Fairfax Financial Holding is primarily engaged in property casualty insurance and reinsurance. The company has an excellent track record (compounding book value at 23.5% annually since current management took over in 1985), and thus rarely trades below book value. However recently, the stock has stagnated due to short term lackluster performance (caused in large part by its defensive equity hedges) and concerns about insurance losses (principally Hurricane Sandy). Additionally, last week Fairfax was removed from the MSCI Canada index for failing a trading volume test and the stock tumbled to a 3-year low. Although it has recovered somewhat from the indiscriminate selling, it still sells below book value (price of $353.90 against a book value of roughly $360) and presents an excellent opportunity to buy into a company with a great long-term track record of creating shareholder value.
Cost of Float
Insurance companies make money by investing float, i.e., the money it receives from premiums before paying losses associated with the risks insured. So two economic factors of key importance, the cost of float and the investment returns earned on the float.
Float provides financial leverage for Fairfax without the traditional covenants or maturities of debt. Insurance underwriting adds economic value to the company if it is able to obtain the use of float at a cost (defined as the underwriting loss as a percentage of float) less than that of alternative sources of funds. In this regard, Fairfax's record since present management took over is below:
| Underwriting profit (loss) | Average Float | Benefit (cost) of float | Average Long Term Canada Treasury Bond Yield | |
| 1986 | 3.5 | 29.8 | 11.6% | 9.6% |
| 1987 | 1.0 | 54.8 | 1.8% | 10.0% |
| 1988 | 0.4 | 72.1 | 0.5% | 10.2% |
| 1989 | -13.3 | 80.8 | -16.5% | 9.9% |
| 1990 | -12.5 | 137.1 | -9.1% | 10.8% |
| 1991 | 5.3 | 180.7 | 2.9% | 9.7% |
| 1992 | -16.9 | 183.6 | -9.2% | 8.8% |
| 1993 | 2.1 | 320.4 | 0.6% | 7.8% |
| 1994 | -16.9 | 683.6 | -2.5% | 8.7% |
| 1995 | -40.9 | 913.2 | -4.5% | 8.3% |
| 1996 | -50.6 | 1,423.1 | -3.6% | 7.6% |
| 1997 | -56.2 | 2,683.5 | -2.1% | 6.5% |
| 1998 | -311.4 | 5,303.3 | -5.9% | 5.5% |
| 1999 | -617.1 | 8,545.7 | -7.2% | 5.7% |
| 2000 | -698.8 | 7,905.5 | -8.8% | 5.9% |
| 2001 | -972.1 | 6,898.8 | -14.1% | 5.8% |
| 2002 | -31.9 | 4,402.0 | -0.7% | 5.7% |
| 2003 | 95.1 | 4,443.2 | 2.1% | 5.4% |
| 2004 | 134.8 | 5,371.4 | 2.5% | 5.2% |
| 2005 | -333.9 | 6,615.7 | -5.0% | 4.4% |
| 2006 | 198.2 | 7,533.4 | 2.6% | 4.3% |
| 2007 | 238.9 | 8,617.7 | 2.8% | 4.3% |
| 2008 | -280.9 | 8,917.8 | -3.1% | 4.1% |
| 2009 | 7.3 | 9,429.3 | 0.1% | 3.9% |
| 2010 | -236.6 | 10,430.5 | -2.3% | 3.8% |
| 2011 | -754.4 | 11,315.1 | -6.7% | 3.3% |
| Weighted Avg since inception: | -2.8% | 4.7% | ||
As can be inferred from the above table, Fairfax has grown its float at a rate of 26% compounded annually, and, critically, the float has cost 1.9% less than the Canadian government's cost of capital.
As illustrated in the table below, the growth in float has slowed considerably in recent years.
| Ten year period | Compound Annual Growth in Float |
| 1986 - 1996 | 47.2% |
| 1987 - 1997 | 47.6% |
| 1988 - 1998 | 53.7% |
| 1989 - 1999 | 59.4% |
| 1990 - 2000 | 50.0% |
| 1991 - 2001 | 43.9% |
| 1992 - 2002 | 37.4% |
| 1993 - 2003 | 30.1% |
| 1994 - 2004 | 22.9% |
| 1995 - 2005 | 21.9% |
| 1996 - 2006 | 18.1% |
| 1997 - 2007 | 12.4% |
| 1998 - 2008 | 5.3% |
| 1999 - 2009 | 1.0% |
| 2000 - 2010 | 2.8% |
| 2001 - 2011 | 5.1% |
Although concerning, the slowing growth is partly due to soft conditions in the insurance markets combined with disciplined underwriting at Fairfax (as well as the anchor that large numbers naturally create). I quote Fairfax's 2011 annual report "on average we are writing at about 0.67 times net premiums written to surplus. In the hard markets of 2002-2005 we wrote, on average, at 1.5 times. We have huge unused capacity currently and our strategy during times of soft pricing is to be patient and be ready for the hard markets to come." This underwriting discipline (refusal to knowingly write unprofitable business) is key to achieving low-cost float. The growth will come as pricing improves, and not before.
Investment Record
In addition to cost of liabilities, the other important component of insurance economics is the return on assets. Below is a table reproduced from Fairfax's 2011 Annual Report illustrating the performance of the investment company's (Hamblin Watsa Investment Council) assets managed relative to their benchmarks.
| 5 Years | 10 Years | 15 Years | |
| Common stocks (with equity hedging) | 8.70% | 15.80% | 14.20% |
| S&P 500 | -0.20% | 2.90% | 5.50% |
| Taxable bonds | 13.30% | 12.50% | 10.40% |
| Merrill Lynch U.S. corporate (1-10 year) bond index | 6.00% | 5.80% | 6.30% |
There's not much to say here, their record is unequivocally outstanding. Currently due to macro economic concerns, Fairfax has hedged approximately 100% of the market value of its equity portfolio. This has resulted in substantial unrealized losses the last two years.
Management
The excellent historical performance aside, management also appears to be unusually shareholder friendly. The CEO owns a large portion of the company, a stock buy back program is in place, a small dividend is paid, and management appears to have an unwavering focus on long term results. They can think independently and have swum against the tide many times in the past, including the most recent 2008 financial crisis where they profited enormously from credit default swaps.
On the con side, they initiated an unusual lawsuit against several hedge funds for an alleged short attack on the company's stock. This strikes me as expensive and possibly unhealthy for the capital markets. I don't have all the background here, and I understand the important role that confidence plays in the reinsurance markets (which could receive signals from a falling stock price), so I reserve too harsh of judgement. But I would be interested in knowing management's reasoning here (and not the boilerplate reasoning presented in public filings), because I figure a better approach would be to just buy back large amounts of stock if other people want to sell at unreasonable prices and have the last laugh that way. In any case, this appears to be mostly behind the company now. Also on the con side, there are allegations of improper tax treatment at one of its US subsidiaries. Based on what I have read, the IRS has not disputed the treatment and it is currently not under review.
Other comments
On 11/15/2012 it was announced that Fairfax will be removed from the MSCI Canada index for failing a liquidity test. This created a substantial amount of selling (for reasons unrelated to underlying business performance) by index funds and other indexers. It's worth noting that the official re-balancing of the index occurs 11/30/2012, so it is possible that the selling pressure will remain for several weeks to come.
Conclusion
Over a 27 year period, Fairfax has consistently (a) grown float at a low cost, and (b) earned superior rates of return on invested assets. These two factors have combined to allow Fairfax to compound book value 23.5% annually for the same 27 years. Without getting into specific valuation calculations in this post (e.g., given point (a), I believe it appropriate to capitalize the economic cost advantage - using conservative assumptions - as a proxy for a conservative premium to insurance capital (see PaineWebber's "Float Based Valuation" of Berkshire Hathaway) investors should be willing to pay), it seems clear to me that Fairfax should sell at a substantial premium to book value. However, due to lackluster recent performance (partly a result its defensive hedge positions), and the recent removal from the MSCI index, Fairfax currently sells for slightly less than book value. It is my opinion that the risk of loss from an investment in Fairfax at current prices is very low and the potential upside is substantial if Fairfax remains disciplined on the underwriting and investment fronts.
Disclosure: Long FRFHF
