Description
Fairfax is one of those bi-polar stocks. There is a formidable list of blue-chip value investors who are long the stock. At the same time there is very high short interest. The longs think it’s dramatically undervalued while shorts like Jim Chanos of Kynikos Associates think “its worth zero” (as quoted in a recent issue of Value Investors’ Insight). One side has to be right and I’m squarely in the long corner on this one.
Some of the well-known Long Value Investors who own Fairfax include:
Longleaf Partners/Southeastern Asset Mgt: 23.6%
Peter Cundill & Associates: 8.7%
Brandes Investment Partners: 5.2%
Markel Gayner Asset Management: 5.2%
Torray Corp.: 2.6%
Reed, Conner & Birdwell: 1.8%
5 of the 6 value investors named above are tracked by Portfolio Reports – a publication by the same folks who put out Outstanding Investor Digest. Some 85.8% of shares are held by institutions and the daily volume on the stock is very low. On most days, under 20,000 shares change hands.
Here is some of the short-interest data:
Aug.15, 2005 2,090,540
Jul. 15, 2005 2,081,941
Jun. 15, 2005 2,206,550
May 13, 2005 2,464,208
Apr. 15, 2005 2,757,007
Mar. 15, 2005 2,531,785
Feb. 15, 2005 2,900,298
Jan. 14, 2005 2,869,452
Dec. 15, 2004 2,691,041
Nov. 15, 2004 2,906,103
Oct. 15, 2004 2,685,063
Sep. 15, 2004 2,445,677
Aug. 13, 2004 1,874,478
For August, 2005, the average daily volume was 19,562 and the days to cover was over 100. Clearly, this is a significant short position by any measure. The savvy longs believe that Fairfax’ intrinsic value is significantly higher than its present $164.79 price. If they are right, the large short position is only going to help get them to intrinsic value (and beyond) sooner.
Business Background:
Based in Toronto, Fairfax has been under present management since 1985. The stock ended the year 1985 at C$6.00 and recently traded around C$198 – a 19.1% annualized rate of return excluding dividends. Fairfax’ CEO and senior management are very good value investors. Even the shorts would not disagree with that statement. Like Berkshire Hathaway, they went into the insurance business as a means of getting access to cheap capital to invest. The 19.1% annualized return over the last twenty years was made possible due to doing very smart value investing with their insurance float – which has grown considerably over the years.
Fairfax has had a history of buying P&C insurers on the cheap. Usually these were poorly managed insurers that Fairfax specialized in turning around. Like Berkshire, Fairfax fixates on underwriting discipline and profit – even at the expense of dramatically shrinking premium volumes. And usually they’ve bought these underperforming insurers at bargain basement prices. The secret sauce Fairfax adds to these acquisitions is exceptional underwriting disciple and exceptional investing prowess released on the newly acquired float. This mantra has worked exceedingly well for 20 years. Infact through 1998, when the stock closed the year at C$540, the annualized return for 13 years (since present management took over in 1985) was an astounding 41.4%.
The Crum & Foster and TIG Acquisitions Saga:
In the late 1990s, the company made two large acquisitions that, in hindsight, caused significant anguish and indigestion. In 1998, Fairfax acquired Crum & Foster Insurance for US$565 Million. At the time of acquisition, C&F had a book value of US$765 Million. The US$200 Million discount to book gave them a decent margin of safety against adverse claim development. In addition, they capped their downside by buying a US$400 Million stop-loss policy against adverse reserve development and uncollectible reinsurance recoverables. If you read page 76 and 77 of the 1998 annual report (http://www.fairfax.ca/Assets/Downloads/AR1998.pdf) you’ll note that the company highlighted the future risks and benefits related to C&F quite candidly. A large part of the C&F acquisition was paid for by issuing a million shares of FFH stock at C$475/share.
In 1999, Fairfax acquired TIG Holdings for US$847 Million – a US$280 Million discount to book value. In addition, to get further downside protection they bought a US$1 Billion excess loss cover from Swiss Re. It covered adverse events at TIG and other Fairfax subsidiaries. The TIG acquisition was paid mostly by issuing 2 Million shares of FFH stock at C$500/share in 1999 (net proceeds were C$959 Million).
The Swiss Re US$1 Billion Cover put Swiss Re on the hook for the first US$1 Billion of adverse reserve development and unrecoverable reinsurance losses. In exchange, Fairfax was on the hook to pay a fraction of that loss as it was incurred and ceded.
Both these acquisitions worked out poorly for Fairfax. Here are the gory details:
1. Losses ceded to Swiss Re and the cost to Fairfax.
1999: US$251 Million (US$35.3 Million paid by FFH to Swiss Re)
2000: US$272.3 Million (paid US$167.2 Million paid by FFH to Swiss Re)
2001: US$203.8 Million (paid US$143.6 Million paid by FFH to Swiss Re)
2002: US$5.4 Million (paid US$4.2 Million paid by FFH to Swiss Re)
2003: US$0
2004: US$0
2005Q1Q2: US$0
Total Premium Paid to Swiss Re: US$350.3 Million. About $267.5 Million of the cover remains.
2. Underwriting Gains/Losses at Fairfax:
1998: (C$311.4 Million/US$249.12)
1999: (C$617.1 Million/US$493.7)
2000: (C$698.8 Million/US$559 Million)
2001: (C$972.1 Million/US$778.7 Million)
2002: (C$5.7 Million/US$4.6 Million)
2003: US$87.7 Million
2004: US$108.4 Million
2005Q1Q2: US$99.2 Million
3. Runoff Gains/Losses at Fairfax:
1998: C$0
1999: (C$54.2 Million/ US$43.36 Million)
2000: (C$43.3 Million/US$34.6 Million)
2001: (C$27.4 Million/US$21.9 Million)
2002: (C$117.7 Million/US$94.2 Million)
2003: (US$110.0 Million)
2004: (US$193.6 Million)
2005: (US$151.9 Million)
From 1998 through Q22005, aggregate insurance underwriting losses (including payments to Swiss Re and Runoff) are about US$2.8 Billion. This includes 9/11 and all the hurricanes through 2004 as well. This US$2.8 Billion loss was offset by the following gains and cash inflows:
2000-2004 Net Fairfax Stock Issues: US$341 Million
From 1998 through 2004, Fairfax’ had very substantial realized gains, interest and dividend income on its investment portfolio – more than offsetting the big underwriting losses. They also IPO’ed 2 subsidiaries and those cash receipts helped as well.
Net after-tax income (including all underwriting losses and investment realized gains) is as follows (in Millions):
1998: US$310.0
1999: US$99.4
2000: US$109.9
2001: (US$276.8)
2002: $263.0
2003: US$271.1
2004: (US$17.8)
2005Q1Q2: US$40.2
TOTAL: $799 Million
Net net, Fairfax earned about US$800 Million in the last 7-8 years including the fallout from 9/11, all the hurricanes and 2 big acquisitions that didn’t work out as expected. They were unable to fix TIG as a going concern and put it in runoff managed by their stellar runoff group, TRG. Crum & Foster is fixed with a 106.5% combined ratio for 2004. In addition, they raised, net of stock buybacks, $341 Million from issuing FFH stock from 2000 through 2004.
Fairfax has about 16.1 Million shares outstanding and a book value of US$184.46 or US$3 Billion as of 6/30/05. Book value/share exceeds the price/share. This only makes sense if the reported book value is false. And the two major ways it could be false are:
1. Reserves are understated and/or
2. The $7.3 Billion recoverable from reinsurers is not necessarily recoverable.
If the combination of 1 & 2 exceeds US$3 Billion then Chanos would be right in his “worth zero” commentary. The facts, however, do not support that thesis:
1. At a recent Wesco annual meeting, Charlie Munger made a comment that you cannot gauge the intrinsic value of any P&C insurer or any bank purely from an examination of the financials. He went on to say that you have to also understand the nature and ethos of management. He was alluding to the wide discretion insurers and banks have in setting aside reserves. If they are set too low, they overstate earnings and book value.
Prem Watsa and his team have been running Fairfax for about 20 years now. I have carefully followed the company for over five years now and have looked at their annual reports going back to 1990. My conclusion is that this is an extremely smart and ethical individual. Watsa is a graduate of the famed Indian Institute of Technology in Madras in India. IIT, Madras gave Mr. Watsa a distinguished alumnus award in 1999 (http://www.iitmaa.org/alumni-awards/Premwatsa.html). The company has always communicated clearly and candidly to its shareholders – especially any bad news. It has always underpromised and overdelivered. And it has always laid out the bad news and their errors candidly.
2. If the reserves and recoverables are way off base, then it must be because Fairfax senior management are crooks and/or incompetent. Even the shorts would quickly dismiss the incompetence question. This is a stellar management team that made a mistake on two acquisitions. As soon as they found out how bad it was, they laid it out for shareholders and took the hit on their balance sheet. They also stopped writing new policies on one of the acquisitions and put it into runoff. Being wrong on reserves can easily happen with long tail P&C Insurers – even ones who are ultra-smart and conservative. Nothing sinister about that.
3. PricewaterhouseCoopers has certified the quality of the reserves and recoverables every year after full due diligence and uninhibited access to documents and employees. In addition, the insurance rating agencies have been gradually upgrading Fairfax’ ratings – and these agencies have had, like PwC, extensive access going well beyond the public filings.
4. Their annual reports going back to 1985 are listed on the website. I’d suggest that the reader go back and read every letter to shareholders since 1985. Before investing in Fairfax, one must understand the nature of management. If one carefully studies the communication and actions of this management team, one will walk away in awe of them. They took a tiny nondescript insurer with about C$17 Million in revenues in 1985 to a business that generated US$5.8 Billion in revenue last year. And they did it honestly with the highest standards of integrity. And the road was bumpy – and they reported the bumps as they happened. In many cases, Fairfax planned well for future bumps and tempered their effect – like the stop loss insurance and Swiss Re cover.
4. Markel and Fairfax were joined at the hip in the early years before going their separate ways. For a number of years, after they separated operations, Steve Markel and Prem Watsa sat on each other’s boards. And in 2004, Markel participated in a private placement of Fairfax stock along with Longleaf. Markel are not only exceptionally good value investors, they know P&C insurance space cold. In addition they’ve known this particular business and this particular management team for 20+ years. The folks at Markel understand Fairfax and its management a lot better than someone like Chanos or myself – and they are long on Fairfax. Markel’s other P&C Insurance investments are Berkshire Hathaway and White Mountains. Their actions suggest that they believe Fairfax will deliver stellar results on their investment.
5. Recoverables from reinsurers was over US$10.2 Billion on 12/31/01 and book value then was over US$170/share. Today reinsurance recoverables are down by nearly $3 Billion while book value is up to about $185/share. These recoverables will continue to decline over time as claims are incurred and paid.
6. The major rating agencies also have extensive access and have made favorable remarks and alluded to ratings upgrades as Fairfax has already turned the corner and beginning to fire on all cylinders.
Intrinsic Value – Over US$330/share
Fairfax has a book value of about US$185/share. Net of the reinsurance receivables, Fairfax has investable float of about US$7.2 Billion as of 12/31/04 or US$447/share. Thus investable assets are over US$600/share (float plus book value). Much of this is sitting under-invested in treasuries and high-grade bonds. Over time, as markets give the savvy Fairfax investment team an opportunity to invest some of these funds in compelling undervalued equities, book value would jump substantially. Even at an average 5% return, it would add $30/year in book value. And at a still reasonable 10% return, it’s a staggering $60 in additional book value annually. Anyway you look at it, Fairfax’ intrinsic value is over US$330/share – and possibly could go much higher over the next few years.
An alternate method to arrive at intrinsic value is to take the market value of its publicly traded subs, add net holding company assets and then assume a 1.5x book value multiple on the wholly owned subs. This is one of the methods used by Scotia Capital (a Canadian investment bank) to arrive at intrinsic value.
Northbridge (59.2%) – 29.5 Million shares @ C$32.50 US$767 Million
Odyssey Re (80.8%) – 52.15 Million Shares @US$24.28 US$1266 Million
TOTAL US$2033
16.1 Million shares
TOTAL Public Subsidiaries (A) US$126.27/share
Holdco:
Cash & Marketable Securities: US$499.5 Million
Other Assets US$112.3 Million
LT Debt (US$1780.6 Million)
Preferred Shares (US$136.7 Million)
Total Holdco (B) (US$1305.5 Million) ($81.10/share)
NAV (A+B) US$45.17/share
Stub: BVPS (12/31/04)
US & Asian Insurance Companies
Crum & Foster US$60.39
Fairmont US$10.48
Falcon US$2.02
First Capital US$2.50
Total US Insurance US$75.30
Runoff
US Runoff US$46.08
European Runoff US$33.74
Total Runoff US$79.82
TOTAL Stub US$155.12
1.5x BV of the wholly owned stub is US$232.68 plus $45.17 is US$278/share
This approach does not capture the potential investment gains from the large float and underwriting profits nor the fact that Odyssey Re and Northbridge themselves are undervalued and will do well over time as well. It is fair to say that intrinsic value is comfortably north of US$278/share – a 68% premium to the present stock price.
As with all VIC postings, I feel under no obligation to update this or any other writeup or to provide any data on present holdings or any past or future trades. Please always do thorough/independent due diligence on this and all other investment ideas posted on VIC.
Catalyst
Value is its own catalyst. However in the case of Fairfax there are a few others:
1. Probable future significant investment gains
2. Continued underwriting profits (Katrina may extend the ins. hard market)
3. Rating Agency Upgrades
4. Any combination of 1-3 above would cause a significant bump up in stock price – leading to a very significant short squeeze and even higher prices. I don’t see most of the savvy value longs unloading any meaningful positions under $250-300/share – leading to severe pain for the shorts. This could be the mother of all short squeezes.
Over time, Fairfax will trade up to its intrinsic value – which is north of US$278/share.