Ari Chester, partner at McKinsey & Company, discusses the sources of disruptive innovation in commercial insurance with David Schapiro, Co-Founder of Planck.
I have known about Ari Chester and his work for a long time but didn’t speak with him until we were preparing for this interview.
A partner at McKinsey, he serves P&C and life-insurance clients and has a particular interest in commercial specialty and wholesale markets. Our conversation about the sources of disruptive innovation and how they could affect commercial insurance was inspiring and eye-opening. Though Ari was happy to share his perspective on the industry, our conversation is not intended to be an endorsement of Planck by Ari Chester or McKinsey.
David Schapiro (DS): What led you to think about where disruptive innovation could come from in commercial insurance?
Ari Chester (AC): There’s a lot of change in the industry. There are new or escalating risks, like cyber or climate change. And emerging digital channels and artificial intelligence could be game-changing. But I think most of the disruption in commercial insurance will come from potential shifts in the value chain. Who does what will be more important than the what itself.
DS: Could you explain what you mean by the “what” compared to the “who does what”?
AC: By the what I’m referring primarily to insurance products supported by today’s insurance processes. There’s a lot of innovation here, but one could argue that this change reflects ongoing evolution, not a revolution.
Look at new risks, for example. The London Market and the US surplus lines market generally cover any niche risk that needs specialized coverage. And eventually, new or specialized risks mature to become standard coverage. Examples include environmental or employment practice liability or even cyber—they start off as exclusions, make their way into surplus lines or other wholesale markets, then mature into mainstream, established products. As experience accumulates, coverage and capacity become widely available for any reasonably insurable risk.
As products mature, everything from pricing to wording to expertise can eventually become widely replicated. It’s hard for any carrier to provide a unique product over the long term.
Another way to innovate is in the service. But service also becomes table stakes since, over time, most companies—not to mention brokers—offer similar services. There are some exceptions. In workers comp and high-limits property, for instance, when insurers attain a significant enough scale and have a network or a critical mass of resources, their offerings are not easily replicated. They also benefit from higher volumes of data and experience, which give an edge to their pricing and risk selection.
Despite these examples, the industry is quite efficient. New risks emerge, and a few companies offer a product or service that’s initially unique, but eventually becomes widespread. This type of evolution doesn’t represent disruption that fundamentally shifts the industry’s economics.
DS: So it’s hard to disrupt the industry with what an insurance company offers. What about your other point, the “who does what”?
AC: If we see major disruption in the industry, it will probably come from who does what. The core challenge in today’s industry is what I call the 60/40: the transaction cost is about 40 cents of every dollar.
It’s hard to see the loss ratio going much lower than an average of 60 points throughout the cycle. If it gets lower, one could argue that the insured has less need for risk transfer, not to mention that updated regulations could recalibrate and reduce the underwriting margin. The bigger challenge is the 40 points of expense needed to deliver the insurance product. That cost is split across intermediaries and the carrier’s internal expense base.
DS: Could you elaborate on the 40 percent?
AC: There are two major elements of the 40 in today’s market: distribution expense and operating expense. Simplistically, about half of the 40 goes to intermediation—compensation for brokers and agents. This commission is well earned: intermediaries ensure that pricing and terms are competitive across companies. Brokers and agents bring competitive parity to the market and maximize value to the end client.
At the same time, there’s duplication in the services provided by the brokers, agents, and carriers. Sometimes, it feels like insureds are essentially paying twice for similar services and transactional support.
There are also significant frictional costs in the hand-offs between parties.
The other half of the 40 is the carrier expense base. There’s a lot of subscale—nonoptimized—operations in the industry. Insurance requires a diverse set of functional capabilities, and most insurance companies do these in house.
True, insurance companies work with TPAs, external counsel, tech providers, and sometimes MGAs and MGUs to support underwriting and claims. But despite having access to these third parties, there’s an instinct to do it yourself and serve as a one-stop shop.
DS: How could disruption come from who does what?
AC: The industry can reduce the 40 cents of transaction cost by rethinking who does what in the value chain.
There are three main ways: insurers could do more and consolidate roles across the value chain. Alternatively, companies could do less and rely more on other entities in the industry, which will reduce duplication and may allow for more scale. Finally, there may be new collaboration models or fundamental reinventions of how core functions are performed.
DS: Could you share some examples? You mentioned how intermediaries’ roles can overlap with insurance companies’ roles. Should insurance companies be doing more or less?
AC: It depends! Both more or less could lead to efficiency gains.
For instance, we could see the carrier expand its role, particularly by providing more direct distribution.
Of course, this is easier said than done given the complexity of commercial insurance, not to mention the many benefits that intermediaries provide.
Looking to personal lines as an analogy, less than 10 percent of homeowners insurance in the US is sold directly. And homeowners insurance is less complex than most commercial products. Therefore going direct is difficult.
However, we see more companies building direct capabilities, and not just micro and small risk. More insurers are building direct insurance for midmarket and even large corporate customers. Technology provides visibility and interactivity that supports this shift.
There’s another scenario in which the role of the broker could be greater and insurance companies may do less. Brokers already provide service, work on claims, and sometimes handle underwriting. They are closer to the client and often have a deep appreciation of risk needs, although brokers are typically hesitant to take on balance-sheet risk or regulatory exposure.
MGA and MGU models, as well as delegated authority, are well established in the US and European markets. We can imagine brokers being even more active in playing an MGA or MGU role.
Taking it a step further, they may increasingly tap ILS as alternative capacity instead of insurance-company capital. In this scenario, the insurance company may be disintermediated.
Conversely, the shift could actually benefit the insurance company. Insurance companies could be more efficient, cut costs, and rely on brokers and agents for distribution and to execute transactions. In short, they could operate more like reinsurance companies.
Then there’s another idea: the role of the carrier and broker could increasingly converge. There would be significant savings if a single entity could do the origination, underwriting, and administration. Can insurance companies acquire or start brokerages? Can brokers cautiously venture into risk taking? Can there be more formal JVs across the value chain? If we do choose a convergence scenario, it would be important to keep competitive tensions embedded within the value chain to prevent any single company from excessively influencing pricing or terms.
DS: Where else do you see disruptive shifts from who does what?
AC: Another major shift could be more collaboration between industry entities. One analogue to this approach is how the major credit card companies became de facto industry–wide payment-clearing networks.
Another analogue is the credit union space within the US banking sector, which includes thousands of smaller institutions; in this industry, there are credit union service organizations, CUSOs, which provide common services at scale.
In today’s insurance industry, this is what RMS, AIR, and others have done in property catastrophe modeling. With such vendors, there is a single source of third-party insight that is widely used, and these vendors have more scale in building insights around weather modeling than any single company could build on its own. It makes sense to centralize this capability.
Imagine if this model were successfully applied to policy administration or other common operational practices—imagine if smaller companies in the insurance industry were to establish service organizations like credit unions do.
DS: The shifts you have described still involve the players in today’s value chain. Could this change?
AC: Absolutely. Another disruptive model is if more insureds collaborate to pool risk and build common insurance-related services. This is not a new idea; it’s a group captive structure. But there aren’t many group captives today outside of healthcare.
The disruption here would be scaling and expanding the group captive structure across more industries for commercial insurance-risk needs.
The insureds benefit from self-retention and self-interest and risk pooling and scale across partners. This could disintermediate insurance carriers, but it could also be an opportunity for carriers, which could provide expertise and benefit from fee-based support models.
Today’s commercial insurance industry actually has roots in group captive structures. For example, Swiss industrial companies formed Winterthur in the late 1800s.
Yet another disruptive model would be more integration with external ecosystems, especially for distribution and risk-related services. The best example is Tesla’s plan to embed insurance within its vehicles and have State National provide the insurance infrastructure.
Whether or not this particular venture succeeds, you can see how it could make sense. Some in the industry are skeptical, with critics pointing out that the driver, more than the vehicle, is what underpins loss experience. Very true. But this critique is moot once you have continuous, real-time data about exposure—which would mean much less underwriting and much more direct measurement.
This is an example of how insurance could be increasingly embedded within ecosystems, particularly as the IoT becomes more widespread. In this scenario, there would still be a need for traditional insurance companies because non-insurance companies typically have little appetite for balance-sheet risk or for handling insurance regulation.
There’s one more disruptive change in who does what: the expanded role of artificial intelligence. In this scenario, artificial intelligence will have attained enough scale to allow for a greatly reduced need for hands-on human involvement across distribution and transaction execution. Of course, commercial insurance remains complex and heterogenous across risk types and service needs. For the foreseeable future, humans will be heavily involved in the value chain. Even when AI reaches scale, humans will still be involved in shaping and teaching the algorithms and actively monitoring portfolios. Even so, the 40 points of transaction cost could be greatly reduced as AI matures and becomes more widely applied.
DS: These are intriguing ideas, but how do we know if they will work? What would need to change for them to be disruptive?
AC: There are three questions we ask when we evaluate the potential for disruption: Can the new model work? Is it significantly better? Who will lead it?
There are many reasons why each possibility we’ve talked about may not work or might fall short of creating a significant change. Still, each of the developments could work for some niches or segments of the industry.
The more interesting question is: Which companies will lead the industry? As the value chain shifts, scale and exclusivity will be important.
For instance, let’s say that the group captive structure gains traction or that there is more integration with third-party ecosystems. There will only be a handful of insurance companies or brokers that can build the expertise and scale to effectively support those specific efforts.
DS: What kind of culture will companies in the industry need to be successful in the future?
AC: Doing any of what we’ve talked about requires thinking differently about today’s value chain. That change is uncomfortable because it often represents a shift of value: there will inevitably be a few companies whose current business model does not survive.
It could be a Blockbuster-Netflix moment for the industry. Thinking differently about the value chain also requires a longer-term view of ROI.
The shifts we’re talking about are not projects that can easily translate into positive earnings next year. They require a longer-term commitment to develop.
DS: If you were to advise one of the commercial carriers about disruptive innovation, what would you say?
AC: What is your investment horizon and risk appetite? Does your company have the culture and drive to disrupt the current model?
I’ve worked with a few large companies who express interest in disrupting the industry, but when they start to imagine the art of the possible, they think incrementally: “If we could get a 2 or 3 percent improvement in value within today’s model, that would be a huge deal. Anything more than that would be too disruptive, too risky, not believable.”
We also see new entrants that think differently and set bold aspirations but often fail to appreciate the grit and nuances of dealing with insurance. They think in terms of a 200- to 300-percent improvement, but when you pick apart the business plan, it’s far removed from reality.
We also see large companies that want to disrupt but don’t have a stable foundation of good performance. It’s like applying machine learning before mastering algebra.
To sum up, few companies have the perfect blend of high aspiration, stable performance, and insurance expertise.
We challenge companies to ask themselves: Do you really have the appetite for disruption? And if you do, why would you be the natural winner?
Ari Chester, Partner, McKinsey & Company
Ari Chester leads global strategy work for McKinsey’s insurance practice.
He serves clients in P&C, life, and wealth-management, including (re)insurance companies, brokers, TPAs, start-ups, and public-sector institutions.
He advises clients on topics including strategy, underwriting improvement, business-to-business sales and marketing, advanced analytics, and artificial intelligence.
In addition to insurance, Chester serves clients in retail banking and fintech, including core account processing, channels, and payments.
In 2012, he cofounded Ingenuity, McKinsey’s advanced analytics solution focused on insurance. He currently co-leads McKinsey’s support for the London Market Group to produce the 2020 London Matters report.
Splitting his time between London and the United States, Chester has also worked in McKinsey’s Dubai, Lagos, and Brussels offices.
Take a look at our next CEO Series interview about underwriting in the evolving human and algorithmic landscape.