2023 | 2024 | ||||||
Price: | 170.00 | EPS | 5.5 | 6.5 | |||
Shares Out. (in M): | 140 | P/E | 32 | 25 | |||
Market Cap (in $M): | 23,767 | P/FCF | 30 | 25 | |||
Net Debt (in $M): | 2,371 | EBIT | 814 | 932 | |||
TEV (in $M): | 26,138 | TEV/EBIT | 0 | 0 |
Sign up for free guest access to view investment idea with a 45 days delay.
Verisk – Back to its Roots (Market cap: $27bn, EV: 30bn) (Please excuse the format)
THE STORY…
If you pull out your homeowner insurance policy, you'll notice ISO as the copyright owner of your policy. Verisk's core legacy business, ISO, has a fascinating history stretching back over 50 years and still going strong. At the time, the regulator mandated that all insurance companies report their data to them. However, due to complex state regulations, it became expensive and an enormous effort for insurers to file state-wise data. Moreover, the regulator required the data to be obtained only through its chosen intermediary (Verisk), which could do a quality check and standardize reporting. As a result, Verisk was born (then ISO) in 1971 as a utility for insurers.
Due to its dual-sided network advantage, Verisk ventured into framing less technical, easier-to-read insurance policy language. The regulator welcomed the move as it lowered their workload and reduced ambiguity. Today, if an insurer wants to launch a policy, it can simply borrow the policy language from Verisk, inform the regulator, and shorten its go-to-market time. The only alternative would be to do a lengthy in-house process. It's remarkable for such monopolies to withstand the test of time. But how did Verisk pull off such a feat? First, it had the expertise to grind through every regulation, bill, and case law in the country over five decades. Every year it makes over 3,000 filings incorporating these studies. And in turn, the regulators treat these filings as industry reference policy forms. No wonder today Verisk claims to have a presence in more than 200m policies across 26 insurance lines.
At the beginning of the century, Verisk made successful inroads into data analytics and new risk markets. Verisk's management team, from 1996 to 2013, deserves credit for replicating ISO-like success into analytics. Notably, they pivoted to a for-profit corporation (more on culture later) and established a new identity – Verisk, with the launch of an IPO in 2009. All through this journey, it purchased two marquee assets – AIR Worldwide (in 2002) and, Xactware (in 2006), which expanded its product lines to Catospheric Modeling and Antifraud claims information, respectively.
But suboptimal capital allocation started creeping up when Scott Stephenson (CEO, 2013-2022) attempted to play the Verisk playbook in non-insurance sectors like Energy and Finance. While these verticals enjoyed similar unique data assets like insurance, they turned out to be either more transactional or lacking moat. As a result, the growth never picked up there, dragging down RoICs, and distracting Management's attention.
However, on the brighter side, newly promoted CEO Lee Shavel has displayed the same attributes as the previous generation of the management team. For example, as CFO (2017-2022), he openly addressed the issue of dilutive returns from his predecessor's acquisitions. In addition, going back to its roots, he's on an accelerated path to untangle the entire non-insurance businesses, making Verisk a pure-play insurance analytics firm.
What you'd read below
I. The Problem at Verisk
Despite outperforming S&P 500 over the previous ten years (annualized return of 14% vs. 11.8%), Verisk has underperformed its peers by more than 400% under its former CEO, Scott Stephenson (2013-2022). At the same time, a poorly designed compensation incentive allowed the insiders to meet their targets at the expense of deteriorating RoICs. A large part of the underperformance stemmed from poor capital allocation strategies, especially Energy (22% of revenue).
Source: TIKR, Bloomberg
Everything was kind of hunky-dory for Verisk until 2014. It was generating a comfortable 25%+ RoIC. But then it got struck by goodwill and intangibles from the 2015 Wood Mackenzie acquisition. Subsequently, a bloated balance sheet couldn't bear the burden of slowing train. Hence you can see the deteriorating metrics from 2015 onwards compared to its glorious past.
Source: Company filings, TIKR
During the 2018 investor presentation, the current CEO Lee Shavel (then CFO) candidly put it in open that large acquisitions (>$100m - Wood Mackenzie, Sequel, PowerAdvocate, G2, and LCI) yielded a measly 4% RoIC. I suspect Wood Mackenzie was the main culprit for such poor returns.
Source: 2018 Investor Day Presentation
The gravity of the situation becomes readily apparent when compared to the outstanding M&A deals (done by previous CEOs) that produced an IRR of 19% before 2015.
…But underneath, we still have an outstanding Insurance business
If we carve out the Insurance business (74% of revenue), we see underneath an exceptional company that has only missed the 7% organic growth mark just by a whisker in 3 out of 12 years.
Source: company filings
II. Why is the situation attractive now?
The following changes within the firm at the behest of the activist shareholder and shareholder outreach program bode well for long-term investors.
Kim Stevenson (from NetApp, Cloudera, and Riverbed Technology) should bring much-needed digital transformation expertise,
III. Jeff Dailey (age 65), 15+ years as the CEO of Farmers Group (a subsidiary of Zurich Insurance Group), should bring his expertise in the transformation of a customer-centric and innovation-focused organization
LTI targets now include the Holy Grail "ROIIC":
…The Patterns are similar to MSCI, S&P Global, and Thomson Reuters
A common playbook is seen at other firms with similar economics, i.e., when they sell non-core assets or change their strategy. Most of the time, it has resulted in an acceleration of organic growth and operating profit expansion. Herein I'm outlining three well-known case studies:
MSCI outperformed
MSCI revamped its internal apps (investing $20 million in 2015) to provide clients with new use cases and contexts. It was triggered by changes to its Board of Directors and a push from an activist shareholder. As a result, MSCI's organic growth rate jumped from 7.5% in 2014 to 10% in 2019, resulting in 1500bps of margin expansion.
Source: Bloomberg, company filings
Thomson Reuters boosted its organic growth and margin expansion after it decided to sell Refinitiv, get a new CEO, and make investments to modernize and overhaul the direction. As a result, organic growth has jumped from 3.7% in 2019 to 6.0% in 2022, and the EBITDA margin from 25.3% to 40%.
Source: Bloomberg, company filings
S&P Global outperformed after the sale of non-core McGraw-Hill in 2012
Source: Bloomberg, company filings
Let's discuss the business now before I tie up the catalysts with the situation
III. BUSINESS SEGMENTS
Verisk reported in three segments until now, but it will be only insurance in the future. Insurance is the largest segment contributing 74% of sales yet 83% of EBITDA.
Since the other businesses will exist no more, I will discuss only the Insurance Segment.
Source: Company Filings
Insurance Segment
Underwriting makes up 70% of the Insurance Segment, while Claims account for 30%. Despite the fact that the Underwriting segment has a high share of decades-old legacy businesses, it continues to grow at a faster CAGR than the Claims solutions.
We have data segregation for these two subsegments (Underwriting & Claims) starting only from 2017, as previously Verisk used to club its entire business (including non-insurance) under Risk Assessment (RA) and Decision Analytics (DA).
Source: Company Filings
Underwriting & Rating Solutions
There are mainly three businesses under it. First, the underwriting business consists of (1) Verisk's flagship ISO business parts (Industry Standard Insurance Programs, Property-Specific Rating, and personal lines underwriting solutions). Second, the Extreme Event is mainly Catastrophe modeling (AIR Worldwide). Lastly, the Software solutions (Like Sequel).
1. Underwriting Solutions (estimated 50% of total revenue, MSD growth):
Verisk stopped reporting the individual data for a few of the above sub-units in 2011 and then in 2017 as a part of segment realignments. Nevertheless, with the risk of not being entirely accurate (but a close approximation), we can construe the historical trend.
Policy Language continues to be the most important, comprising ~38-40% of the entire insurance revenue and growing at a 5% CAGR (2006-2017). P&C sector lags behind in the tech stack. Alarmingly, the insurers have used ancient delivery techniques (such as physical forms, some pdfs) for information flow between Verisk and regulators. This is why Verisk sees so much room for development in areas like dynamic forms and algorithms for loss data, which can enable real-time monitoring.
Statistical data has been a small 2% portion and hence losing relevance. In comparison, the Property-specific rating segment has also dwindled from 17% of revenue in 2006 to 11% in 2017 due to a slower 3% CAGR.
It is worth noting that RMS was acquired by Moody's at a stunning 36x EV/EBITDA multiple in 2021. There are two sources of revenue from this industry – 1) CAT modeling software/analytic offerings and 2) CAT bond issuance. While RMS is the market leader in the steady software offering, Verisk (AIR) is dominant in CAT bond issuance (considered more transactional). Hence Verisk might command a slightly lower multiple (if broken down or taken private), but it still is a marquee asset.
Claims Solutions
This segment provides 1) Claims analytics (fraud detection, loss prediction), 2) Loss quantification tools (Xactware), and 3) Aerial imagery solutions (Geomni).
Source: Company Filings
The business was worth $117m in 2010 and was growing at a CAGR of 50% (2006-2010).
Source: Company Filings
IV. COMPETITIVE LANDSCAPE
Verisk is a recession-resistant mission-critical analytics provider to the P&C industry, with limited competition for most of its products. First, I will discuss fundamental competitive advantages for both segments, followed by the changing landscape.
Underwriting Solutions
How differentiated data assets make it a monopoly
Anyone with an actuarial background can appreciate that insurance companies are very data-intensive. Verisk, for example, maintains a database of over 29.7bn statistical records and processes over 2.6bn insurance records yearly. This makes Verisk one of the world's largest private databases, with 19 petabytes of information. In addition, fascinatingly, a large portion of data feeds comes from its own customers, creating a network effect. No other vendor can provide this (yes, no competition!). Even if someone tries to recreate it, it will be impossible to gain the trust of entire P&C insurers to have them share their data. Trust!
Why Verisk is mission-critical for Insurers
Insurance is one industry where companies don't know the cost of goods sold (loss costs) until after they've sold their policy. And the lost expenses account for ~80% of the annual premium. Hence controlling them is the most vital function to remain profitable. Because of this, Verisk is a mission-critical function for its customers. So, for example, there has not been a single year of negative growth regardless of the economic cycle.
Can an in-house analytics team replace the need for Verisk
All major P&C insurance firms have an in-house analytics team to assess and quantify risk. However, the P&C business is unusual in that no carrier has a substantial market share, which means that no single insurer can construct a complete perspective of all risks using its data. In addition, the P&C industry is known for disparate outcomes. Consequently, being a member of the Verisk ecosystem becomes very advantageous for insurers.
Can startups break this monopoly by circumventing the need to get data from insurers?
Insurtech is a net positive
The P&C market is not immune to the technological shake-up. CB Insights estimates that the top 50 Insurtech cohort has raised more than $11bn in private equity funding since 2017. So if you attend the company's conference calls, there's a high probability of someone asking about the impact of InsurTech on Verisk.
Broadly two categories of InsurTech have emerged over recent years. First, MGAs (Managing General Agents) such as Lemonade, Root, and others that underwrite and distribute insurance (digital insurance experience). InsurTech firms of this sort will likely need data analytics to price and sell insurance in a competitive market efficiently. These businesses would ideally be customers of Verisk and its solutions like LightSpeed (providing a quote in 2 mins with just 2-3 variables).
The second kind of player is in analytics, vying with Verisk to create novel methods to assess risk or fraud. However, they lack the type of data that Verisk possesses and would prefer to collaborate with Verisk. Of course, that would benefit them more than Verisk. However, it is a net good for Verisk since competition has driven conventional customers to use more data and analytics, thereby increasing the TAM.
For context, Verisk added 84 insurers in the last two years and 47 alone in 2021 (the highest in a single year) to its statistical database. A significant chunk came from the InsurTech startups, showing InsurTechs prefer partnering to competing.
Competition with Core processing providers is complimentary
A common pushback for Verisk is whether P&C core system providers like Guidewire, Duck Creek, and Majesco represent a possible threat to Verisk. These core processors are akin to Jack Henry and Fiserv in the banking space and have the advantage of being a platform. This means they can pick and choose third-party modules for their platform. Moreover, these also provide analytics on loss ratios, etc., similar to Verisk. Three of their core activities include Policy, Billings, and Claims. For instance, Guidewire, the biggest of the three, is a leader in Claims, while Duck Creek leads in the policy. It is not uncommon for a single carrier to use multiple vendors.
But there's a profound difference. Verisk provides the entire industry aggregate data (like loss ratios), which goes as input for Guidewire or Duck Creek. And only Verisk has this data. So none of the core processors can gain client acceptance without integrating with Verisk. In fact, Guidewire and Duck Creek take pride in how well their systems are integrated with Verisk so that an insurer can directly pull data through them.
A Core processor for P&C player
Source: Duck Creek
Verisk going offensive with software or getting on the turf of Guidewire/ Majesco
Interestingly, Verisk is becoming more software intense and poaching their turf, albeit currently only in the Life Insurance industry. For example, it acquired Sequel, a policy administrator for the Life Insurance industry. So it will be interesting to watch whether Verisk will expand it into P&C. But the previous CEO was clear that Verisk would be interested only in software for analytic purposes.
"I would only want to see us in software businesses that have the potential to yield data assets that we can use and reuse for analytic purposes. Since the core transaction processing platforms are unlikely to yield any sort of data rights, they're just less likely to be of interest to us."
Previous CEO in 2017
Competition in Property Estimating is moderate
Xactware is the entrenched, dominant player in claims estimation. It competes with Marshall & Swift/Boeckh (MSB) in the replacement cost estimate and Symbility in claims estimation software. A claim estimating platform is at the core of a property insurer's operation, so they would like to partner with a reliable provider. Hence, Symbility could not grow due to its small size despite having better open technology. Moreover, compiling the construction cost data for thousands of line items was the most significant barrier to entry for smaller players like Symbility.
For this reason, Farmers Insurance, the third largest insurer, switched back to Xactware in 2014 within two years of moving to Symbility. But things turned around in 2018 when CoreLogic, which owned MSB, also snapped up Symbility. While replacing a system like Xactware is always disruptive and time-consuming, Verisk will have tough competition ahead.
"It doesn't make sense for us to go against Verisk and say we will be better than Verisk on claims management. They're way too big. They're well-entrenched with ISO, and it's a thing that we say let's focus on underwriting."
V. Financials
TAM: Verisk has grown its revenue at a consistent 7% Y-o-Y organic CAGR since 2007, demonstrating excellent pricing power (+3-5%) with 80% subscription-based revenues. Following a few years of maintaining a market opportunity of $7.5bn, Management increased it to $9bn (domestic $6.5bn) in their 2018 investors' day. This was primarily due to the benefits of the rise of InsureTechs.
Source: 2018 Investor Day
I expect Verisk to increase the TAM in their 2023 Investor Day, considering it has ventured into life insurance and expanded data sets. Even the existing domestic TAM of ~$6.5bn looks conservative, considering Verisk is just 30bps of the entire insurance industry's premium.
The US P&C insurance market is the largest in the world, with over $780bn in yearly premiums. Yet, most P&C insurance companies do not make high RoEs. Because the world's most heavily insured market is also the most competitive. An insurer's return is = investment income + premium growth + disciplined underwriting of risks. With investment income under pressure from low-interest rates (in the last decade) and premiums from the competition, only prudent underwriting can help them remain competitive.
Hence, even though the US P&C Direct Written Premium (DWP) has increased at 3.2% CAGR, Verisk has outgrown with a 7.5% CAGR. A $6.5bn TAM would imply ~80bps of the industry's premium in 2021. Consider Duck Creek (a core processor), which recently pegged its domestic TAM at $6bn and global at $9bn. SwissRe projects total global P&C premiums to more than double from $1.8trn to $4.3trn by 2040. However, Verisk is restricted to the US for P&C, which is also expected to continue growing at a historical rate.
Source: Company filings
Can sustainable growth exceed 7%?
The long case for Verisk is whether it can continue to grow at 7% and above for the foreseeable future or falter.
Below are a few initiatives (software, Life Insurance, international, Telematics, InsurTech) growing at double-digits, which can propel its growth rate for many more years.
Verisk entered the Life Insurance space by paying $194m to acquire FAST at 20x 2019 EBITDA, which can have the potential of making it big. This purchase strengthens its software offerings (similar to the Sequel UK acquisition in 2017). FAST offers policy administration software helping core processing of issuing policies (think Guidewire or Duck Creek in P&C). Moving into the Life Insurance space could be a natural extension considering Verisk estimates 40% of its customers (primarily international) also offer Life Insurance. In addition, the platform provides more room to offer proprietary data analytics solutions such as Voice Analyzer (identifying a smoker via a voice sample) and Avocation models (looking at social media data) for Life Insurance.
"InsurTechs are very insistent that everything be cloud-native, tech-forward. They're not even going to look at you, if you're not that way."
Verisk believes there is a ~$2bn TAM for its non-US insurance businesses. Because the domestic ISO business cannot be replicated elsewhere, Verisk has made software acquisitions such as Sequel (UK), ACTINEO (Germany), and others to create a similar ecosystem of data analytics platforms. Sequel acquisition in 2017 (20x EBITDA) gave Verisk a stronghold in the Lloyd's of London market. It happens to be in London, yet it is a worldwide market. Sequel manages policy, starting from the upfront and then underwriting onto claims. It is growing at a solid ~16% CAGR (the UK also includes Energy), and international has grown at 11.2% CAGR. Other notable acquisitions have been Opta (property intelligence risk) in Canada and ACTINEO (market leader for personal injury claims digitalization) in Germany.
Source: Company filings; international also includes Energy which is decelerating
Furthermore, based on the 3-year average loss ratio and nation score, VRSK classifies attractive existing and growing markets. It has highlighted promising prospects with greater loss rates in established markets such as Germany, China, and the United Kingdom.
Margins
The inherent strength of the "Buy once and sell many times" model supports it with one of the highest EBITDA margins of ~50%. Verisk clocked a high insurance-only margin of 55% in 2021 and 53% for 9M in 2022. However, this will normalize by 200-300bps coming out of 2021 as travel expenses return; cloud investments shift from capex to opex, and investments in high-growth areas (Life insurance, FAST, Sequel, Journaya) with low margins continue. In addition, some 200bps of stranded costs from the divestitures of Energy and Financials will impact the corporate margins. Hence the baseline for 2021 and 2022 margins should be considered 50-51%.
The Management has guided 300-500bps margin expansion by FY24, meaning between 53-56% margins. A closer look at the cost structure indicates that there are a lot of low-hanging fruits (like real estate) for Verisk to optimize. Even Verisk has acknowledged benefits from real estate consolidation and increased usage of global talent optimization locations. In effect, I do not rule out a positive surprise in margin expansion above its guided target of 53-56%.
" We've been looking at our source of geographic locations on capital, which is something that I have a lot of experience with coming from S&P."
Source: Company filings
Is Verisk a highly capital-intensive business?
A common pushback for Verisk is that it is a highly capital-intensive business compared to other similar economic businesses. It ended 2021 at 9% capex, and 2022 will also be near that. However, two significant trends are worth highlighting here.
Source: Company filings
First, historically, you can see that the Capex peaked in 2018 at nearly 10%. This was a function of its aerial imagery investment. It was buying a lot of planes and sensors to develop that dataset. What it did subsequently was to contribute those assets into a partnership with Vexcel. This was smart as it eliminated the capital intensity and the OpEx while allowing it to continue accessing the datasets.
Second, probably five years ago, the insurance industry was a little cautious about the migration to cloud environments. Now that is not an impediment. Verisk was early in recognizing this and started investing earlier in software development. However, this led to high usage of hardware and third-party software. But this is changing as Cloud efficiencies accelerate. For instance, for the first time in more than 30 years, Verisk will no longer own a mainframe server. In addition, by 2023 it will be shutting down all of its data centers as it has migrated fully to AWS. This means that it will shift expenses from Capex to Opex.
Who has the best business model?
Coming back to our moot question of high capex, since GAAP allows allocating expenses between Opex and Capex – the best metric to judge is EBITDA-CAPEX – for an apples-to-apples comparison or to rank a business.
So, in the table below, Verisk is already the 5th best. Over the next 1-2 years, there will be 1) 300-600bps margin expansion guidance, 2) Closure of data centers due to moving to the cloud (reducing Capex), and 3) Energy divestiture. Hence 41% today has the potential to become 48% = 41 (today) +5% (margin expansion guide) + 2% (Capex reduction). It will make it the BEST, just below MSCI (unless SPGI & Moody's also improve from here).
Source: Company filings, TIKR
As a follower of Michael Mauboussin's approach to expectations investing, I look at reverse DCF to see what's priced in the current stock price. Then, I take a traditional method of starting from EBITDA and subtracting all cash expenses to arrive at free cash flows for equity shareholders. Unlike many on the Street, this approach treats stock-based compensation (SBC) as real expenses and simultaneously dilutes the shares. The other benefit of this approach is that it consistently identifies expenses like leases otherwise skipped in valuation shorthands.
Verisk is a simple and easy-to-understand mature company. It currently trades at ~25x times FY'24 EV/FCF multiple. The valuations have never been cheap. For instance, to arrive at the current price, the share price already implies a 7.6% CAGR (FY' 24-31) before moderating to 3%. And that operating profit margins expand from 40% to 46%. This would mean it trades at 20x FCF multiple at the terminal year, and an investor will earn an IRR of 7.1%.
The opportunity is two-fold. Firstly, the organic growth accelerates higher than the price-implied expectations, and secondly, the longevity. Businesses like Verisk are built-to-last kinds of unique assets. Unless there's any considerable potential disruption, they can continue the top line for much more years than my assumption of nine-odd years. For instance, at the end of 10 years, its projected revenue of $4.5bn will still be almost half of S&P Global's current revenue.
"I think one of the challenges we all face is that at the valuation levels that we have right now, it is more challenging to get an attractive return on capital simply by acquiring a great business. We need to find ways to be able to extract value by either accelerating the deployment of those data sets and solutions or by enhancing their new methods."
Commonalities with MSCI and FICO
As discussed in the beginning, MSCI and FICO re-rated substantially (2015-2019) after going through a similar process.
A similar change in MSCI's strategy led to a massive jump of 1400bps in its EBITDA margin. Valuation multiples skyrocketed as buybacks exceeded free cash flows. Moreover, this was during the rise of US Govt yields from 2.2% to 2.7%.
Similarly, FICO re-rated when buybacks started exceeding 100% of its FCF.
In my opinion, Verisk also has a similar potential, albeit with low intensity, as the above two cases. Moreover, a strategic deal to go private or be acquired can command tremendous value since these are non-replicable marquee assets. (e.g., Moody's paid 39x EBITDA for competitor RMS)
Despite having roots in being a non-profit entity (until 1997) and extensive M&A-led growth, Verisk has flourished. This is because it has roots in being an employee-centric firm that brought an owner's mentality. For example, until 2009, Verisk was ~25% owned by its Management and employees. Insurance companies owned the remaining stake.
As a result, the churn in senior managers was minimal. For example, even after the acquisition of AIR Worldwide, its founder Karen Clark (who pioneered the probabilistic catastrophe modeling techniques) continued to head as its CEO and later as a Board member. Similarly, Jim Loveland, the founder of 250-employee privately-held Xactware, remained its President & CEO for over ten years.
During its IPO, it was only Berkshire Hathaway which didn't sell a single share. But the focus shifted after 2011; the insiders sold their shares to holding ~2% of stock now. I believe even Berkshire exited in 2017-18.
However, it is encouraging to see Lee Shavel emphasizing "Owner's Mentality" as the core DNA. Verisk has always attracted high-quality actuaries and data scientist talent. The dependence of regulators on Verisk is deepening even after 50 years since it cannot keep pace with data science or attract talent.
Before I close, did I mention that the CEO gets a special $2m+ one-time bonus if the 3-year TSR exceeds the 65th percentile of the S&P 500 index : )
Investor Day in April 2023 and a host of catalysts discussed above.
I wrote this thesis exclusive for VIC. Very open to pushback from the community. If I don’t get membership activated, I can be reached at www.theinvestingprinciples.com for discussion.
show sort by |
Are you sure you want to close this position VERISK ANALYTICS INC?
By closing position, I’m notifying VIC Members that at today’s market price, I no longer am recommending this position.
Are you sure you want to Flag this idea VERISK ANALYTICS INC for removal?
Flagging an idea indicates that the idea does not meet the standards of the club and you believe it should be removed from the site. Once a threshold has been reached the idea will be removed.
You currently do not have message posting privilages, there are 1 way you can get the privilage.
Apply for or reactivate your full membership
You can apply for full membership by submitting an investment idea of your own. Or if you are in reactivation status, you need to reactivate your full membership.
What is wrong with message, "".