacy of published captive numbers. Most agree the numbers are wrong. The number of captives, domiciles and managers are growing. The published numbers do not reflect this. Why the disconnect? History and definitions are factors.

The lack of reporting is the biggest issue. This year I decided to apply some “captive math” to come up with what I hope is a more accurate number. What is captive math you ask. It like other “new math” you read about: unscientific, humorous, and maybe irrational, but hey, Einstein said the definition of insanity is doing something the same way over and over and expecting different results.

Captive math will give us something new to talk about.

A little history

I believe Guernsey was the first domicile and had the first captive, Commercial Insurance Company Limited (CICL) registered in June 1922. They were not called captives back then. Fred Reiss was the first person to use the term captive in 1962. Captives existed for 40 years before anyone called them captives. Guernsey was also the first to offer protected cell legislation in 1997. I think cells are a big part of where the growth is.

Based on this year’s research, the world has 88 domiciles. This is up from the 82 I found last year.

Forty in the US and 48 everywhere else. My definition includes domiciles that either have a captive, have special legislation, or both. It includes domiciles such as Denmark and Germany that have captives but no special legislation. The UK will soon be coming off the board as a domicile with captives and no legislation. The government is introducing a regulatory framework to make the UK a competitive onshore domicile by 2027. It plans to add cell legislation sometime after that.

This year for the first time, I looked for an accurate number of captive managers. I could not find a single published total. Most discuss the 20 largest. Vermont has 23 on its website, so I know it’s more than 20. The Insurance Managers Association of Cayman lists 20, but one appears twice. Between

Cayman and Vermont, I found five on both lists. That’s 37 unique managers in just two domiciles. I have my AI agents working on a total. Stay tuned.

Let’s get back to licensed captives. Counting licences is probably not the best way to track growth. I

have some historical numbers going back to 1980s. A.M. Best said there were 1,000 captives back

then. Wikipedia said there were 5,000 in 2008. That’s growth. This publication reports annually on new licences. Their numbers for the past two years are 6,040 in 2023 and 6,075 for year-end 2024. 35 licences. That is less than one licence for every two domiciles. I have seen published numbers

stating licences peaked in 2017. That’s right. We had more licensed captives eight years ago than

today. Why this numerical consensus for no growth, when everyone in this space seems to think captives are booming? I believe it’s partly driven by the fact that licensed captives have a 50%+ closure rate.

As in the past, I use the State of Vermont’s numbers to support my claim. I see no reason to change. Vermont’s history goes back to 1981 when it licensed its first captive, First Charter Insurance

Company, for tyre manufacturer BF Goodrich. Vermont has issued 1,413 licences as of December 31 ,2025. 667 are active, 706 are closed and 30 dormant. For every new licence Vermont issues, another

captive closes. Captives are not for everyone. Here is my list of the 88 domiciles. They all have captives, legislation or both. Six more than last year.

Could be that Google search has stepped up its game with competition from other AI sources. I believe some regulators and domiciles are also cleaning up their web pages when the AI bots search

them. I had an interesting AI experience when putting together this year’s list. I found a reference to the state of West Virginia having 563 licensed captives with a hyperlink to the data. You can imagine my surprise when I clicked on the link and it took me to an article I wrote last year that did not even mention West Virginia. West Virginia has no active captives. Don’t trust. Verify.

I think we can agree that the number of licensed captives is not an accurate reflection of growth. My first thought was that most of the growth was in segregated and protected cells. Those cells reside in manager-owned, broker-owned and carrier-owned cell captives. In my search for cell numbers, I also found some serious growth in group captives.

The largest group captive manager has 53 member groups serving 7,700+ members and $5.7 billion in premiums. These are groups sharing risk not individual protected cells. Is that 7,700 captive owners? Guess it depends on who you ask. I bet the members would tell you they own captives. By one definition, that’s more captive owners under one umbrella than all licensed captives combined. This year I want to make a best guess on a total captive number. I want to include the segregated and protected cells contained in the licensed captives but where to start. I began looking for input. I

started with the 6,075 licensed number from above. I know this number is low as it doesn’t represent

all the domiciles. Zurich Insurance Company publishes an annual “Spotlight on Captives”. Its estimate

for 2025 was 8,000 captives and $50 billion in premium. I like the premium approach, but I think this premium number is light. EY’s 2024 Global Insurance Outlook believes captives represent nearly 25% of the overall commercial insurance market. I think that is a good number. Let’s start with that.

The global commercial insurance market is anywhere between $900 billion and $1 trillion dollars.

Let’s go with $950. 25% is $237.5 billion.

The parent company of this publication has done the best job I’ve seen aggregating captive numbers.

It even includes a cell estimate. I collaborated in writing this. Its 2025 report of 2024 numbers recognised 77 domiciles, 6,075 licensed captives 3,753 cells and $240.8 billion in premium. Everyone seems to have a different licensed captive number. I think that’s a definition and reporting issue, I

believe their cell captive number also to be low. Light doesn’t mean wrong. It only publishes the numbers it receives. Not everyone reports. Bermuda, Puerto Rico and Kentucky are missing from its

list. I know for a fact they all have cell captives. I was president of a rent-a-captive in Bermuda for several years. I believe we had 60+ segregated accounts. I know Bermuda’s number is not 0. I have seen industry reporting of 2,000+ cells for Bermuda. I have also seen reported numbers for Puerto

Rico of 300 and Kentucky with 49. That’s almost 40% more cells in three domiciles. Definitions and reporting matter.

The world of captives is not just commercial insurance. Medical stop loss (MSL) is one of the fastest- growing segments in captives. The largest MSL group captive manager has more than 4,000 employers in four group captives premium is estimated to be about $2’5 billion. That averages out to $500,000 per member.

Life insurance reserves in XXX and AXXX captives is another big number. Triple X “captives” are not

really captives, but they are licensed as captives by their domiciles. Vermont, Delaware, Arizona and South Carolina host most of them. Iowa just added life reinsurance to its updated legislation. XXX captive reserve credits totaled more than $240 billion dollars in 2024. I believe XXX reserves is strictly a creation of US accounting regulation.

Captive math

OK, so here is the fun part. How do we take these inputs and turn them into a plausible number? I am going to need a little “captive math” to get there. A humorous, irrational thought process to come up with a number. Captive math is my estimate of a number we all know exists. Unfortunately, not all

domiciles and no one at the IRS is willing to share all they know. Maybe we can change that. The 6,075 licensed captive number includes 262 risk retention groups (RRGs) with an estimated premium of $5 billion or $19 million per RRG. These are all 2024 numbers. If we subtract 262 from the licensed total, we have 5,813 captives remaining. If we assume the remaining 5,813 have the same average premiums as the RRGs, 5,813 x $19 million equals $110.44 billion in premium. The total premium for the 6,000-plus captives and 262 risk retention groups is $115.4 billion.

If you are willing to accept that captives represent 25% of commercial insurance premium, the premium in my licensed captive calculation represents a little less than half of that total. There must be more average premium per licence than my first estimate would suggest. I also believe there are more unreported licenses. Some of that premium is also sitting in some large group and single-parent captives. Some are also hidden in protected cells and segregated cells.

The largest group captive manager has 56 captives with $5.7 billion in premium running through them. That’s more than $100 million in premium per captive. The large MSL group captives I mentioned earlier have $625 million in each of their four captives. That’s $8.2 billion in premium for 60

licensed captives. That’s more than $136 million per licence. The largest single-parent captives by premium are dominated by major energy and industrial firms. The top ten account for more than $4.4 billion in premium. That’s $440 million per licence.

Best guess

One of my earlier inputs was from Zurich. It projected 8,000 licensed captives. That is 20 to 25% more than the commonly quoted licence numbers. I can go with that. $237.5 billion in commercial premium running through 8,000 licensed captives. That’s almost $30 million per licence. 8,000 and $30 million are two nice round numbers. Unfortunately, we still need to account for the premiums in life and MSL captives.

Life captives had $240 billion in reserves in 2024. As you might guess no one reports on how many

XXX and AXXX captives there are. The number of life insurers in the US is between 700 and 900 depending on your definition. Do you include annuity companies that sell life insurance, health insurers that do, or non-profit fraternal organisations such as the Knights of Columbus that sell life

insurance to its members. I am going to go with 800 and assume 10% have XXX reserve facilities.

That’s 80 more captives and $240 billion in reserves or $300 million in reserves per life captive.

Experts believe 25% of employers’ stop-loss premium now runs through captives. That’s $8-$10 billion in premium. The largest group captive stop-loss captives have average member premiums of $500,000. That’s 18,000 employers and $9 billion in premium. I am going to guess that is at least another 100 captives.

My estimate for total captive premium is $486.5 billion if you include commercial, life and MSL premiums. My licence estimate is 8,180; 8,000 commercial, 80 life and 100 MSL. The reported cell number is 37,503*. Since I found 40% more in three domiciles, I believe the cell numbers are double what is being reported. I will go with 8,000 cells. That is consistent with my licence estimate. Once cell for every license. Those are big numbers. That’s the growth everyone is talking about.

Did you get value from this story?  Sign up to our free daily newsletters and get stories like this sent straight to your inbox.