Description

Thesis – Long FNHC, Short HCI

Situation –

Florida is the 3^rd largest state, with 21m residents and is expected to hit 26m by 2030. The Homeowners insurance market is estimated to be ~$9.4b, yet it remains highly fragmented with national players comprising less than 20% of market share and a number of smaller specialists dominating the market. The state backed insurer, Citizens, meanwhile is a fraction of its former size and the current political climate seems to indicate further movement toward more privatization.

Despite non-life insurance companies typically being preferred in a sluggish environment (given lower sensitivity to economic growth and negative correlation to interest rates), FNHC’s Florida focused peer average is down ~20% YTD vs. the SPX up +9%. FNHC shares are down -30% YTD and -25% since Late Feb when the Maison deal was announced. HCI shares are down -20% YTD and -12% since late Feb.

Both FNHC and its peers have seen their shares trade off substantially as Q4’18 results were relatively messy, and what is usually a seasonally strong Q1, was impacted by a freak hailstorm in Central FL. Concerns around increases in next season’s reinsurance renewals were the biggest topic on last Q’s earnings calls. However, the market is ignoring the potential rebound in the sector due to its attractive positioning as a defensive amid a slowing economic environment, and more importantly the positive benefit state legislative reforms. The insurance reform legislation should not only immediately help loss ratios, but will likely also create numerous other positive effects, such as allowing increased growth in South Florida’s most densely populated Tri-County region, potentially lower reinsurance costs in future years and a higher likelihood of M&A given the FL market could start to look attractive again to national carriers.

Long FNHC

FedNat’s shares will rebound as their FL business should start to turn around now that rate hikes have worked their way through FedNat’s book, unprofitable business lines have been fully exited and recently passed AOB[1] reform, translates into lower losses, but also re-opens the door in terms of growing in South Florida. In fact, on the Q1 call, Mgmt explicitly said that these new reforms have renewed their interest in growing their FL book again. Despite being based Broward (the heart of Tri-County), Tri-County is only ~24% of FNHC’s book vs ~40% for HCI.

FNHC’s Mgmt team has demonstrated strong strategy and execution skills in terms of both growing their book outside of Florida while controlling both risk and costs. FNHC successfully expanded to 4 other states and has established productive relationships with AllState and Geico. In Florida, Mgmt optimized FedNat’s book by shedding unprofitable lines along with a considerable amount of exposure to the riskier areas of FL (namely Tri-County which is prone to much higher instances of scams and associated costly litigation), while experiencing only a modest drop in premiums and maintaining a top 5 position in terms of marketshare.[2] Although FedNat typically focuses on hurricane zones 1 and 2 (where need is greatest), the company specifically targets the more modern, higher quality houses which have more advanced hurricane mitigation features.

Mgmt has also focused considerable efforts on cutting OPEX, which have already started to bear fruit, and will likely impress investors once FNHC laps recent deal activity (and related expenses). Meanwhile thanks to OPEX and reinsurance synergies, the Maison acquisition should also provide a further tailwind to FNHC’s expense ratio and ROE. Any dilution to book value from Maison is expected to be earned back over the next 2yrs.

Once the Maison deal is completed, FNHC can also resume repurchases which can be particularly impactful given FNHC trades below book. Meanwhile, if there are not any significant hurricanes in 2019, this would result in substantial gains for FNHC as it can use excess cash to repay debt or repurchase shares. Further, given a market cap of sub $200m and trading below book value, FNHC also represents a decent takeout target for other insurers looking to gain scale or further diversify geographically.

Short HCI

Unlike FedNat, HCI’s mgmt. is far from impressive. They make little effort to hide their strategy of insuring shacks with the lowest values and highest premiums. While this strategy was admittedly a boon during the low storm activity seen prior to 2017, it is not sustainable. Not only is policy retention a major headwind, but this type of strategy also faces major issues in terms of both rates and reinsurance costs once losses begin to escalate. It would not be surprising if HCI continued to see losses creep from recent storms (particularly Irma), as a result of over-indexing to Tri-County and under reserving for losses. HCI has already seen its surplus erode from $270m at YE’16 to ~$220m at YE’18.

HCI’s expansion efforts have been a failure as the company never successfully grew outside of Florida, despite multiple initiatives to underwrite in other various over recent years including TX and AL. Mgmt also failed to expand within Florida once the state backed insurer (Citizen’s) was done shedding most of its more attractive policies. HCI’s Mgmt has effectively taken a bottom feeder approach to its strategic initiatives, making low ball all stock bids for FNHC in late 2017 and again in early 2018. Effectively, they missed the boat for scooping up another insurer during the AOB crisis over the past 3 years. During this time, HCI saw premiums erode as the company suffered declines in policy counts due to attrition of existing customers and inability to generate organic growth within FL.

Without completing a meaningful expansion, HCI has attempted to diversify its business via investments in real estate and technology, both of which Mgmt has seemingly little expertise. The company created a separate segment, Greenleaf Capital, to invest in commercial real estate. But with virtually all of its real estate holdings in in Florida, this still leaves HCI exposed to not only storms but also more sensitive to an economic downturn than its peers. Stripping out the value of the real estate business, HCI trades at a significant premium to FNHC, despite having inferior growth potential and less marketshare.  

Valuation – FNHC is attractively valued relative to both its peer HCI, the rest of the sector as well as historical levels

•          FNHC trades at less than 0.8x P/B ($13.24/$16.98) vs. HCI which trades at ~1.8x P/B ($40.40/$22.37), and vs peer average of ~1.3x[3].

o    This is also significantly below FNHC’s 3yr and 5yr Avg P/B of 1.1x and 1.37x respectively and the 3yr and 5yr peer average of 1.5x and 1.8x respectively

§  Note, on a Price/Tangible Book Basis, FNHC Tang P/B multiple is virtually identical at 0.8x vs. peer average of ~1.8x

o    Adjusting FNHC for their Maison deal results in a Pro-Forma P/B of ~0.9x (based on a PF book value of at least ~$15/share[4]) vs. HCI which excluding its real estate, trades at Pro-Forma P/B ratio of ~2.0x[5]

•           On a P/E Basis, FNHC trades at ~9.4x last year (2018) EPS, and 7.6x 2019e EPS vs. HCI at 16.0x and 12.7x respectively, and its peer average of ~11.5x and 11.0x respectively.

§  On a historical basis, FNHC averaged ~13x Fwd P/E over the last 5 yrs vs. sector at ~11x

o    Adjusting for the real estate business, HCI’s underlying insurance operations trade at a P/E of ~11.3x for 2018 and ~9.0x 2019 respectively.[6]

Other Highlights

•           In terms of exposure, FNHC and HCI relatively balance each other out (with a heavy FL concentration),

o    Admittedly FNHC retains some Gulf exposure particularly now that Maison adds some extra presence in the Gulf and Texas

•           In terms of expense ratio, FNHC’s total expense ratio remains a work in progress at ~39% in Q1’19 vs. ~38% in both 2018 and 2017 as cost cutting efforts have been offset by a smaller premium base (due to the exit of non-HO lines), but FedNat is still achieving a far better ratio than HCI and trends point to this gap continuing to widen over the next few years

o    FedNat has kept the expense ratio below 39% for each of the past 3 Qs and this trend should continue to improve as FNHC grows premiums, laps recent deal activity (and related expenses[7]) and continues to focus efficiency.

§  Looking at FNHC’s core Homeowners segment shows a more drastic improvement, FedNat’s HO expense ratio was 36.7% in Q1’19 an improvement versus 40.3% in 2018 and 40.5% in 2017.

o    HCI posted a net expense ratio 47.6% in Q1’19, up from 43.5% a year ago (YoY increase due to the lower Net premiums earned) and 36.6% in Q1’17/2yrs ago.

§  For 2018, HCI reported an expense ratio of 44.5%, vs 42% in 2017

o    In terms of other similar residential focused P&C peers that focus on FL, HRTG was ~38% in 2018, UIHC was ~45% and UVE was 33.4%

§  Peer Avg Expense ratio was ~40.3% in 2018, 41.1% in 2017, and ~35% in ’14 and ‘15

•           HCI currently yields ~4.0% vs. FNHC ~2.6% and peer average of ~2.6%.

o    However longer term HCI’s dividend appears far less stable and has less room for growth given shrinking premiums and surplus (see below for more detail) along with a 1.6x coverage ratio vs. FNHC’s 3.7x

•           FNHC has superior management, with current CEO Michael Braun well regarded in the industry.

o    HCI’s CEO is seen as somewhat controversial and has a reputation for relying more on financial engineering, than quality underwriting standards and conservative reserving.

§  In the past HCI has found ways to keep its share price levitated but the recent spat of storms combined with legacy AOB issues may be the event that causes the truth to be told.

FNHC Background – Federated National or “FedNat” (ticker FNHC) is a residential P&C insurer focused on single family homes in the state of FL. The company also has a growing presence in the Gulf (namely TX and LA), as well as in SC. FNHC recently announced the acquisition of Maison, the P&C insurance segment from 1347 Property Insurance Holdings (ticker PIH) earlier this year.[8] FedNat also acquired the remainder JV interest in Monarch[9] insurance in a deal announced in late 2017. FNHC’s core demo is the more modern (built after 1994), higher quality houses which have more advanced wind/hurricane mitigation features, as well as lower all other peril (ie non-CAT) losses. FedNat has ~20% of homes in this class statewide. The company utilizes both a linear model and manually reviews every bound risk to ensure accuracy of underwriting criteria.

FNHC has a market cap of ~$190m and writes about ~$600m in premiums and earns ~$350m. Florida Homeowners (HO) is ~80% of the company’s total premiums, while Non-Florida HO is ~15% of total and the remainder is non-HO lines. Once completed, the Maison acquisition will add ~15% in total premiums to FNHC and result in ~30% of the PF Homeowner’s book being outside of FL (specifically adding exposure in TX and LA, along with ~7k policies in FL). Mgmt expects the deal to increase ROE ~3-5ppts and sees EPS accretion of +10% as “very achievable”.

In addition to expanding outside of FL, FedNat is in the process of exiting its non-Homeowners lines, specifically its Automobile and commercial general liability lines, due to poor profitability (both of which they announced they would exit in Q4’17). Because of overhead expenses and adverse development, these two segments remain a drag on profitability as they are unwound, but this is expected to be over by YE’19. Once completed, FNHC will be a pure play HO P&C insurance company[10].

Cutting OPEX has been a constant theme over past year or so. FNHC sought to cut costs and boost profitability following the exit of unprofitable non-HO lines and after rejecting 2 bids from HCI. In 2018, FNHC cut 105 positions (~25% of staff) generating over $6.0m in annual savings as a result of Mgmt’s initiatives to maximize operational efficiencies. PF Maison, Mgmt expects ~$5m in reinsurance synergies and OPEX save of 25% of Maison's G&A (Mgmt estimated Opex save of ~$2.5m in the first 12 months, increasing to over $3.3m post 2020). The recent refinancing (with 10yr notes), not only gave a longer horizon but also lowered their weighted avg. interest rate by 170bps and provided FNHC with more favorable debt covenants.

The exiting of the non-HO businesses has also weighed on overall premium growth over the past ~2yrs. FedNat’s total gross written premiums are down vs 2016, but its core Homeowner segment has continued to increase premiums increase thanks to growth outside of Florida (which has seen ~50-75% YoY growth over the past few quarters) offset by LSD declines in the FL HO segment as FedNat has shed riskier policies in areas plagued by fraud.

•             For 2018, FNHC’s total Gross written premiums were $568m vs. $603m in 2017 and $605m in 2016

o    However, excluding the exited businesses (Auto and CGL) total gross premiums were $554m in ’18 vs. $548m in ’17 and $522m in ’16[11]

o    Thanks to growth Ex-FL, the core HO segment saw premiums increase to $540m in ‘18 vs. $537m in ‘17 and $512m in ‘16

§  Non-FL HO premiums were ~$81m in ‘18 vs. $54m in ’17 and $35m in ’16

o    FedNat’s Florida HO business saw premiums decline to $459m vs. $482 and $477m, but it's important to note that this is due in large part to Mgmt opting to shed riskier policies, partly offset by rate hikes on the remaining policies

•            As of Q1’19, FedNat’s FL book is continued to shrink, down to ~240k policies representing ~80% of total policies vs. 265k policies a year ago (then representing ~89% of total policies in force) and 273k 2yrs ago

o    Mgmt noted that over the last 3yrs, they've dropped well in excess of 40k policies in FL and dropped $20b of exposure, yet said their premiums in Florida only dropped by ~$10m.  

o    The impact on premiums from the decline in policies was partly offset by rate hikes, as FedNat has taken an aggregate 21% compounded homeowners rate increase since 2017 including took a 10% rate hike in 2017, ~5.2% in 2018 and another 4.6% increase which took effect in April 2019.

o    Shedding these policies has helped FNHC, as they recorded favorable prior year development in Q1 of ~$1.6m per Q.

§  Recall peers such as UIHC and UVE saw unfavorable development lead to large write downs (and declines in their share prices) over the past two Qs.

HCI Background: HCI, based in Tampa, is a miniature sized P&C insurer (market cap <$400m) with aspirations of becoming a palm tree version of Berkshire Hathaway. Its principal subsidiary, Homeowners Choice Property & Casualty Insurance Company, Inc. ("HCPCI") provides P&C insurance to single family homeowners in Florida. The company also offers existing policy holders new products like its recent flood endorsement. HCI, through its GreenLeaf Capital segment, owns office and retail space including two shopping centers; and, owns and operates one full-service restaurant and two marinas. In addition, the HCI dabbles in cloud software, most of which is focused on geographic type functionality (for industries such as insurance for purposes of compiling property information, mapping claims and generating visualizations and similar reports.)

Essentially HCI has benefitted from both timing coupled with an aggressive strategy, as they entered the FL HO market at the right time, with the company being established in 2006 and IPO’ing in 2008, which was right after FL ended a brutal streak of tropical storms[12]. Meanwhile this storm activity had caused many of the larger national insurers to either scale back or leave the state entirely (State Farm, Allstate, Liberty Mutual’s Safeco), ceding market-share to smaller regional players. Furthering the opportunity to gain share, it was not long after this that the state began looking to shrink the burden on taxpayers from its State-run Insurer Citizens (which was created in 2002 and had swelled to 1.5m policies by 2006). In 2013, the FL state legislature passed a bill intended to reduce the insurance policy count of Citizens, by shifting business toward private insurers (effectively allowing private insurers to takeover policies if they could provide similar coverage within 15% of the price offered by Citizens.)

HCI began doing take-outs of single family homeowner (HO) policies from Citizen’s in 2007 and continued until 2014. Substantially all of HCPCI’s premium revenue since inception has come from the policies acquired in these and other assumption transactions and their subsequent renewals[13]. Now little opportunity remains for acquiring decent quality business in the state as the market has shifted considerable over the past decade.[14] Citizens now stands at only a quarter of its size compared to 2012, and effectively only retains the riskiest policies and peer insurers have looked to cull the fat off their books by increasing rates on higher risk areas such as the South FL Tri-County area which has been ravaged by fraudulent claim activity.

HCI is transparent in terms of its strategy of looking to insure shacks with highest premium and lowest value, which means they outperform when there are no storms but suffer when weather turns against them as these homes are more vulnerable and thus typically take larger relative losses. This strategy has been a boon given the low storm activity seen in recent years, but it can pose a major issue for rates and reinsurance once losses begin to escalate. Prior to 2017 and Irma, Florida storm activity had been quite favorable, with Hurricanes Mathew and Hermine the only major hurricanes in the ~10 years post 2005/Wilma. As a result, reinsurance costs were also favorable, as they had been held down by low rates and an influx of capital.

With take-outs no longer fueling growth, Mgmt. expected M&A to drive growth in 2017. They made two bids for FNHC in Q4’17 and again in Q1’18. After these bids were rejected, Mgmt changed its tune, and said on its Q2’18 call that HCI was going to “re-concentrated” on the insurance biz in Florida, while touting their “TypTap” (data) technology as giving them an edge over competitors particularly in terms of avoiding the AOB minefields in Tri-County. They also said they would focus more on expanding outside of Florida and expanding their flood insurance business. The lack of M&A was blamed on price, as the CEO said that a gap remained between what sellers were asking and what any rationale buyer would be willing to pay.

Since 2017 premiums have continued to shrink at a low single digit rate. In Q1’19, gross premiums earned were  -10% below their level 2yrs ago, while net premiums eared were ~19% below Q1’17 levels. As a result, HCI’s expense ratio, has deteriorated to 44.5% in 2018, vs the low ~30% seen back in 2014 and 2015.

Capital allocation is a key issue for HCI. Lacking scale, the company needs to do an acquisition to offset their shrinking book. Yet the “gap” in M&A pricing that Mgmt referenced is unlikely to narrow now that AOB reform has passed, and thus HCI is more likely to overpay given they missed their window for a cheap acquisition pre-AOB reform. Meanwhile the sustainability of HCI’s capital return is starting to look questionable as surplus continues to shrink. HCI announced a $20m repurchase plan on Dec 4th 2018 and pays out ~$11m annually in dividends. HCI had stat surplus of $220m at YE’18 and given HCI’s insurance sub paid out $20m in dividends during Q1’19 and a Q1 combined ratio was just north of 100%, surplus was likely sub ~$200m in Q1’19 vs. $273m at YE’16 (Pre-Irma).

Risks –

•           Hurricanes

o    FNHC is exposed to storms and hurricanes impacting the Gulf Coast

§  Even in Texas, FNHC is more concentrated on coastal areas, with plans to target mostly Tier 1 and Tier 2 coastal counties, especially in and around the Houston area.

•           No Hurricanes don’t guarantee success either, as 25 Florida insurers failed from 2004-2014, with 13 failing between 2007-2011 (a hurricane free period, including six of 18 companies approved for take-outs and Seven non-takeout companies)

o    FL insurers virtually all choose to avoid AM best ratings and instead opt for Demotech ratings, an agency that is widely criticized as being too lenient and which basically grants every insurer the same A or A+ rating.

§  By contrast another small ratings agency, Weiss Ratings routinely assigns these very same companies grades of C, C-, C+ and D.

•           Non-Cat losses are a particular threat

o    AOB scams and related litigation around inflated claims have lead to adverse reserve development over recent years

§  As seen by both UIHC and UVE over the past two Qs

o    “Non-PCS” events such as the recent Brevard County Hailstorm, while not devastating, can ruin what is typically a good Q seasonally and weigh on both earnings/surplus.

§  Note, similar to AOB in 2014-2018, Sinkholes were a particular problem for the FL P&C industry in the late 2000s and early 2010s, while the sinkhole mini-crisis was remedied by changes to insurance contracts, it illustrates the potential threat of less ‘obvious’ risks in the P&C business. Especially given that such risks also can be typically concentrated in certain areas (as Sinkholes were a phenomenon that occurred predominately on the Gulf coast and in parts of Central FL)

o    Reducing TriCounty exposure due to AOB has also had other secondary effects, as FNHC had a higher concentration in Brevard County. Likewise, Hurricane Michael really hit FNHC, as they have 10% marketshare in the Panhandle vs. 5% in state overall

Catalyst

• FNHC's Closing of Maison Deal followed by Share repurchases below book value 
• Continued unfavorable reserve development by HCI
• 2019 Reinsurance Renewals and Hurricane Season 
• Trade war and macro economic concerns persisting (leading to more attention toward defensively positioned sectors such as P&C)
  • Increased coverage by Sell side 
  • M&A activity in the sector