00:00 - the values are rocketed and the values are in the tens of billions. Now, >> they're projecting data centers with a value of 53 billion by 2028. >> It's not uncommon for that figure to be around or to exceed 80%. [music] >> Hello everyone, welcome to Insurance Business TV and the Q3 edition of our construction insurance power panel. Let's meet the experts who are going to take us through the latest market developments. They are John Faker, chief commercial officer, construction and

00:28 - infrastructure at AON. Chris McTear, director UK Construction, Infrastructure & Surety Team at Marsh and Theo Knottenbelt, Associate in the owners and developers team in UK construction at Gala. So welcome everyone and to start with something newsy. Gateway 2 backlogs are finally starting to clear, but industry commentary suggests Gateway 3 could become the next major pressure point alongside a government consultation on creating a single construction regulator and the regime's move from the HSC to MH CLG in January.

01:01 - So from an insurance and shorty perspective, is the market bracing for a repeat of the gateway 2 disruption at gateway 3 or has the sector learned enough to avoid it? Uh John, let's start with you. >> Yeah, thanks Paul. um plenty of thoughts on this. So, look, no one can guarantee there won't be an issue for gateway 3, but it's unlikely we'll see the disruption on the same scale um as the early gateway 2 experience and certainly the insurance and security market is generally positioning for managed known

01:29 - risk rather than a fresh shock. Uh it's certainly worth just pointing out to those listening uh just to explain the gateway 2 and gate gateway 3 are mandatory stop-go regulatory checkpoints introduced under building safety act for higher risk buildings. So gateway 2 requires building control approval of detailed safety and design plan before construction starts and gateway 3 requires final as build inspection and a completion certificate for anyone can occupy the building. So going back to the question itself, Paul, uh, look,

02:01 - gateway delays falling under both two and three, they're usually not insured events under a standard C or DSU. They're more regulatory program risks. Um, however, delays to gateway 2 could reduce natural program flow because more time is spent frontloading uh the design and documentation and concentrating on the critical milestones. So just in relation to that in itself you know from a DSU perspective where you're normally looking and when you talk to contractors and employers and you start to really

02:32 - understand where the float is and you know how much fat have you got in a construction program suddenly if you're frontloading everything and with gateway 2 that has an actual knock-on effect which means there may not be so much um within that program to make up in the event of an actual insured event which could one trigger the C and then in you know internally the DSU. So there's certainly that to actually make uh make an actual thought process with regard to that and

00:03:02 - this makes projects sort of less resilient to in short events e.g. physical damage major defects uh and from that a defect or damage event that might previously have been absorbed within the program may be more likely to tip the project into DSU as I say because there's no slack around gateway dates and occupation deadlines. I mean underwriters respond by challenging DSU periods and indemnity periods because again when you look at a construction period indemnity period should reflect that construction period but more and

03:32 - more you see the indemnity period not so much half but a lot less the construction period itself. So from that they're ensuring that they reflect the realistic gateway and the remediation timing and with that probing deficiencies uh dependencies planning uh gateways grid key materials and then sometimes excluding or sub limiting non-damage delay exposures tied to regulatory issues. I mean I've got lots and lots of bits and pieces here maybe Paul to sort of um to really mention I mean seriously we've also got particular

04:03 - bits and pieces but I think really just to sort of give an over roof an overall view of that look we expect friction and delay around gateway 3 and insurers will underwrite and price around it but we're not assuming a systemic a systemic shock of the same nature as the first gateway 2 backlog basically just as an overall sort of um summary of this we expect friction and delay around gateway 3 and insurers will underwrite and price around it but we're not assuming you know particular shock of the same nature

04:34 - as the first gateway to backlog. >> Okay. Well, I think an excellent explanation there by John Theo and I guess you know he brought up the issue of delays. I guess time frames are going to be front of mind for clients. >> Yes, 100%. And um I think to sort of expand on what John was saying from a maturity perspective um the market has largely adapted to the new regulatory environment and now recognizes that gateway approvals can extend the project timelines. Um this additional time is increasingly being

05:07 - factored into underwriting and project planning. The disruption experienced following gateway 2 and that introduction was largely driven by the fact that the approval times were underestimated uh and significant backlogs emerged before measures were introduced to improve processing and capacity. So hopefully um they get it a bit better uh first time round but I think a longer term concern however is contractor appetite. So, we're seeing contractors uh particularly those operating across multiple sectors

05:41 - reducing or withdrawing their involvement in areas such as high-rise residential uh student accommodation uh and hotels due to the additional regulatory burden and uncertainty uh associated with the gateway process. While uh while the industry is adapting, there is a risk that regulation could reduce this pool of contractors willing to operate in the sectors in which the gateway regime applies. So that's there's the delay aspect as you say and the cost aspect, but there's also other

06:14 - impacts that sort of leak into other areas of the market, not just specifically insurance, but you know contractors and where they want to operate. >> Yeah. And Chris, any examples so far of delays? >> Well, as John said, there's been really good lessons learned from gateway 2, and we're not anticipating those to be merited in gateway 3, but we are seeing examples of buildings that are completed effectively uh at practical completion, but have been unable to satisfy the golden thread rule through the

06:48 - documentation. And those delays can be significant if a contractor has to go back and revisit some of the as built works to ensure that they match with the documentation. Um, in those circumstances, most C policies have a small automatic extension period, maybe 30 days, maybe 60 days that we can go to. But if the period extends beyond that, we're seeing examples of insurers wanting full chapter and verse for the reason for the delay. And that's because construction insurers generally dislike

07:22 - covering uh vacant completed buildings while they wait for some administration elements to be completed. So there can be a challenge if those delays are uh very long-term but we're expecting it to be more at the margin rather than the uh normal cost of events. >> I've got you Chris as well. I want to ask you about what is an ongoing issue for the sector which is insolvencies. uh construction has been the most insolvency prone sector in the UK. Of course indeed providers are

00:07:53 - reportedly tightening scrutiny in response making bonds harder to secure. So from your perspective is insolvency risk now they're the dominant underwriting concern on projects ahead of the perils that it traditionally sat alongside. >> Um I think the construction market as far as contractors are concerned have always operated on very narrow margins. the profit that they get out of these even multi-million pound projects um is comparatively small and that with the additional risk that employers are

08:24 - looking to put onto them caused sometimes by regulation and caused sometimes by the fact that the employer is not in the construction market themselves. Um it means that more risk ends up sitting with the contractor and that pressure over the long tail period that a construction project tends to work 3 years 5 years can result in insolvency issues. cash flow is king for contractors and it's not always easy for them to achieve and historically we're all aware of some very high pro examples

08:58 - uh of contractors who can't uh continue because of its solvency issues and when that happens that's a genuine problem for everyone involved in the project for subcontractors and for the employer um and it's why you'll find brokers quite often saying for larger projects and owner controlled in insurance program, a standalone project policy is the way to go. And that's because it covers contractors of any tier involved in the project and that makes it easier to remove one contractor and bring in

09:32 - another without a massive delay to the coverage. Whereas if you're on an annual policy, an administrator gets appointed for the insolvent company and it's quite common for them to cancel the annual policies very shortly after their appointment. And that then means the cover that the employer is relying on comes to a halt as well. That's a particular issue for professional indemnity type coverage clearly. Um and so again this would be the reason why a lot of contractors recommend uh the use

10:04 - of a project policy if they can to have the risk standalone. Um planning issues of course don't always translate uh very well that way. So they can be dealt with separately but I think essentially the uh the issues around solvency are a constant issue for contractors >> and just dive into the shorty side for us as well Theo if you don't mind what's happening there >> so for the providers uh similarly to construction uh or risks and those areas solvency is always been at

10:40 - the heart of risk assessment so this is not necessarily a new concern. But what has changed is the scrutiny applied to bond wordings and the underlying contracts they support. So sureties are conducting more detailed reviews of contractual provisions to understand how obligations can translate into bond claims following a contractor failure. So this reflects a broader trend towards more contract amendments which have increased the potential exposure under bonds. So combined with the effects of sustained

11:12 - inflation over recent years, the cost of completing projects following an insolvency has risen significantly. As a result, claim severities are often materially higher than they would have been historically. So historically uh Surety might have expected aggregate claims arising from a contractor insolvency to represent around 40% uh of its total bond exposure. Today it's not uncommon for that figure to be around or to exceed 80%. Uh and this is driven by both higher completion costs um and contractual provisions that can

11:48 - make bonds more susceptible to loss in an insolvency situation. So while insolvency remains a central underwriting concern, the focus is increasingly uh on the scale of the resulting exposure uh rather than simply the likelihood of uh contractor failing. >> Yeah, some excellent uh explanations there from Chris and Theo and John, if you don't mind, just talk to us about how big an issue insolvency has been. >> Yeah, Paul, I mean, there's been a significant amount of contractors that

12:20 - have fallen into administration. I mean this actually began probably back in 2023 and uh if anybody tuned into previous podcasts I did actually go into the details that in 23 2023 UK uh company insolvencies reached a 330y year high so that was 25,000 cases including over 1500 administrations continued into 2024 and 2025 and we thought it slowed down dramatically until the exception of armor more recently and I think we're still to see the true effects of Surety's position with regard to that. But whilst

12:54 - construction is perceived and as you mentioned in in the question to be the in insolvency prone sector it is in fact similar risk to retail and hospitality. It's true that insurers uh became more prudent precoid. However well-managed companies currently don't have the same kind of problem that uh we you know that has been mentioned certainly from that you know insolvency prone sector that we're sort of talking about. So insolvency risk has moved much closer to the center of underwriting conversations

00:13:25 - but is not completely displaced uh the traditional technical peril based concerns and I think really just to sort of just adding to what Theo um Theo has mentioned certainly in relation to that and Chris before that is that regardless of everything we're talking about there are however new entrance that are willing to consider bonds. So the landscape whilst we're talking about this and making maybe a sort of dampening picture, it's not as bleak as maybe we all sort of suggesting. >> And of course, yeah, we're having a

13:55 - construction round table here and I'm sure you guys were probably expecting that there would be a data center question coming your way at some point. Of course, data center construction has become one of the largest single drivers of new capacity in the global construction insurance market. Insured values on subro projects reportedly running into the billions. And of course a couple of guys on this call will be well aware we have dedicated facilities like AON's data center life cycle program Marsh's Nimbus

14:21 - which are racing to keep pace. So uh Theo from your perspective is the London market able to absorb this scale of risk or is capacity becoming the binding constraint on how these projects get built? >> Um I wouldn't say capacity is a sort of a binding restraint. I think data centers and developments in the UK are generally just replacing other historical industrial asset constructions. So, and as you said, you know, there's these really large figures involved, which is obviously a

14:53 - big reason why, you know, these things make press. Um, but the concentration of those values is often not as bad as it appears. So for example, if you got a 10 billion pound project, um each data hall might be around 1 billion and it's a phase development over a number of years. Uh so there's good separation and there's also good separation between those halls. So from an aggregation perspective, uh the main issue around that is really um if for example they were all built on a flood plane, but obviously no one's

15:25 - going to be doing that. But um where there is an aggregation of values is maybe more on the operational side but also um I think in other areas of the world where there you know where there's more severe weather events and if having all these high value data centers in one area you know could can pose serious aggregate risk but we don't really have that uh that issue in the UK. Um I think the growth um the growth of data centers is obviously a very hot topic as you said. Um but it is important to note

00:16:02 - that the market have worked on larger and more complex projects for decades. Um data centers are more straightforward than a lot of specialist projects that we deal with uh and have dealt with historically. So it is just currently a very trending industry sector uh that's booming in the market. Um, and it's obviously creating a lot of noise. >> Speaking of noise, John, just speak to us about AON's DCLP, which I mentioned there. And indeed, is there is there plenty of capacity out there at the

16:32 - moment? >> Thanks for the plug, Paul. Um, and certainly there is. I mean, capacity is absolutely a constraint on these, we can call them mega projects, but isn't just a hard ceiling that stops them being built. The London market can absorb multi-billion dollar data center construction risks. just has to assemble them as highly engineered multimarket programs rather than as plain vanilla placements. So you know you can't have a conversation regarding construction projects without data centers being

17:01 - mentioned and whilst in early 2025 the values the data centers were in the hundreds of millions the values are rocketed and the values are in the tens of billions now. So you mentioned the A on DCLP so our data center life cycle program which started in July uh 2025 uh and at that stage it was a $1 billion value for projects when launched. Uh we now go on 20 you know to July 2026 that capacity has increased to $5 billion due to the increase in values and the pace that these data centers are

17:35 - globally starting to really you know ramp up if you like. So you know we we've had four different revisions in relation to that. It's a global facility or global program as well. So you know we've identified the need for the program and the DCLP provides construction and operation cover for data centers on a global basis. So having done my little um sort of stint in relation to the DCLP having spoken with insurers many are actually running data centers and will continue to do so.

18:04 - But what you need to remember we are in a soft market. I think that's the first time we mentioned that during this podcast. So, you know, and insurers do have targets. So, however, at the same time, they're looking to balance their books. So, yes, data centers, many insurers want to ride the data centers, but they're also looking at other projects and certainly insurers are happy to look at those across the whole sector of the construction uh book to make sure they start to see how they can

18:30 - balance these projects with regard to data centers. So, put simply, that the value of data centers are increasing. However, the London market continues to absorb the scale of this risk. Whilst the values are increasing, the full value is not always at risk with phased handovers and tight and aggressive time scales meaning that full value is not always required. The DCLP is certainly continuing to grow to meet demands of the many data centers and to date the capacity is still there and as these figures increase so will DCLP.

19:01 - >> Yeah. And Chris from your perspective I mean John's kind of alluded to it there. the data centers, you know, that the whole sort of uh area of it is so rapidly expanding. How do you think the markets might deal with this in the future? >> I think there will be challenges for the markets in the future. If you look, for example, just at the Asian Pacific region, um they're projecting data centers with a value of 53 billion by 2028. And that's not very far away. Um and uh the London market is very capable

19:31 - of absorbing very large projects as Theo has said and has done for a number of years. The individual value of these projects can easily be a billion pounds. Um and in the same way as AON have facility, the Marsh Nimbus facility is able to deal with these mega projects and provide DSU and BI type coverage on the end of that which um means that it becomes a risk that the employer and the contractor can carry together knowing that insurers are on board for it. As the value of these projects will

20:06 - inevitably increase as everything in AI seems to expand at a huge rate. Um there may be issues in the future. Um and when that happens I think uh as Theo said earlier the market will start considering earlier the maximum loss scenarios rather than the total value of the entire project because the phase sectional handover um the individual policy limits or uh detailed value engineering might be much more important than just the total value of the overall completed system. Um I say the London market is very capable of leading these

20:45 - projects. Um but the market has to work for the future to find a middle ground to find a way that uh these risks can properly be written and managed for the satisfaction of everybody without unrealistic policy structures. And the key focus for Marsh and I'm sure for AON Galagas is that uh the quality and the breadth of the policy cover must not be traded off just because you've got a very large number involved. The policies have got to work otherwise they're not worth the money the substantial amounts

21:20 - of money that insurers will be expecting in return for them. to expand on the theme Chris if I can. Planning and judicial review risk is emerging of course as a live issue for data center and major infrastructure projects as seen with the quash consent for the IVA hypers scale site. So is that translating into real delay and business interruption exposure for insurers or is it still an isolated case? uh hypers scale was intended to be the UK's largest uh data center and was quashed I think around environmental

21:52 - reasons as opposed to um insurance issues if you like. Um and it's important to examine the risks that are carried by an employer in those circumstances and those that are carried by the contractor. Um for a contractor, their risks are really around um the ones that covered under constructional risks policy and the delay and startup policies and they're triggered by physical loss of damage events. Um the judicial review itself doesn't constitute a physical damage event and certainly not one on site because the

22:25 - site had not started yet. And so those planning issues don't really translate into a delay in startup or business interruption type exposure which construction insurers can manage. Um the risk of delay through uh a planning challenge or judicial review remains with the client and it's been a risk that's been there for a number of years. Um I can remember uh previously working for a contractor when we were appointed on um the Stonehenge bypass and that was 20 odd years ago and that's still in the

23:00 - long grass through constant challenges and the other guys will have similar examples that they know of. Um I understand that there is uh some limited cover available in specialist uh real estate markets um to protect against things like abortive costs where a planning uh application fails. It's not really my area but u it as I say it's a specialist zone and I think that cover is likely um to be more limited with the greater propensity that there is for electional challenge on that project like everybody else insurers

23:40 - want to manage their risk and they do that by uh either increasing the premium or reducing the coverage on risks that are much more likely to come home for them. So as I say with the IVA case, it's important to consider the development transactional risks for uh both projects and separately for investors and where they sit. Um but it continues to be a challenge across all parts of the industry. >> Anything to add on this one, John? >> Yeah, absolutely Paul. Um you know I think between the three of us when we

24:10 - were discussing this particular question uh and obviously the whole issue in relation to delaying BI or business disruption we did actually just touch on the fact that offjem has recently published this consultation uh on proposed reforms and this is in relation to data centers. So you know I thought it's just worth just at least putting it on the radar in relation to the fact it's very you know it's very new. We still need to really understand you know more about it but the whole consultation

24:36 - considers that a data center commitment fee you know which um which again is new and how you ensure that you know will that be in covered under a delay in startup policy will be BI I think there's lots of questions really just to come out of that but it ties in quite nicely with the question that you asked and I think really just a very just to put out some bullet points in relation to that this refundable commitment fee be payable when the projecting accepts its connection offer,

25:04 - the fee would be returned when the project energizes. So, you know, again, it's, you know, we're still in this early stage in relation to data centers and fully understanding the full uh the full values. You know, Theo and Chris have both touched on the extent of the values of data centers, but you know, don't forget the DSU, the lane startup and how the these are big figures, big figures that we're talking in relation to not just the um the actual value, but the you know the loss

25:29 - if these data centers don't get up and running and they are very aggressive in their time scales to make sure that they do hit that. the proposed fee range could be £237,500 to be exact or £712,500 per megawatt. So, you know, this these are big figures and I think that's going to start focusing people's minds when it when it comes to the uh the data centers themselves. So, we just thought we'd throw that one in there because, you know, developers and investors also need to think how that works when they're

25:57 - actually investing in these data centers and the potential loss of revenue um if they don't actually um start the project and they therefore lose out on this commitment fee. >> Well, I just want to throw one final question at all of you if I can. I'll just run through it very quickly. Of course, post 2025, changes to UK skilled worker visa thresholds have tightened access to overseas trades people just as the sector faces a wave of retirements with the temporary shortage list due for

26:25 - review in in mid 2026. So, is that labor shortage starting to show up as a genuine delay and quality risks in C and DSU claims? But it's still more of a cost pressure than an insurance one, Theo. >> Um, thanks Paul. I think it is difficult to pinpoint exactly which claim is to do with labor shortages but I think that there is no doubt that the industry is facing you know massively growing pressure to attract and retain uh a skilled workforce. Um and addressing this challenge will be critical if the

26:58 - sector you know is to meet its future demands. Um labor shortages inevitably increase the risk of poor workmanship. So as you said there is that risk there and it can also cause delays and productivity issues but I think it's important to note sort of what the defenses that you can you can also implement to combat this and I think the first line of defense should always be a robust quality assurance and quality control processes. So where effective inspection, testing and signoff procedures are in place and consistently

27:34 - followed, many defects and associated delays can be identified and addressed before they become material issues. So that said, um the construction industry does actually have a strong track record for adapting to challenging circumstances. So obviously um CO 19 the response during this pandemic demonstrated how quickly contractors can revise their programs, working practices and site operations to maintain progress under these challenging conditions. I think also if labor shortages persist uh a similar level of innovation will be

28:11 - required. So we're already seeing greater adoption of modern methods of construction including modularization, automation, robotics, 3D printing, uh and increasingly sophisticated plant equipment. And when we were speaking about this earlier, John very rightly uh mentioned AI. So you know over time these technologies have the potential to reduce reliance on labor intensive activities uh have the potential to improve productivity and help ensure the projects can continue to be delivered successfully despite any um continuing

28:46 - workforce constraints. >> John Theo mentioned that the productivity pressures there are they are they key? Absolutely, Paul. I mean, again, as Theo's mentioned, AI seems to be something that I think will certainly be coming more and more into this when we talk about the shortage, but I mean, I think realistically, you know, the changes to the UK's skill uh worker visas is showing up more as a cost and productivity pressure than is a distinct uh C DSU loss driver. But it's moving up the list of contributing

29:16 - factors on delay driven claims. So I mean I think when we start looking at this you know there are fewer more expensive overseas candidates coming through at the same time as domestic retirements which is what you sort of picked up within the within the question. So I think in practice certainly from a UK perspective on UK projects the concern is twofold. Repair reinstatement works taking longer to resource and price and labor day rates are higher than the assumptions baked into older C pricing models. And then

29:46 - this inflates the size of fiscal damage claims and extends policy response timelines. Surveys, remedial work, snagging all have an impact in in relation to that. And I think we can really just sort of sum this up in, you know, in a very sort of in in a way that we say yes, labor shortage is contributing to quality design and rework, but it's being consumed under the more familiar headings of workmanship and design management, not relabeled as its own insured risk. >> Yeah. And and then just briefly expand

30:16 - on that for us, Chris, if you don't mind. How are those staff shortages impacting claims? >> Theo and John are right, I think, in in their starting point with this. Um, it's been a constant issue uh recent years to get qualified staff. Um and whilst a C policy will have uh post uh physical loss of damage event will have cover for things like additional cost of work in your additional cost of construction of unbuilt parts in order to try and regain the program. Um that's very difficult to do if you haven't got

30:47 - the qualified staff to allow you to do that repair and to drag them in from other projects is often difficult too. So in a complicated claim, you can even have uh a limited number of experts and that can cause issues around the qualified experts finding the solution to the problem. And if you're not careful of that in the extreme circumstance, you can end up with a kind of well the repair element is too difficult. So let's knock it down and start again. Um because we know how to do that. Now clearly that's not an ideal

31:20 - position for insurers because the costs associated with something like that are huge and get bigger very quickly and if there's a delay in startup t cover attached to the back of it that's a scenario nobody wants to have. So I think that um the contractors are improving their uh flow of new staff in there likely to be a short-term uh deficit of the pool but quite quickly with the uh elements that Theo talked about earlier that pool is likely to be replenished. >> Great thoughts again Jensen indeed huge

31:54 - thanks for your time. Remember that this expert panel will be back together again one more time this year. So keep it right here at insurance business TV. >> [music]