Use access key #2 to skip to page content.

TWGP: Tower Group, Inc.



April 10, 2010 – Comments (18) | RELATED TICKERS: TWGP.DL

I've made it no secret that I tend to pick outperforms/underperforms on CAPS after just a quick look at the financial statements.  One can quickly determine if a company is obviously undervalued or not after just a quick look.  As the market has come up and the obvious bargains have become fewer and fewer, I've now taken it upon myself to further research my opportunities, understanding how the companies make money and how they can grow business. 

I've taken an interest in the financial sector, one that I avoided for the most part in the past.  I've specifically begun to learn a lot more about banking and more recently, insurance.  I've decided to highlight a recent CAPS pick that I added once before but closed out early just to bank accuracy.  It's hard to keep a pick open long if you don't understand its business.  Upon taking a further look, I discovered that I rather like this company.  

What company am I talking about? Tower Group, Inc. (TWGP), as you've no doubt noticed from the title.

In my screens lately, I've been looking for low PB, low PE, and high RoE.  Together, low PB and high RoE make quite a good combination.  Given the same level of earnings, companies that trade at a discount to book value will tend to have lower return on equity than a company trades at a premium to its book value.  It's hard to earn a ridiculous return on equity if there's a lot of equity to begin with.  If a company has low PB and high RoE,  that means you're getting a bargain company that earns money at a rather nice rate.  Low PE is probably a given if you have the low PB and high RoE combination working, but I like to look for it anyway.

Anyways, here goes...


TWGP offers property and casualty insurance to small and medium sized businesses.  It generally does business in segments untouched by the big boys, since these areas tend to require special underwriting expertise that requires time and skill.  The big boys seem to have a more one-size-fits-all approach, which is profitable in general but cannot be used properly to serve certain smaller segments.  TWGP seems to target these types of opportunities.

Why is all that important? Property and casualty insurance is somewhat of a commodity.  You pay money for coverage.  Pricing power is intense in this industry and margins aren't always great.  Operating in a business segment that is largely avoided by the big boys is a huge plus.  I believe a good analogy would be large mutual funds being unable to own small caps because the impact to the bottom line just isn't there.  I believe small caps contain the best values, but someone with billions to invest would be better served investing in mid and large caps, since a high rate of return on so little money amounts to a paltry rate of return.


I first became interested due to these numbers:
PB: 0.94
PE: 7.98
RoE: 15.78%

The numbers aren't amazing, but I like the PE numbers the best.  In an economy where businesses' earnings have been shelled since 2007, check out the earnings growth for the past 5 years:

2005: $1.03
2006: $1.82
2007: $1.93
2008: $2.47
2009: $2.76

Based on the earnings growth, I would say that the company does indeed know what it's doing and has grown business at a nice clip without any major hitches.  TWGP is only projected to earn $2.83 in 2010, but that's still an increase without a dip in a very long time.

Combined ratio:

Most who know about Berkshire Hathaway know that Buffett skillfully invests the float from the various insurance businesses quite masterfully.  Not every company has Buffett to deploy capital and make such strategic moves, so I'm more interested in an insurance company's combined ratio.  Combined ratio is the loss ratio (premiums - losses) plus the underwriting expense ratio (cost of attaining new policies).  A ratio under 100% indicates profitability.  

Gross combined ratio has averaged 83.32% over the past 5 years and net combined ratio has averaged 86.02% over the past 5 years, with the combined ratio never exceeding 90% in any of the years.  


There have been several acquisitions in the past year and many others in the past.  I obviously prefer organic growth, but I would say that TWGP has done well acquiring businesses based on its track record.  TWGP seems to want to maintain its edge in the more profitable niche segments while achieving economy of scale.  

Dividend Policy:

The company current pays a $0.07 quarterly dividend, which amounts to a 1.3% yield based on the close price of $22.03 on Friday 4/9/10.  This dividend started at $0.025 in 2004 and has been steady and slowly increasing.  The payout ratio is quite tiny, so I expect the dividend to continue growing in the future.

Risk Factors:

There are many risks in general in the insurance industry.  Terrorism or a major catastrophe could cause earnings to fall off a cliff and even go very negative at any given moment.  That seems less likely with TWGP, but it the risk is still there. 

Three times in the past 10 years, TWGP has participated in major dilution.  I believe this has been to acquire similar businesses, because EPS growth has not slowed despite the dilution.  Management is quite solid, but there's a chance of a major acquisition in the future that doesn't quite pan out.

For the most part, I haven't seen any major risk factors that aren't common to all other insurers as well.   


I like what I see.  EPS growth has been fantastic, even increasing the past 2 years while other businesses have seen their earnings fall off a cliff.  As TWGP grows, I expect EPS growth to slow at some point and for combined ratios to eventually climb as margins eventually fall.  This is inevitable, since the property and casualty segment is very competitive and the margins from its niche segment cannot last forever.  While it may be far away, the company will eventually will eventually grow large enough to have to sacrifice its margins a bit to grow its business.

I believe that a company still growing its earnings without hurting margins should be valued at much higher than a PE of 8.  Considering that this company is still growing, I believe this company is easily worth $35-$40 and will exceed those levels sometime in the future.  It may take some time to get there, but I believe TWGP will continue to increase earnings and dividends and that the market will eventually price this company higher.

Disclosure: Long TWGP.

18 Comments – Post Your Own

#1) On April 10, 2010 at 10:55 AM, TMFBabo (100.00) wrote:

Oops, loss ratio = losses/premiums. 

Report this comment
#2) On April 10, 2010 at 11:17 AM, XMFRosetint (48.32) wrote:

Very nice post, Bullishbabo. I wish we'd see more of these types of blogs here.

I found TWGP a while back - I don't recall when, exactly, but I was very impressed by their business. It's currently a drag of about 16 points on my CAPS score, though. I don't particularly mind, because I think the company is sufficiently cheap to far outpace the S&P in the future.

Best wishes,


Report this comment
#3) On April 10, 2010 at 12:49 PM, Option1307 (30.65) wrote:

Good thoughts!

I've been debating pulling the trigger on several insurance companies the last few months, but haven't been able to convince myself that they are a sold play going forward from these levels quite yet. I agree that they are offering somewhat of a decent value play, but many have stable/declining revenue streams and/or declining EPS the next few yrs. Personally, the insurance sector has always been something I've avoided as there just seems to be so many potential problems. I'm just not sure I can do it!

I've been looking into several different areas of the insurance sector, namely AGII, SUR, VR,and SBX. Although, I'll check out TWGP now thatyou brought it up.

Btw, check out this blog from Ultracontrarian a few months back, some decent thoughts there about the sector in general.

Report this comment
#4) On April 10, 2010 at 9:29 PM, SockMarket (34.23) wrote:

excellent blog babo (+1).

For whatever its worth, if you selected for stocks with a price/book of <1.2 & an ROE over 15% you would have underperformed the market by about 2% over the last two years. (Unfortunatly I cannot test for a longer period of time). 

Obviously though you manage to do quite well, regardless of what the model says, so I would say that your analysis does quite a bit.


Report this comment
#5) On April 11, 2010 at 12:23 AM, XMFRosetint (48.32) wrote:

Danielthebear - is that for the market as a whole? While a purely quantitative analysis is interesting, I think investors should consider the difference between the book value of an industrial company and an insurance company.

In general, the book value just by the numbers of an industrial company is a lot less valuable than one of an insurance company, because an industrial company will tend to have lots of "hard" assets - tractors, railroads, power plants, pipelines, whatever. Some companies will have massive amounts of inventory. The book value will often either significantly overstate or understate the true market value of these assets. Not so with an insurance company, their assets primarily consist of cash & securities where a market value is readily available, or if not, data for comparable securities are. As such their book values, in my opinion at least, are more reliable and more useful than a non-insurance company's would be. You'll have exceptions, obviously, where a company has invested in Level 3 assets, but this will be disclosed in their SEC filings. TWGP had 0.7% of its securities classified as Level 3 assets as of December 31, 2009, for example.

You can find all of TWGP's investments by level on page F-28 of the most recent 10-K.

Report this comment
#6) On April 12, 2010 at 2:59 AM, TMFBabo (100.00) wrote:

I probably won't blog on them, but I've flagged AWH, ENH, FSR, NYM, PTP, and VR to analyze next.  I'm basically screening by PE, PB, and ROE and looking at the 10 year numbers along with annual reports to see if the nice TTM numbers are sustainable.

@HallShadow: Thanks.  I agree, I wish we saw more of these blogs too.

I'm impressed by TWGP as well and I expect our green thumbs to be worth a good chunk of points in the future.  Other companies might seem to have more attractive ratios and such, but TWGP's track record of growing earnings without hurting their ratios is most impressive.

@Option1307: I think insurance companies might be getting a bad reputation because some of them insured against financial losses (credit default swaps!!) and got destroyed.  There are some gems out there for anyone who looks.  I liken it to safe, conservative banks being oversold just because other banks made too many risky loans.

There are good companies in every bad sector and bad companies in every good sector.  I'm finding that insurance isn't that scary to analyze and own if you learn what to look for.  

@danielthebear: What you said is why I believe value strategies don't catch on more than they should.  I believe that low PE and low PB stocks will outperform over long periods of time, but there may be droughts of several years in a row when a value strategy does not outperform the general market.  It's these droughts that turn off people who are skeptical of value strategies, I would think.

I believe low PB and high ROE will generally be a winning combination, especially with financial companies.  I do believe book value is much more important with insurance companies and banks, as HallShadow pointed out.

Report this comment
#7) On April 12, 2010 at 10:34 PM, SockMarket (34.23) wrote:


is that for the market as a whole

no. it is for the market as a whole - OTC stocks & CO's with less than a 1B market cap. but close.


I think investors should consider the difference between the book value of an industrial company and an insurance company. 

maybe so, but not for the reasons you listed. Lets take 2 balance sheets as examples. TWGP has 3.3B of total assets. of which $165M is cash and cash eq. most of it is long term investments (ie not stuff they can divest quickly, like subsidiaries in other businesses).

Then lets take PG, which is about as typical as it gets. of $135B in assets $37B is PP&E

An insurance co's balance sheet is different, slightly, but not in the ways you claim.


As for L3 assets if you can't accuratly value them (the definition of L3) you can't say how much is really in L3...




I am not disagreeing, however I would say this: the screen projected a model that outperformed in good times and underperformed in bad. By a similar amount each time. 

Obviously the last 2 years were not typical for the markets but it appears that the strategy, w/o any analysis simply magnifies the move of the market. I am a bit surprised by this myself, but I can't get significantly better answers by tinkering slightly with the model which would indicate that it has at least been true for the past couple years.

Report this comment
#8) On April 13, 2010 at 6:48 AM, TMFBabo (100.00) wrote:

@danielthebear: I have been screening for financials instead of the whole market using those parameters, but I actually still like those parameters for the rest of the market as well (compared to non-value strategies).  I guess I should've mentioned that in my earlier comment.  It sounded like I was in disagreement with you, although I didn't mean to sound that way.

When I do screen the whole market, though, I add other criteria.  I generally screen for low valuation, some sales/EPS growth, and some form of financial health. 

I haven't really done any analysis or modeling that you've mentioned; I just have a strong belief that value outperforms over the long haul.  It's the only "style" that makes sense to me and I'll probably continue to pursue it for as long as I invest in equities. 

Report this comment
#9) On April 13, 2010 at 8:30 AM, JakilaTheHun (99.92) wrote:

Great write-up, bullishbabo! 

I should probably look into this one more.  I green thumbed it on here after an initial glance, but never dug into it all that deeply.  

Report this comment
#10) On April 13, 2010 at 6:01 PM, XMFRosetint (48.32) wrote:

Danielthebear -

Perhaps I should have noted that my definition of industrial company is one with lots of hard assets. Utilities, railroads, shipping companies, etc. I threw inventory in there thinking about a potential liquidation candidate - CRC - because the inventory could be worthless. My definition of an industrial company is probably different than most, so it's my mistake that you would misinterpret that. I am, however, willing to analyze PG with you anyway.

How do you get those numbers? You're wrong, and the most recent reports of both companies can back me up on this.

As per PG's most recent report, they do have $135 billion in assets. That much is true. I have no idea how you're getting your PP&E numbers - I can only get that if I take PP&E and add all of PG's current assets to it except for cash, which doesn't really make sense. Either that, or you're ignoring depreciation. For the purposes of this post, I'm just going to use Net PP&E, which is $19 billion.

$19 b / $135 b =  14.07%

So a bit over 14% of PG's assets are PP&E, but only if you include intangible items. If we back out PG's $89 billion of intangible items, we get total tangible assets of $46 billion, which would mean 41% of PG's tangible assets are PP&E. That's a lot higher than you suggested, even using a fraction of what you calculated their PP&E as.

Now let's take a look at an example of the type of company I had in mind - Union Pacific. This is a great example of my definition of industry. It invests boatloads of cash into hard assets with long lifespans and that provide reasonable returns. If we take a look at its balance sheet, we can see that it has over $42 billion in assets. Of those assets, a whopping $37 billion of them are PP&E. That's a huge amount - it shows that 88% of its assets are fixed. This is the type of company I was referring to in my last post. To argue that a company like this is similar at all to an insurance company would be a joke, I think you can see that.

But PG's balance sheet isn't similar to an insurance company's either. If you'll recall, I calculated that 41% of their tangible assets are PP&E. Let's take a look at TWGP's balance sheet. They have just over $3.3b of assets, that much is true. Let's go ahead and back out intangibles from that number. If we do that, we get just over $3b. 

Now, where you went wrong, was counting long-term investments as investments in subsidiary companies. If you take a look at the SEC filing, the company breaks them down for you. The lion's share of them - nearly $1.8b - are in fixed maturity securities. You know, notes and bonds which of course pay them interest. After those are out of the way, you have a combined $112mm in available-for-sale equity securities, and short-term investments. Then, as you said, are the cash & equivalents, of which TWGP had over $164mm as of the most recent 10-K. Let's round them to $270 mm ($110 mm securities, $160mm cash). Add that to the $1.8b, and you get about $2.07b. 

So, ultimately, of TWGP's $3b of tangible assets, about 69% of them are in cash & securities. What about the remainder? Surely they must be in my so-called "hard" assets? Well, let's just see. According to the balance sheet, fixed assets net of depreciation account for just over $66mm. Which is equivalent to . . . 2.2% of tangible assets. The assets not accounted for under cash and investments and fixed assets are the working capital of the insurance business. You know, premiums receivable and that sort of thing.

I should probably count investment income receivable among cash & investments, but I've already finished the post and I'm not getting paid for or investing based on this analysis, so take it for what you will. I rounded most everything in this post, so the numbers might be off by a couple hundred basis points in either direction. Sources are below. 

Best wishes,

Scott - Union Pacific - Tower Group - Procter & Gamble

Report this comment
#11) On April 13, 2010 at 6:17 PM, XMFRosetint (48.32) wrote:

It occurs to me, after reading my post, that I should have said "current assets of the insurance business," instead of "working capital of the insurance business," in the next-to-last paragraph. Please take this post as a correction of sorts.

Best wishes,


Report this comment
#12) On April 13, 2010 at 6:23 PM, SockMarket (34.23) wrote:


no worries, I guess I just took it wrong. In any case some screening that may interest you:

(all screens are over the last 6 mos because that has been a normal time period relative to the past 2 years). The following are the returns of $100 invested with the given paramaters:

1) Finance only stocks, over $1B market cap, price over $5/share: $110

2) above paramaters + book value*1.3>price, ROE 5 year avg>15%: $100

3) #2's paramaters + 5 year EPS growth > 10%: $97

4) #3's paramaters + EPS must be positive for the past 3 years: $98


Over 2 years:

1) same paramaters as previous #1: $84

2) same paramaters as previous #2: $68.3

3) same paramaters as previous #3: $76

4) same paramaters as previous #4: $70.3


I have the same opinion about value stocks. I try to look for stocks that are trading below historical P/E ratios and have reasonably good growth rates. Naturally this underperformes the benchmark by some $20. 

I would tend to say that backtesting value isn't a good idea because in general the baby is thrown in with a whole bunch of bathwater and you can't really tell the screener to sort through it all. 

Report this comment
#13) On April 13, 2010 at 6:37 PM, XMFRosetint (48.32) wrote:


Thanks for the screen results. Can your screener tell us how many stocks fell into each group? That information might be useful - perhaps there's an abundance in one and a lack in another. Then again, maybe not.

Best wishes,


Report this comment
#14) On April 13, 2010 at 11:27 PM, SockMarket (34.23) wrote:


I got my data from google finance. go take a look...

Rails are a cheap example as their whole business is PP&E but I suppose if those are what you consider industrial companies then, yes that would be true. Also what you said for tangible assests is more or less accurate, not so for book value, however. 

It all depends on your definitions. What you said is technically wrong, what you meant is right.


as for the screener I am not sure, although I could check if necessary. There is enough in each group (save maybe the last) to keep a single stock from having massive effects on the results, if that is what you are driving at. 

Report this comment
#15) On April 14, 2010 at 3:50 AM, theboo1 (< 20) wrote:

Babo - Excellent write up

Report this comment
#16) On April 14, 2010 at 12:51 PM, JimBobbly (20.49) wrote:

Babo, this is my first visit to your blog - just wanted to say I enjoyed reading it, thank you.

Report this comment
#17) On April 14, 2010 at 6:33 PM, Lordrobot (87.57) wrote:

PE Ratio is a fool's paradise. There is absolutely no relationship historically between companies with High PE or low PE as having predictive value for stock performance. It is one of Ken Fishers's classic example of investor myths that get propagated with zero evidence to back them up. 

That fact does not diminish your skill as a stock picker. In fact, your skill is remarkable to find junk stocks that will go down even in a gale force upward bull trend. This interests me since your assertion here is that your reasoning is symmetrical. But it might not be symmetrical at all. Your equation for finding overvalued stocks may not necessarily apply to finding undervalued stocks. I would advise a symmetrical analysis to test your theory although in my cursory analysis of it, I don't think you can match your downside predictors. 

I have never really thought about a reverse wave predictor but I have looked at your stock picks and you may have stumble on it. You cite the major trend as having pushed companies into the land of overvalue but the major trend is decidedly not over yet. The history of recession says the bounce up matches the depth. So in this case, one would expect an 85% up side from the depths and we are only 49% up. So we have a ways to go. But these small caps you have picked have clearly run out of steam. Isn't that just taking advantage of the large cap shift which inevitably carries greater value in recovery?

Enjoyed your post.  

Report this comment
#18) On April 19, 2010 at 10:36 AM, TMFBabo (100.00) wrote:

@Jakila, @theboo1, and @Jimbobbly: Thanks.

@Lordrobot: I am a believer in what Fisher calls a myth.  I've heard of many papers that "probably" show that low PE outperforms over the long haul (I have the names of some), but I'm unable to access them because I'm not a paying member of some database site.  I refuse to pay such fees just to back up my point, so I will say to you that I have no proof but I do believe low PE outperforms.

Does Fisher say anything about low PB? I can cite Fama and French and some others on low PB outperforming.

Finding junk is easy: there are some stocks that lose money all the time and only gain equity (temporarily) by issuing more shares to unsuspecting people.  The other style is to just find overvalued companies: companies that make very little money or no money at all who trade at valuations that just aren't justified.  You can also usually see lots of dilution to raise cash, dwindling sales and margins, and more.  I also try to see to it that they have very little book value to begin with, so an asset sale isn't really an option either.

There are similarites in my green thumbing and red thumbing approaches, but there are differences as you've suspected. 

I agree that many small caps have run out of steam.  Many that I bought in early 2009 that were incredibly undervalued are now fairly valued or even overvalued already.  However, I believe that there are still stocks to buy in a market that isn't ridiculously overvalued (and no, the market is not ridiculously overvalued). 

I will continue to buy companies that I think have a chance of going up 50 to 100 percent.  The majority of the companies I like are still micro caps actually, but larger small caps such as TWGP (~1B market cap) are finding ways into my portfolio.

Reverse wave predictor? Sounds like technical analysis.  The only aspects of technical analysis in which I believe are simple volume information and a tiny bit on momentum (moving averages).  I'm not a believer of technical analysis, I'm afraid.  I'm much more interested in valuing companies and I don't believe a chart pattern says anything about what a company's worth. 

Report this comment

Featured Broker Partners