PIH - 1347 Property Insurance Holdings

[muscleman]

Hi Aman1,

    Thank you! I think my math is off because I should have assumed the 124% is 1/8 of a two year horizon instead of 1/2, because they have hit the $4 M reinsurance cap, so future claims won't affect the next few quarters.

    However, I think their reported 75% combined ratio excluding the catastrophe is artificially low. Please take a look at schedule p part 3 on page 34 that I attached below. For the 2014 losses, only $2.7 M were paid in that year, and the cumulative losses for 2014 claims came to $3.4 M in 2015. This is quite common among insurers and nothing wrong for them to take time to process these payments, but this also means the 85% combined ratio excluding the catastrophe is lower than the actual number. If we assume something similar would happen. 3.4 / 2.7 = 125%, so there might be another 25% not yet paid. 2016 Q1's losses are  6.6 M. Excluding the $4 M catastrophe, the losses are 2.6 M. 2.6 * 125% = 3.3 M. So I think the actual combined ratio excluding catastrophe is 81%. If there is one catastrophe every two years, and it always hits the $4 M limit, then the long term average combined ratio would have been 124 * 1/8 + 81 * 7/8 = 86%.

 

    Note that there are some other caveats to this calculation. If you look at page 5 here, you will notice that the combined ratio is usually a lot higher in Q2 and Q3 than Q1.

    http://www.t11capital.com/wp-content/uploads/2016/02/T11-PIH-research.pdf

 

    I found that last year's combined ratio for the  whole year was 91%, and there was a $1.5 M catastrophe payment. Last year's non-catastrophe payment was 9.9-1.5 = 8.4 M. Assuming they actually needed to pay 8.4 * 125% = 10.5 M, and last year's net premium earned was 25.9 M, the adjusted combined ratio would be (10.4 + 6.5+7.2) / 25.9 = 93% excluding the catastrophe. I think this may be a more realistic non-catastrophe combined ratio. We know that catastrophes may not always hit the $4 M limit, but if we just assume so, and we use that figure, then 124 * 1/8 + 93 * 7/8 = 96.8% would be the average long term combined ratio.

 

    I agree that with these conservative assumptions, the combined ratio is still below 100%, so it may still be a good business. When do you plan to meet the management? Could you please take the math to them and let them comment on that?

 

    Also please let me know if any part of my math is wrong.

 

 

 

Thank you!

 

2015_Annual_Statement_-_signed_jurat.pdf

[Guest MarkS]

Prior to meeting with management you might read these links.  Clearly weather plays a part in the high rates. But weather alone can't explain why we also have some of the highest cost medical insurance, worker's compensation insurance and auto insurance. Prior to our most recent insurance commissioner, the previous three commissioners all went to prison for various crimes. We have elected public officials while they were actually sitting in a jail cell.  I remember one particular race for governor. Our choice was between a known criminal and a former Grand Wizard of the Klu Klux Klan. People drove around with bumper stickers that said "Elect the crook, it's important."  When I compared Louisiana to a banana republic, I actually meant it.

Thanks

Mark

 

http://theadvocate.com/news/neworleans/neworleansnews/15535868-63/unitedhealth-pulling-out-of-obamacare-coverage-in-louisiana-in-2017-affecting-nearly-29000-customers

 

 http://www.usagencies.com/blog/louisiana-drivers-paying-some-of-highest-auto-insurance-rates/

 

 http://theadvocate.com/news/acadiana/13748683-148/commissioner-race-is-about-louisianas

 

 http://www.claimsjournal.com/news/southcentral/2015/02/11/261507.htm

 

 http://www.lpb.org/index.php/publicsquare/topic/09_11_-_extreme_premiums_louisiana_insurance_rates/

 

[Aman1]

Thanks Mark, very useful - I appreciate the links, just having a read now.

 

Muscleman - hard to be precise on the numbers because the big x-factor of weather means all our numbers are prone to wild misses. I think the most valuable take-away is that there's pretty clear logic as to why it shouldn't trade below book value.

[Guest MarkS]

The Harbert familly of Birmingham, Alabama, just filed a 13D. The "purpose" section is interesting reading.

http://www.sec.gov/Archives/edgar/data/1591890/000091957416013418/d7158484_13d-a.htm

Thanks

Mark

[Aman1]

It's nice to see someone else agree with the investment thesis and buy a sizable position.

 

I'm not particularly impressed with the activism points they've raised to be honest - everything listed is basically the company's existing strategy, and in Larry Swets Jr of KFS and Gordon Pratt as Chairman (who is also a sizeable shareholder), they already have shareholder champions in meaningful board positions. The governance issue that could be raised is the deal to end the KFS management services agreement, however the terms weren't particularly terrible on PIH in my view as a minority shareholder. It's worth flagging that the recent KFS annual letter noted that if PIH did not find uses for their capital, they would push for capital returns (buybacks are a no-brainer here).

 

The Harbert 13D comes across more as a value investment parading as an activist position - maybe it's something the fund managers can write about as a case study down the road, as the odds are that ultimately the shares will trade materially higher in a few years and they can point to any acts (such as a buyback or M&A) as something they influenced, when really it looks like the standard plan as things stand.

 

Personally, I like and respect the PIH management - they are now in two markets (LA and TX) and are likely to grow policy count and revenues this year by 40-45% while writing profitable business and generating decent book growth over the prior 3 years. A Florida entry is inevitable, and because of their big base of contacts with agents in FL, it would make sense to grow organically there to maximize return on capital. It would also make sense to wait for a catastrophe to harden rates before entering that market after a few softer years - so I'm less interested in seeing them rush to make an acquisition here just for the sake of spending cash that is burning a hole in the pocket.

 

The only thing I would do that is different from the current stated strategy is take $10-15m of the company's $27m unrestricted cash and use it to tender for shares, this would allow them to retire 20-30% of outstanding shares and would be highly accretive to shareholders (ideally bought below book value) while leaving decent growth capital in place. That said, I'm not overly worried about management either squandering or hoarding cash over the medium term; this is a sensible management and shareholder friendly board that I'm sure will deliver value over the medium term.

 

[muscleman]

Thanks Mark, very useful - I appreciate the links, just having a read now.

 

Muscleman - hard to be precise on the numbers because the big x-factor of weather means all our numbers are prone to wild misses. I think the most valuable take-away is that there's pretty clear logic as to why it shouldn't trade below book value.

 

 

 

Have you met management?

 

I wonder what risks are hidden in those issues MarkS pointed out. What do you think?

 

[muscleman]

Prior to meeting with management you might read these links.  Clearly weather plays a part in the high rates. But weather alone can't explain why we also have some of the highest cost medical insurance, worker's compensation insurance and auto insurance. Prior to our most recent insurance commissioner, the previous three commissioners all went to prison for various crimes. We have elected public officials while they were actually sitting in a jail cell.  I remember one particular race for governor. Our choice was between a known criminal and a former Grand Wizard of the Klu Klux Klan. People drove around with bumper stickers that said "Elect the crook, it's important."  When I compared Louisiana to a banana republic, I actually meant it.

Thanks

Mark

 

http://theadvocate.com/news/neworleans/neworleansnews/15535868-63/unitedhealth-pulling-out-of-obamacare-coverage-in-louisiana-in-2017-affecting-nearly-29000-customers

 

http://www.usagencies.com/blog/louisiana-drivers-paying-some-of-highest-auto-insurance-rates/

 

http://theadvocate.com/news/acadiana/13748683-148/commissioner-race-is-about-louisianas

 

http://www.claimsjournal.com/news/southcentral/2015/02/11/261507.htm

 

http://www.lpb.org/index.php/publicsquare/topic/09_11_-_extreme_premiums_louisiana_insurance_rates/

 

I understand that more people suing causes auto insurance to be expensive, but what's the cause for expensive home owner insurance? Don't they have similar hurricane frequency as Texas?

[muscleman]

http://www.sec.gov/Archives/edgar/data/1591890/000138713116005804/pih-8k_060616.htm

 

 

Does anyone understand the new reinsurance terms here? It is very confusing to me.

Is the new cap 170 or 175 million? What is the 25 million figure? It sounded like if losses exceeded 170 million, there is this additional 200 million coverage, so what is this 25 million figure for?

 

Note that this 10 million increase in premium increase will wipe out their existing annual profits.

 

 

For each catastrophic wind event occurring within a 144-hour (six day) period, we can receive reinsurance recoveries up to $170 million in excess of $5 million retention for the first and subsequent event until our limit is exhausted. For the second event of the season, $170 million in excess of $5 million is also available since we have purchased a traditional property catastrophe program that includes one full reinstatement for each layer.

 

We have also purchased a second layer of coverage that can be used for the first event for losses above $175 million per event, in an amount up to $200 million. If any portion of this $25 million in coverage is not used from the first event, the remaining portion becomes available for additional events. This $25 million program coverage applies in total for all events occurring during the reinsurance period and includes a franchise deductible of $125,000 per event. This $25 million program would also provide $3 million of potential recovery in excess of $2 million retention and $22 million in potential recovery in excess of $175 million retention for a second event occurring during the treaty year; subject to the Company’s limit not having been utilized on the first event. In addition, we have again purchased a reinstatement premium protection program which reimburses the Company for reinstatement premiums due following a first event and further limits the Company’s potential net loss.

 

The Company estimates that its total cost of reinsurance will be approximately $23.0 million for the 2016-2017 treaty-year. The total cost for the 2015-2016 treaty-year was approximately $13.2 million. Although the total cost of our reinsurance program increased by approximately $10 million when compared to last year’s program, the Company’s risk adjusted rate decreased year over year based upon AIR catastrophe modeling statistics. The estimated cost and amounts of reinsurance are based on our current analysis of Maison’s exposure to catastrophic risk and will ultimately be determined based upon the exposures we insure as of September 30, 2016. From year-to-year, both the availability of reinsurance and the costs associated with the acquisition of reinsurance will vary. Any catastrophic event or multiple catastrophes could have a material adverse effect on the Company’s results of operations, financial condition and liquidity.

[Aman1]

Hi muscleman,

 

I did have a call with management but I spent most of the time on the background of management and some more basic questions around the business model and capital allocation. We ran out of time so I wasn't able to focus more on the legal regime in LA unfortunately. However, I would note that the company's own presenation (March 2016 version) on TX vs LA doesn't show a material difference in pricing:

 

Slide 10 - shows LA as having 1.8m households and $2.6bn homeowner premiums written. This implies an average policy price of $1,444

 

Slide 11 - shows TX as having 8.8m households and $11.4bn premium = $1,295.

 

The  2014 data on LA home insurance does however show an enormous range of pricing with some insurers premiums at 4x the level of this. I've attached the insurance comparison guide produced by the state of LA. Ironically this is quite an impressive effort by the State, not something I'd see in the UK! Though I appreciate Mark will have a far better grip on the realities of state infrastructure.

 

I've been away on holiday so thanks for flagging the 8-K on reinsurance renewal.  Bigger picture, this doesn't look particularly out of line in terms of structure of the prior year (see slide 17). With one main block of reinsurance ($125m with $4m excess on first event and $1m on the second event) with a $15m top-up.

 

As the policy base grows (they are expecting Dec policies of 40-50k vs current 30k), the absolute amount of reinsurance would have to go up to match that. So this rising to $175m + $25m if anything seems reasonable. The first $5m of the $175m policy is the "excess" with the reinsurance effective for the remaining $170m up to $175m loss. At that point, if an excess went over this level, it would go into the $25m second trache protection. That's my read anyway.

 

I have gross premium growth of c.40% through 2016 which would mean c.$20-30m of additional gross premium through the policy year (i.e. through H117). The higher $10m of reinsurance is coming out of this additional policy growth.

 

Anyway, these are all ball-park figures but hopefully useful.

 

Mar_16_Presentation.pdf

2014-homeowners-rate-guide.pdf

[muscleman]

Hi muscleman,

 

I did have a call with management but I spent most of the time on the background of management and some more basic questions around the business model and capital allocation. We ran out of time so I wasn't able to focus more on the legal regime in LA unfortunately. However, I would note that the company's own presenation (March 2016 version) on TX vs LA doesn't show a material difference in pricing:

 

Slide 10 - shows LA as having 1.8m households and $2.6bn homeowner premiums written. This implies an average policy price of $1,444

 

Slide 11 - shows TX as having 8.8m households and $11.4bn premium = $1,295.

 

The  2014 data on LA home insurance does however show an enormous range of pricing with some insurers premiums at 4x the level of this. I've attached the insurance comparison guide produced by the state of LA. Ironically this is quite an impressive effort by the State, not something I'd see in the UK! Though I appreciate Mark will have a far better grip on the realities of state infrastructure.

 

I've been away on holiday so thanks for flagging the 8-K on reinsurance renewal.  Bigger picture, this doesn't look particularly out of line in terms of structure of the prior year (see slide 17). With one main block of reinsurance ($125m with $4m excess on first event and $1m on the second event) with a $15m top-up.

 

As the policy base grows (they are expecting Dec policies of 40-50k vs current 30k), the absolute amount of reinsurance would have to go up to match that. So this rising to $175m + $25m if anything seems reasonable. The first $5m of the $175m policy is the "excess" with the reinsurance effective for the remaining $170m up to $175m loss. At that point, if an excess went over this level, it would go into the $25m second trache protection. That's my read anyway.

 

I have gross premium growth of c.40% through 2016 which would mean c.$20-30m of additional gross premium through the policy year (i.e. through H117). The higher $10m of reinsurance is coming out of this additional policy growth.

 

Anyway, these are all ball-park figures but hopefully useful.

 

 

 

Thank you arman. I always look out for potential signs of trouble and play Devils advocate.

I wonder what you think when you read this in the above 8-k link.

 

The total cost for the 2015-2016 treaty-year was approximately $13.2 million. Although the total cost of our reinsurance program increased by approximately $10 million when compared to last year’s program, the Company’s risk adjusted rate decreased year over year based upon AIR catastrophe modeling statistics.

 

 

 

 

What does this mean? If they adjusted their risk model recently due to the recent frequent catastrophe, then it would imply that in all their previous years, they underreserved.

 

I do understand your rational that with more business, they need to increase their coverage,but the above language makes me feel like they are not being honest with themselves.

 

 

 

[Aman1]

Hi muscleman,

 

I don't have a strong view on the language used. I agree that coverge has increased but premiums have gone up by more, so ceteris paribus the cost per unit has not fallen, implying that the reinsurance could have a higher expected loss implicty priced in. However, there's not enough information on this.

 

Personally, it's not something I myself put in the category of "alarming", or even "troubling" for that matter. Overall reinsurance has been agreed, the coverage implies material growth in line with projections and the incremental revenue growth will comfortably cover (at least 2x) the additional reinsurance coverage.

 

Perhaps something for you to ask management on the next public call?

 

 

[muscleman]

https://www.sec.gov/Archives/edgar/data/1591890/000138713116006479/ex99-1.htm

 

Does anyone understand how the ceded losses work? In 2016 Q2, Gross losses 5.6M and ceded losses 3.7M.

I thought reinsurance only kicks in if a catastrophe happens and total loss exceeds 5 M.

 

 

Well........

I think most of the ceded losses came from claims for events in 2016 Q1, so it seems normal.

 

Losses and Loss Adjustment Expenses

The gross loss ratio for the quarter ended June 30, 2016 was 48.0% compared to 32.9% for the quarter ended June 30, 2015. The net loss ratio for the quarter ended June 30, 2016 was 24.5% compared to 48.3% for the quarter ended June 30, 2015. While claims activity was significant for the current quarter, as we continued to experience development on the wind and hail events

which occurred in February and March 2016, we benefitted from recoveries due to us under our reinsurance treaties, due to the fact that, as previously communicated, we met our overall retention of $5.0 million for multiple storms in Q1 2016.

 

[muscleman]

https://weather.com/news/weather/news/gulf-coast-deadly-flooding-latest-news-thousands-rescued-evacuated

 

Lousiana just got badly flooded. Last century there was a catastrophe every two years. Now with global warming, the frequency seems to have increased.

In my previous post, an estimate of one cat per two years gets them a long term combined ratio of 97%. Higher frequency means the combined ratio may not look good, but they can increase prices like other insurers.

[muscleman]

Does anyone know the meaning of this? What is each risk and each event? Does this mean, if a house sets on fire, they pay at most $250 and the reinsurance pays up to $1750?

It seems to me that the definition of each risk and each event can overlap.

 

 

Under the Company’s per-risk treaty, reinsurance recoveries are received for up to $1,750 in excess of a retention of $250 for each risk. The Company ceded $109 and $46 in written premiums under its per-risk treaties for the three months ended March 31, 2016 and 2015, respectively.

 

Under the Company’s excess of loss treaty, for each catastrophic event occurring within a 144-hour period, the Company receives reinsurance recoveries of up to $121,000 in excess of a retention of $4,000 per event.

[muscleman]

Anyone still tracking it?

Louisana flood is said to be worst since Hurricane sandy.

Just as we moved past the catastrophe in Q1, the flooding came in Q3, as they just amended the reinsurance from 5 million retention to 10 million retention.

The first six months broke even, so should I estimate 5 million loss for the second half of the year?

 

 

[Guest MarkS]

Hey muscleman,

 

I don't really follow this company. I also don't know much about the flood insurance world as I've always made sure that our homes were outside the 100 year floodplain.  But from what I understand the insurance is mostly covered by the federal government - and as an aside not particularly good. Typically damage from flood waters are not covered by your homeowners insurance.  If wind rips off your roof and rain destroys your furniture you can collect on your typical homeowners policy. But damage from rising water isn't usually covered.

 

Hope this helps.

Mark

[muscleman]

Hey muscleman,

 

I don't really follow this company. I also don't know much about the flood insurance world as I've always made sure that our homes were outside the 100 year floodplain.  But from what I understand the insurance is mostly covered by the federal government - and as an aside not particularly good. Typically damage from flood waters are not covered by your homeowners insurance.  If wind rips off your roof and rain destroys your furniture you can collect on your typical homeowners policy. But damage from rising water isn't usually covered.

 

Hope this helps.

Mark

 

 

Got it! So if people lose  their homes due to flood, there is nothing they can do? They just lost a big financial asset of their life?

 

I found this for PIH's 10-k. "our manufactured home policies can be endorsed to include coverage for flood and earthquake".

 

"Catastrophes can be caused by various events, including hurricanes, tropical storms, tornadoes, windstorms, earthquakes, hailstorms, flood, explosions, fires and by man-made events, such as terrorist attacks." This is vague.

 

 

But this language probably means wind/hail is the only thing they cover, though for mobile homes they provide flooding coverage.

 

"Maison provides dwelling policies for wind and hail only, and dwelling, homeowner and mobile home/manufactured home policies for multi-peril property risks."

 

[Guest MarkS]

Flood insurance certainly covers some of the loss .  However, I've heard a lot of people complain that their flood policies do not include damage to contents. But that is just what I'm hearing.  It makes a certain amount of sense as the mortgage holders seem to be the primary beneficiary of flood policies. But that's the cynic in me.  Most of the flooding occurred in some of the fastest growing parishes (counties) in the state.  IMHO the uncontrolled growth changed the way water runs off causing damage in areas that never previously flooded . This has lead to lots of homeowners with no coverage. Ordinarily we expirience a period of rapid growth after major storms. People get their insurance settlements and start rebuilding. But I'm concerned that the effects from this disaster may linger.

Mark

[muscleman]

I had a call with IR.

 

A few things to point out.

1. In proxy form, there is no compensation philosophy. IR followed up with me later saying bonus is based on overall profitability. I asked if there are specific profitability goals, and the answer is no. The bonus is paid on discretionary terms. Meaning as long as the board is happy, the board will decide to give the executives their bonus.

2. I asked why do they say although the reinsurance cost has increased from 13M to 23M, it is actually a better deal on a risk adjusted basis. Does this mean the risk is now higher? IR said, well, you know, there is 30% increase in number of policies written. So I said that means there is a 30% increase in risk. Why does this justify a nearly 100% increase in reinsurance cost? IR said the policy composition has also changed. I didn't say anything. But I think if  the policy composition has changed and that justifies the remaining 70% increase in the reinsurance costs, then they would have to also increase their per policy premium by 70% at least, in order to balance the increased risk. But that's clearly not the case here. Per policy premium is almost flat compared to last year.

 

Thoughts?

 

[muscleman]

no one following this anymore?

https://www.sec.gov/Archives/edgar/data/1591890/000138713116007965/ex99-1.htm

 

2016 Q3. Combined ratio 144%. There is always some natural catastrophy in LA. About one every two quarters. :o