Tag Archives: PMLs

Lancashire finds the love

After going ex-dividend in November, investors went mega bearish on Lancashire (LRE.L) when it nearly dropped below the 500p level, as the graph below shows. A previous post highlighted the reasons behind the change in sentiment over the first half of 2014 on the once darling of the specialty insurance sector.

click to enlargeLancashire Insurance Group 2014 Share Price

The firm released its Q4 today and announced another special dividend of $0.50 on top of the regular $0.10 dividend. Driven by stable results, as per the graph below, and by the chatter that Lancashire could be an M&A target, the price today reflects a respectable 160% multiple to diluted tangible book. It was odd that although the firm’s executives joked about having prepared an answer to the M&A question, no analyst actually asked the question in the conference call today!

click to enlargeLancashire Historical Combined Loss 2008 to 2014

One of the big positives from the call today was the news that the firm has restructured their reinsurance programme that protects their book to give them more event coverage with reinstatements (away from previous aggregate cover). This provides more protection to Lancashire from multiple events. The PMLs as at January expressed as a percentage of the calendar year earned premiums (estimated figures for 2015) show the reduced net risk profile of this arbitrage strategy.

click to enlargeLancashire PMLs January 2015

It’s nice to see Lancashire recover some of its shine and it will be intriguing to see if it does become an M&A target in the coming months.

Lancashire is looking unloved

With exposure adjusted rates in the specialty insurance and reinsurance sector continually under pressure and founder/former CEO, Richard Brindle, making an unseemly quick exit with a generous pay-out, Lancashire’s stock has been decidedly unloved with the price trading well below the key £7 threshold highlighted in my last post on the subject in February. Although we remain in the middle of the US hurricane season (and indeed the Napa earthquake is a reminder that its always earthquake season), I thought it was a good time to have a quick look over Lancashire’s figures again, particularly as the share price broke below the £6 threshold earlier this month, a level not seen since early 2011. The stock has clearly now lost its premium valuation compared to others in the London market as the graph below shows.

click to enlargeLondon Market Specialty Insurers Tangible Book Value Multiples August 2014

Results for H1-2014, which include full numbers from the November 2013 acquisition of Cathedral, show a continuing trend on the impact of rate reductions on loss ratios, as per the graph below.

click to enlargeLancashire Historical Combined Loss 2006 to H12014

The impact of the Cathedral deal on reserve levels are highlighted below. The graph illustrates the consistent relative level of IBNR to case reserves compared to the recent past which suggests a limited potential for any cushion for loss ratios from prior year reserve releases.

click to enlargeLancashire Historical Net Loss Reserves

The management at Lancashire have clearly stated their strategy of maintaining their discipline whilst taking advantage of arbitrage opportunities “that allow us to maintain our core insurance and reinsurance portfolios, whilst significantly reducing net exposures and enhancing risk adjusted returns”. In my last post, I looked at post Cathedral gross and net PMLs as a percentage of earned premiums against historical PMLs. More applicable figures as per July for each year, against calendar year gross and net earned premiums (with an estimate for 2014), are presented below. They clearly show that the net exposures have reduced from the 2012 peak. It is important to note however that the Gulf of Mexico net 1 in 100 figures are high at 35%, particularly compared to many of its peers.

click to enlargeLancashire PMLs July 2010 to July 2014

There is of course always the allure of the special dividend. Lancashire has indicated that in the absence of attractive business opportunities they will look at returning most, if not all, of their 2014 earnings to shareholders. Assuming the remainder of 2014 is relatively catastrophe free; Lancashire is on track to make $1-$1.10 of EPS for the full year. If they do return, say, $1 to shareholders that represents a return of just below 10% on today’s share price of £6.18. Not bad in today’s environment! There may be a short term trade there in October after the hurricane season to take advantage of a share pick-up in advance of any special dividend.

Others in the sector are also holding out the prospect of special dividends to reward patient shareholders. The fact that other firms, some with more diverse businesses and less risky risk profiles, offer potential upside through special dividends may also explain why Lancashire has lost its premium tangible book multiple, as per the first graph in this post.

Notwithstanding that previously Lancashire was a favorite of mine due to its nimble and focused approach, I cannot get past the fact that the sector as a whole is mired in an inadequate risk adjusted premia environment (the impact of which I highlighted in a previous post). In the absence of any sector wide catalyst to change the current market dynamic, my opinion is that it is expedient to pass on Lancashire here, even at this multi-year low.

The game of chicken that is unfolding across this sector is best viewed from the side-lines in my view.

Arthur opens the US Hurricane Season

After Hurricane Arthur briefly made landfall in North Carolina on Thursday night, a weakened storm is now heading north. I thought this would be good time to have a look at the probable maximum losses (PMLs) published as at the Q1 2014 results by a sample of specialist (re)insurers, first presented in a post in June 2013. That post went into some detail on the uncertainties surrounding the published PMLs and should be read as relevant background to the figures presented here.

Despite predictions of an above average 2013 Atlantic hurricane season, the number of named hurricanes was the lowest since 1982. Predictions for the 2014 season are for a below average number of hurricanes primarily due to cooler sea temperatures in the Atlantic due to the transition to El Niño (although that is now thought to be slower than previously anticipated). The graph below includes the 2014 predictions.

click to enlargeHistorical Atlantic Storms & Hurricanes I like to look at PMLs as a percentage of net tangible assets (NTA) on a consistent basis across firms to assess exposures from a common equity viewpoint. Many firms include subordinated debt or other forms of hybrid debt in capital when showing their PMLS. For example, Lancashire has approximately $330 million of sub-debt which they include in their capital figures and I have show the difference with and without the sub-debt in the percentages for Lancashire in the graph below on US wind PMLs to illustrate the comparison.

Whether hybrid debt comes in before equity or alongside equity depends upon the exact terms and conditions. The detail of such instruments will determine whether such debt is classified as tier 1, 2 or 3 capital for regulatory purposes under Solvency II (although there are generous transitional timeframes of up to 10 years for existing instruments). The devil is often in the detail and that is another reason why I prefer to exclude them and use a consistent NTA basis.

As per the June 2013 post, firms often classify their US wind exposures by zone but I have taken the highest exposures for each (which may not necessarily be the same zone for each firm).

click to enlargeUS Wind PMLs Q1 2014 These exposures, although expressed as percentages of NTAs, should be considered net of potential profits made for 2014 to assess the real impact upon equity (provided, of course, that the expected profits don’t all come from property catastrophe lines!). If for example we assume a 10% return on NTA across each firm, then the figures above have to be adjusted.

Another issue, also discussed in the previous post, is the return period for similar events that each firms present. For example, the London market firms present Lloyds’ realistic disaster scenarios (RDS) as their PMLs. One such RDS is a repeat of the 1926 Miami hurricane which is predicted to cost $125 billion for the industry if it happened today. For the graph above, I have assumed a 1 in 200 return period for this scenario. The US & Bermudian firms do not present scenarios but points on their occurrence exceedance probability (OEP) curves.

As it is always earthquake season, I also include the PMLs for a California earthquake as per the graph below.

click to enlargeCalifornia EQ PMLs Q1 2014 In terms of current market conditions, the mid-year broker reports are boringly predictable. John Cavanagh, the CEO of Willis Re, commented in their report that “the tentacles of the softening market are spreading far and wide, with no immediate signs of relief. We’ve seen muted demand throughout 2014 and market dynamics are unlikely to change for some time to come. The current market position is increasingly challenging for reinsurers.” Aon Benfield, in their report, stated that “the lowest reinsurance risk margins in a generation stimulate new growth opportunities for insurers and may allow governments to reduce their participation in catastrophe exposed regions as insurance availability and affordability improves”. When people start talking about low pricing leading to new opportunities to take risk, I can but smile. That’s what they said during the last soft market, and the one before that!

Some commentators are making much of the recent withdrawal of the latest Munich Re bond on pricing concerns as an indicator that property catastrophe prices have reached a floor and that the market is reasserting discipline. That may be so but reaching a floor below the technical loss cost level sounds hollow to me when talking about underwriting discipline.

To finish, I have reproducing the graph on Flagstone Re from the June 2013 post as it speaks a thousand words about the dangers of relying too much on the published PMLs. Published PMLs are, after all, only indicators of losses from single events and, by their nature, reflect current (group) thinking from widely used risk management tools.

click to enlargeFlagstone CAT losses Follow-on: It occurred to me after posting that I could compare the PMLs for the selected firms as at Q1 2014 against those from Q1 2013 and the graph below shows the comparison. It does indicate that many firms have taken advantage of cheap reinsurance/retrocession and reduced their net profiles, as highlighted in this post on arbitrage opportunities. Some firms have gone through mergers or business model changes. Endurance, for example, has been changed radically by John Charman (as well as being an aggressive buyer of coverage). Lancashire is one of the only firms whose risk profile has increased using the NTA metric as a result of the Cathedral acquisition and the increase in goodwill.

click to enlargeUS Wind PMLs Q1 2013 vrs 2014

Lancashire…so much to answer for.

My bearishness on the reinsurance and specialty insurance sector is based upon my view of a lack of operating income upside due to the growing pricing pressures and poor investment income. I have posted many times (most recently here) on the book value multiple expansion that has driven valuations over the past few years. With operating income under pressure, further multiple expansion represents the only upside in valuations from here and that’s not a very attractive risk/reward profile in my view. So I am happy to go to the sidelines to observe from here.

So, what does this mean for my previously disclosed weak spot for Lancashire, one the richest valued names in the sector? Lancashire posted YE2013 results last week and disappointed the market on the size of its special dividend. As previously highlighted, its Cathedral acquisition marked a change in direction for Lancashire, one which has confused observers as to its future. During the conference call, in response to anxious analysts, management assured the market that M&A is behind it and that its remains a nimble lead specialist high risk/high return underwriter dedicated to maximising shareholder returns from a fixed capital base, despite the lower than expected final special dividend announced for 2013.

The graph below illustrates the past success of Lancashire. Writing large lead lines on property, energy, marine and aviation business has resulted in some astonishingly good underwriting returns for Lancashire in the past. The slowly increasing calendar year combined ratios for the past 5 years and the lack of meaningful reserve releases for the past two year (2013 even saw some reserve deterioration on old years) show the competitive pressures that have been building on Lancashire’s business model.

click to enlargeLancashire Combined Ratio Breakdown 2006 to 2013

The Cathedral acquisition offers Lancashire access to another block of specialist business (which does look stickier than some of Lancashire’s business, particularly on the property side). It also offers Lancashire access to Lloyds which could have some capital arbitrage advantages if Lancashire starts to write the energy and terrorism business through the Lloyds’ platform (as indicated by CEO Richard Brindle on the call). Including the impact of drastically reducing the property retrocession book for 2014, I estimate that the Cathedral deal will add approx 25% to GWP and NEP for 2014. Based upon indications during the call, I estimate that GWP breakdown for 2014 as per the graph below.

click to enlargeLancashire GWP Split

One attractive feature of Lancashire is that it has gone from a net seller of retrocession to a net buyer. Management highlighted the purchase of an additional $100 million in aggregate protection. This is reflected in the January 1 PML figures. Although both Lancashire and Cathedral write over 40% of their business in Q1, I have taken the January 1 PML figures as a percentage of the average earned premium figures from the prior and current year in the exhibit below.

click to enlargeLancashire PMLs January 2010 to January 2014

The graphs above clearly show that Lancashire is derisking its portfolio compared to the higher risk profile of the past two years (notably in relation to Japan). This is a clever way to play the current market. Notwithstanding this de-risking, the portfolio remains a high risk one with significant natural catastrophic exposure.

It is hard to factor in the Cathedral results without more historical data than the quarterly 2013 figures provided in the recent supplement (another presentation does provide historical ultimate loss ratio figures, which have steadily decreased over time for the acquired portfolio) and lsome of the CFO comments on the call referring to attritional loss ratios & 2013 reserve releases. I estimate a 68% combined ratio in 2014, absent significant catastrophe losses, which means an increase in the 2013 underwriting profit of $170 million to $220 million. With other income, such as investment income and fee income from the sidecar, 2014 could offer a return of the higher special dividend.

So, do I make an exception for Lancashire? First, even though the share price hasn’t performed well and currently trades around Stg7.30, the stock remains highly valued around 180% tangible book.  Second, pricing pressures mean that Lancashire will find it hard to make combined ratios for the combined entities significantly lower than the 70% achieved in 2013, in my view. So overall, although Lancashire is tempting (and will be more so if it falls further towards Stg7.00), my stance remains that the upside over the medium term does not compensate for the potential downside. Sometimes it is hard to remain disciplined……

CAT models and fat tails: an illustration from Florida

I have posted numerous times now (to the point of boring myself!) on the dangers of relying on a single model for estimating losses from natural catastrophes. The practise is reportedly widespread in the rapidly growing ILS fund sector. The post on assessing probable maximum losses (PMLs) outlined the sources of uncertainty from such models, especially the widely used commercial vendors models from RMS, AIR and EqeCat.

The Florida Commission on Hurricane Loss Projection Methodology (FCHLPM) was created in 1995 as an independent panel of experts to evaluate computer models used for setting rates for residential property insurance. The website of the FCHLPM contains a treasure trove of information on each of the modelling firms who provide detailed submissions in a pre-set format. These submissions include specifics on the methodology utilised in their models and the output from their models for specified portfolios.

In addition to the three vendor modellers (RMS, AIR, EqeCat), there is also details on two other models approved by FCHLPM, namely Applied Research Associates (ARA) and the Florida Public Hurricane Loss Model (FPHLM)developed by the Florida International University.

In one section of the mandated submissions, the predictions of each of the models on the number of annual landfall hurricanes for a 112 year period (1900 to 2011 is the historical reference period) are outlined. Given the issue over the wind speed classification of Super-storm Sandy as it hit land and the use of hurricane deductibles, I assume that the definition of landfall hurricanes is consistent between the FCHLPM submissions. The graph below shows the assumed frequency over 112 years of 0,1,2,3 or 4 landfall hurricanes from the five modellers.

click to enlargeLandfalling Florida Hurricanes

As one of the objectives of the FCHLPM is to ensure insurance rates are neither excessive nor inadequate, it is unsurprising that each of the models closely matches known history. It does however demonstrate that the models are, in effect, limited by that known history (100 odd years in terms of climatic experiences is limited by any stretch!). One item to note is that most of the models have a higher frequency for 1 landfall hurricane and a lower frequency for 2 landfall hurricanes when compared with the 100 year odd history. Another item of note is that only EqeCat and FPHLM have any frequency for 4 landfall hurricanes in any one year over the reference period.

Each of the modellers are also required to detail their loss exceedance estimates for two assumed risk portfolios. The first portfolio is set by FCHLPM and is limited to 3 construction types, geocodes by ZIP code centroil (always be wary of anti-selection dangers in relying on centroil data, particularly in large counties or zones with a mixture of coastal and inland exposure), and specific policy conditions. The second portfolio is the 2007 Florida Hurricane Catastrophe Fund aggregate personal and commercial residential exposure data. The graphs below show the results for the different models with the dotted lines representing the 95th percentile margin of error around the average of all 5 model outputs.

click to enlarge

Modelled Losses Florida Notional Residential PortfolioModelled Losses FHCF Commercial Residential Portfolio

As would be expected, uncertainty over losses increase as the return periods increase. The tail of outputs from catastrophe models clearly need to be treated will care and tails need to be fatten up to take into account uncertainty. Relying solely on a single point from a single model is just asking for trouble.