Undervalued Insurers?

[ericd1]

I've found a couple of companies that appear to be under-valued insurers, but I'm not very knowledgeable in this sector.

 

Arch Capital Group (ACGL)  4B mkt cap - P/E 4.7x  P/B 1.0

First Mercury Financial Corp (FMR)  185M mkt cap - P/E 5.5  P/B 0.6

Meadowbrook Insurance Group (MIG) 475M mkt cap - P/E 9.4  P/B 0.9

 

They appear to have improving operations, but perhaps I'm not looking at the right numbers.

 

I'd appreciate any comments, or insight that can help learn about the industry. I'm thinking there may be an opportunity with one of these.

[JAllen]

Hi Eric,

 

One key metric of insurance companies is the "loss reserve development".  Insurance executives have great discretion in reporting the estimated future payouts of each years losses (as evidenced by First Mercury).  So, each year the Co. reviews its business and says something like: based upon this type of business and this amount of business we wrote, we estimate that in the future we will have to payout $X.  This is a measurement of management conservatism.  This is another way of management being able to report how well they are writing insurance, their core business, and hopefully making money. 

 

I took a quick look at First Mercury's loss development info located here: http://bit.ly/dyqs27 (p. 68 of 2009 annual report).  It shows that FM has been grossly deficient in accurately estimating its losses on each year's business in most years of the past 10.  This means that in any given year we don't know how much money FM has REALLY made or lost.  It's partly up to the management to decide how much money they made writing insurance that year.  An insurance analyst would want to see "cumulative redundancies" indicating that management over-estimated eventual payouts of the insurance they wrote in each year which might indicate that the company actually made a little bit more that year than they reported.

 

I took a minute to add the total deficiencies and net reserves at 12/31/09 (from the bottom of the page, net of reinsurance) The total deficiencies of the last 10 years divided by the end of year net reserve is 15% and taking each year's alone comes out to 20%.  (Since they've reported redundancies the last two years when their premiums were larger it reduces what I think would be the weighted average reserve development-(the 15% one) but remember they reported redundancies after one and two years out in the prior years also! :) )  See how the reserves just grow and grow one year out, two years out etc.??

 

What this means is that they consistently under-report their insurance losses each year by 15-20% (probably closer to 20%) so the current year's financial results shouldn't be fully relied upon or they should at least be adjusted downwards (insurance profit down, loss ratio up).

 

Others can chime in with more nuanced views of individual insurance lines.  And, I know that there are more conservative insurance companies out there so maybe I would read everything that has been written about the ones that are so popular on the board already (FFH, MKL, FMMH come to mind) that have been pre-analyzed for you.

 

 

[twacowfca]

Good explanation, pof.  Eric, ACGL should be excellent in important respects, including favorable reserve development.  It's a great Bermuda Re company,and also a Bill Gates favorite.  :)

[goldendad]

Well, I've thought about this sector a fair bit.  First, the companies you identified are not particularly cheap compared to peers.  I've eyeballed the P&C insurers in Value Line (standard and small/mid-cap) and there are a lot of them (dozens and dozens) with 5-9 PE's with P/BV in the 0.5-1.0 range.  I think this is result of: (1) a recognition that insurers have the vast majority of their assets in bonds and as we all know the yields on bonds are very low these days, so I think the market is assigning a lot less earnings power to the float, and (2) a soft insurance market where a lot of companies have capacity to write more business so it is a competitive market (i.e. pricing power is weak for insurers right now).  Reason #2 jives with my personal experience on rates for my business insurance and personal lines.  My dad always says that when an insurer is trading below book that it means that people don't trust the book value.  Here I think it means people are discounting the value of assets because of the anemic bond returns on the float.  I also think since so many bonds are trading above par now and insurers (well, life insurers anyway) buy so many long term bonds to match-up with their liabilities, the market is mentally calculating and incorporating the risk of what will happen to the book value when someday rates rise again.  That would knock a lot of the book values down.

 

One insurer I've seen mentioned quite a bit on another board is ASI.  (No position for me.)  It trades at $16, earns $2/sh, BV=$29.  Is it too low because book is so much higher?  Maybe, but the earnings power has been consistently low . . . at book it would be at 15x which would be way to high for an insurer.  (Let alone a Bermuda reinsurer, which I have a distaste for myself, both because of the difficulty in trusting a Bermuda company and the fact that everything is great for reinsurers right up until a catastrophic event occurs).  Anyway, just my thoughts on the matter.  I don't own an insurance equity these days, just some of the exchange traded debt that I picked up during the crisis that I am still holding.

[ericd1]

Thanks for the comments..I know there's a wealth of insurance experts here! I was reviewing several "value screens" and these popped up. Thanks pof for the lesson...I knew reserves would impact the scheme of things, but didn't really know what to look for...now I do...thanks...Both Mercury and Meadowbrook showed good growth in revenues and earnings (20%+ last 5 years), but if they aren't reserving properly there's a problem.

 

 

 

 

 

[Packer16]

We had pretty good discussion about these a few month back.  I think were able to  come up with a few names that had the following good characertistics:  above average BV growth, a long track record (at least 10 yrs), around book value and positive redundancies in the reserve triangles.

 

Packer

[Uccmal]

There is a really interesting Life/PC insurer that appears to be improving its lot:

 

p/b = 0.23

p/cf = 2.2

Market Cap = 26 billion

Partly owned by a long term shareholder interested in reducing its position at a profit.

 

I dont have the time to analyze it.  Owned by a couple of renowned value investors.

[zippy1]

There is a really interesting Life/PC insurer that appears to be improving its lot:

 

p/b = 0.23

p/cf = 2.2

Market Cap = 26 billion

Partly owned by a long term shareholder interested in reducing its position at a profit.

 

I dont have the time to analyze it.  Owned by a couple of renowned value investors.

Let me guess, AIG?

[PlanMaestro]

Has anyone run AIG's total outstanding shares after the government conversion?

[Uccmal]

No, But I pulled the book value from a Reuters spreadsheet, and from an S&P spreadsheet.  Book value accounts for the debt.  So, the conversion of shares should make little difference to BV/share assuming the US doesn't profit too handsomely.

 

It is an intriguing proposition but the analysis is cumbersome to say the least. 

[twacowfca]

No, But I pulled the book value from a Reuters spreadsheet, and from an S&P spreadsheet.  Book value accounts for the debt.  So, the conversion of shares should make little difference to BV/share assuming the US doesn't profit too handsomely.

 

It is an intriguing proposition but the analysis is cumbersome to say the least.  

 

I think BV per share will be about the same $44 or so with allowance for the preferreds.  But BV is inflated. They've got some big time latent reserve deficiencies, IMO.  The Treasury gave up for nothing the options that they got in the first AIG bailout stage before they later put in all the TARP money for preferreds.  Wonder why they did that?  Doesn't make economic sense.  Current common holders also get warrants.  Again, doesn't make economic sense for taxpayers.  Chicago politics?

[Uccmal]

No, But I pulled the book value from a Reuters spreadsheet, and from an S&P spreadsheet.  Book value accounts for the debt.  So, the conversion of shares should make little difference to BV/share assuming the US doesn't profit too handsomely.

 

It is an intriguing proposition but the analysis is cumbersome to say the least. 

 

But BV is inflated. They've got some big time latent reserve deficiencies, IMO.

 

That is the problem of analyzing AIG particularly the long tail life policies.  You have too look at every subs reserving.  Now, if Berkowitz stabilizes their balance sheet and forces them to reserve properly then they may look like Geico one day.  The biggest problem is that under reserving can be hidden from the cash flow for awhile. 

 

I dont see the stock price rebounding in any hurry so there is lots of time to wait and see.  It could be Berkowitz's Sears or Salomon Bros.