Tower Group TWGP

The Fund I manage has a long position in TWGP. The position was initiated on October 1, 2013. This is not a recommendation to buy or sell any security. Do your own due diligence.

Tower Group (TWGP) has been in the news quite a bit this year. The company operates in two difficult markets: workers compensation and commercial auto liability. When reserved correctly (like anything else in insurance) these businesses can be very lucrative. When reserved incorrectly or too aggressively, these business lines can cause a world of hurt, something that TWGP has been experiencing.

Things have been unraveling for a while now and the past 18 months have been marked with "reserve strengthening." (This is just another term for "we estimated our liabilities incorrectly and now you the shareholder has to pay.") Things really hit the fan after the company hired an independent actuarial firm on June 30, 2013.

So TWGP announced that reserves would be boosted by $60-$110 million and there would be a large goodwill write-off. TWGP then orchestrated an awful PR campaign.

Order of Events

1. The press release from August 8, 2013 said that financial statements would be released within 30 days.

2. TWGP announced that they were terminating the agreement to acquire American Safety Reinsurance from Fairfax- just a few days after they received a stock notice that they were in danger of being delisted (due to a failure to post financials and a 10Q).

3. On September 17th the company issued another press release stating that they would release financials on October 7th.

4. On September 23rd Michael Lee, TWGP's CEO and Chairman, had a margin call and was forced to sell 1.233 million shares. Overall I saw a lot of commentary on various forums that basically said "Lee is selling shares because he has no confidence in the business - therefore the business is going under." Taking the time to read the actual filing painted a much different picture.

5. Gary Maier, Tower's Chief Underwriting Officer, announced his resignation on September 24, 2013. 

This all sounds bad and the stock got pummeled each and every press release. But let's look at a potential scenario to examine what could happen.

Bad Scenario

It should be obvious that TWGP estimates their reserves very aggressively. This was known beforehand and I found it surprising that a company with legacy line issues and aggressive accounting traded at a premium to book(AFSI is probably headed for a similar issue soon). With that said the company brought in outside actuaries prior to this debacle and established reserve ranges, which can be seen on page 13 of their 10K.

Tower carried reserves at $1,398.9 million and had high/low reserve estimates of $1,770 million/$1,284 million respectively. The midpoint for the reserve estimates is $1,527 million, about $129 million higher than what is currently carried. If Tower had to carry reserves equal to their high estimate, reserves would have increased $371.1 million.

As of 3/31/2013 tangible book value was $843 million. This works out to TBV/share of $14.68(57.4 million shares outstanding). If the company reserves at the high end of estimates book value can be expected to drop to $472 million ($843M - $371M) or $8.22/share. That looks good compared to a share price in the mid 7's. The problem is a dramatic decrease in book value would by extension boost operating leverage. This would probably attract the attention of AM Best and cause a downgrade, the company is already under surveillance with a negative outlook (August 12, 2013) so this wouldn't be welcome news.

I don't know what AM Best would require but at the most simplistic level they would want to see Net Written Premiums to Policyholders' Surplus(NPW/PHS) drop below 3X. Anything above 3X is considered "Unusual." In 2012 the company wrote Net Premiums of $1,721 million. Compared to my low estimate of TBV of $472 million this works out to NPW/PHS 3.6X, definitely "unusual."

These elevated levels leave TWGP with a couple of options: hope that regulators ignore them, raises equity, or borrow a balance sheet and cede premiums to better capitalized insurer(s). The first option seems unlikely. Raising equity is a definite possibility and borrowing a balance sheet was announced on October 1st. There are three reinsurers who participated.

Arch Insurance Agreement

This is a quota share agreement where TWGP cedes  premium to Arch. Based on previous data, Arch is expected to gain $227 million in premiums. In exchange, Arch will pay TWGP a commission between 25%-31.5% ($56.75-$71.5 million) depending on loss ratios recognized by December 31, 2015. At the lowest end, TWGP will cede ~$227 million in premiums and receive $56 million. With operating expenses equal to 35% of gross written premiums the company will take a hit to their combined ratio.

The company has not had a stellar combined ratio in the past several years, so most of the return on book was from investment gains. Near term, this will be ~$10-$25 million hit to returns (35%*$227 million minus ceding commissions of at least $56 million) in 2014.

Hannover Agreement

This is a quota share agreement that is similar to the Arch agreement but it covers the Workers Comp business. This covers 14% of the business and at no point can ultimate ceded NPW exceed $550 million(page 6). They are ceding $284 million for business already written and expect to cede $251.3 million for new/renewal business. Since renewals have already occurred it seems reasonable that 2014 will comprise the $251.3 million tranche.

The ceding commission rate is 25%-30.6% depending on how loss ratios develop by 2014. So old business was written at a 5 point loss(at least, operating expenses were 35% in 2012). So commissions in 2014 could be $62.8-$76.9 million. Based on 35% operating expenses this creates a loss of $25.1-$11.0 million.

Southport Agreement

This agreement has three sub-agreements. The first has TWGP ceding 30% of their workers comp and employers liability business written after July 1, 2013. TWGP will get 19-28% commissions depending on loss ratios. Gross premiums for workers comp was $419 million in 2012, so TWGP will cede $125 million of premium and lose $20-$8.75 million on the ceded business.

The second agreement takes care of $459 million of workers comp and employers liability policies written from 2011-2013. The premium paid to Southport is equal to the unpaid portion of liabilities. I think this will bring down investment income (float is essentially taken off the book and given to Southport to invest and hopefully) but it will have very little impact on BV. I have assumed that this will bring ROE down half a point or roughly $6 million. 

The final agreement is an excess of loss contract for workers comp business written from 2011-2013. The premium paid is the same as the second agreement. Again, I'll assume (perhaps incorrectly) that the impact from this agreement is $6 million.

Altogether the Southport agreement could impact TWGP $32-$20.75 million in 2014.

Ceding Agreements Impact to ROE

Management has guided for 12% ROEs in 2014. Based on a book value of  $1.218 billion this means that management expected to generate roughly $145 million of earnings. This will obviously be significantly lower due to the ceding agreements.

Using the mid-range to my estimates, the Arch agreement will bring earnings down by $17.5 million, the Hannover agreement will bring earnings down by $18 million and the Southport Agreement will bring earnings down by $26 million. Grand total this brings earnings down by $61.5 million. I will round up to $65 million to add a cushion.

The company will start reserving at higher levels as well. I estimate (based on the high/low reserve estimates in the past 10K's) that higher reserves will chew up another $50 million of earnings. This is hardly scientific though and calls to IR have not been returned.

2014 Earnings could be $145-17.5-18-26-50 = $33.5 million

Now granted, this doesn't account for any growth, hardening of lines(workers comp should see 9% rate increases), or lowering of expenses(something that management has guided for). But it sets up the base.


NWP/PHS Impact

The reinsurance contracts will take ~$600 million of premiums away from TWGP. That means, all things equal, TWGP will write $600 million less in net premiums in 2014 compared to 2012, thus NWP should be $1,121 million.


In my worst case scenario TBV drops to $472 million, which would give a NWP/PHS ratio of 2.3X. This would probably be an acceptable ratio for regulators and given that the legacy workers comp book has been passed off to Southport there is a lot less to worry about.

With lower liabilities from the legacy line, and a better NWP/PHS ratio I believe that shareholder dilution is unlikely.

Conclusion

Results are expected on Monday October 7th. With the company trading in the mid 7's I believe that a vast majority of the downside is accounted for and it is likely to prove excessive. There is a tremendous amount of pessimism with TWGP and a lot of it is warranted. The company is being taken to the cleaners but auditors and actuarial firms are making sure that reserves are more than adequate. Right now the downside is an AM Best downgrade, something that I believe is unlikely, but certainly not out of the picture.

My bet here is that a lot of bad news is baked in and a conservative approach would yield a company that is roughly fairly valued. Maybe something better comes up, maybe not but pessimism is a contrarian investors best friend. Monday will hopefully paint a clearer picture. Long TWGP.