Notes on Fairfax Financial Holdings' Shareholder Letters (1998-2017)

These are my own scratch notes on what I found interesting and noteworthy as I read through Prem Watsa's Shareholder Letters (linked here).

Notes on 1998 Letter

I particularly enjoyed this one's commentary on the market environment back in 1998. It is interesting to note that he was very attuned to the market's valuation risk and adjusted Fairfax's portfolio heavily in favor of bonds and cash and away from equities (7% of total portfolio vs. up to 35% in earlier years).

Here was my favorite section:

"We continue to be very concerned about the level of the U.S. stock market as discussed in our 1996 and 1997 Annual Reports – even though the S&P 500 increased again by 26.7% in 1998. The possibility of deflation, mindless long term investing in mutual funds and a lack of investment ‘‘values’’ in the North American stock markets do not appear to bother most investors. In fact, speculation is rampant in the U.S. markets as demonstrated by the ‘‘internet’’ stocks. America Online has a market cap that exceeds the total market cap of the five largest Canadian banks even though AOL has been a public company for only six years. Amazon.com has a market cap that is in excess of Sears (U.S.) even though Sears annually earns more than twice Amazon.com’s revenues. Finally, Yahoo! has a market cap in excess of Boeing even though the latter has revenues of $55 billion compared to Yahoo!’s $190 million. The speculative juices are flowing freely in the U.S. but the music will stop and many investors (speculators!) will not have any chairs to sit on. Caveat emptor!"

How times have changed for Sears and Amazon! But nonetheless, are we surprised about the internet bubble bursting when valuations were at such insane levels?

Notes on 2000 Letter:

In this letter, Prem Watsa highlights his portfolio companies' shortcomings:

Return on equity of 4.1% (measly by Fairfax's historical standards).

Fairfax stock sold for less than book value for over 15 months but allowed Fairfax to repurchase shares at prices immediately accretive to shareholders.

Large insurance losses mostly stemming from newer US operations in Ranger, Crum & Forster, and TIG ($588M in 2000 alone).

Total underwriting combined ratio of 116%, a far cry from cost-less float (100% or better).

Even with unfavorable results, Fairfax has compounded book value by 37% from 1986-2000 and recorded an 18.2% ROE over that time period, not to mention a stock that compounded at 33% (close to book value) over the time. It is refreshing when management (Prem Watsa in this case) is highly transparent about their business, especially when things go sideways. Who wouldn't want a partner that is honest 100% of the time?

The last tidbit I noted was their low exposure to stocks (~6%) in their insurance float portfolio due to the high valuations back in 2000 as well as their discussion of some of the internet company stock blowups of the dot com bubble (AOL -54%, Amazon -80%, Yahoo! -86%, VerticalNet -92%, Red Hat -94%, and the list continues).

Notes on 2001 Letter:

Sustained large losses related to the Enron collapse as well as the 9/11 terrorist attacks, as well as the historically low interest rate environment, all combined to produce a negative return on equity for the first time since 1985.

It seems to be my observation that the property and casualty insurance industry can earn good returns on equity capital in general when prevailing interest rates are moderate to higher (think 80's and 90's) because investment portfolio returns can outweigh soft pricing. However, when low interest rate environments prevail, it is extremely important to underwrite in a conservative manner because it is often tougher to come by high investment returns that make up for bad underwriting. It can be a double edged sword - low investment returns coupled with bad underwriting results can erode shareholder capital (and sometimes at alarming rates if the balance sheet is too highly levered).

Continued to hold large notional amounts of puts on S&P 500 and a basket of tech stocks to hedge against equity market declines, and only had 6% of investment float allocated to common stocks.

Notes on 2002 Letter:

Combined ratio of 100%.

Still only a 7% allocation of investment float to common equities. Total bond and common stock realized gains of over $700M.

Listed Fairfax shares on NYSE - where 2 million of their shares were promptly sold short!

Notes on 2003 Letter:

Over $800M of total realized investment gains.

97.6% combined ratio.

Net debt to equity of 53% from 55% in 2002.

13% allocation of investment portfolio in common stocks, but 50% in either cash or short term maturity investments.

"We have been concerned for some time about the risks in asset-backed bonds, particularly bonds that are backed by home equity loans, automobile loans or credit card debt (we own no asset-backed bonds). It seems to us that securitization (or the creation of these asset-backed bonds) eliminates the incentive for the originator of the loan to be credit sensitive. Take the case of an automobile dealer. Prior to securitization, the dealer would be very concerned about who was given credit to buy an automobile. With securitization, the dealer (almost) does not care as these loans can be laid off through securitization. Thus, the loss experienced on these loans after securitization will no longer be comparable to that experienced prior to securitization (called a ‘‘moral’’ hazard). And here’s the rub! These asset-backed bonds are rated based on their historical loss experience record which will likely be very different in the future – particularly if we experience difficult economic times. Also, in the main, these asset backed bonds are a creation of the 1990s, a period when the U.S. experienced one of the longest economic expansions in its history, followed by one of the shortest recessions."

Notes on 2004 Letter

Large hurricanes, coupled with a conservative investment portfolio (largely in short term investments and cash) essentially eliminated any profits for 2004.

Net debt to equity of 48%, deleveraged from 53% in 2003.

Notes on 2005 Letter

Tough year for insurers with Hurricane Katrina, Rita, and Wilma.

Watsa notes interestingly that over the 7 year period, Fairfax essentially broke even when all premiums, losses, and investment gains/losses were netted out - a disappointing period for both management and shareholders.

Continued to deleverage at the parent company level by issuing equity (approximately $300M) and paying down debt, but shareholder's equity took a beating, resulting in net debt to equity ratio rising to 53% from 49% in 2004.

Watsa largely uses the GDP to Market cap ratio as a general valuation metric like Warren Buffett. When it is far above 1, he tends to use hedging techniques such as buying put options on certain market indices to protect his equity holdings against large market declines.

Notes on 2006 Letter

Continued to deleverage, net debt to equity ratio of 38.9%.

It is interesting to note that even in 2004 and 2005, Watsa was attuned to the housing market bubble that would pop in 2008.

"Finally, we continue to worry about the unprecedented issuance of collateralized bonds, mortgages and loans (we hold none!). The assumption in the marketplace is that ‘‘structure’’ will eliminate or significantly reduce all risks. So a portfolio of 100% non-investment grade bonds, sub-prime mortgages or non-investment grade corporate loans, by sophisticated structuring, can transform into securities of which 80% or more are rated A or above. This has resulted in thousands of collateralized bond issues being rated AAA while fewer than 10 corporations in the U.S. are AAA! We see an explosion coming but unfortunately cannot predict when. As Grant’s Interest Rate Observer said in its December 15, 2006 issue, ‘‘Blame for the distress at the fringes of subprime, we judge, cannot be laid at the feet of the U.S. economy. It should, rather, attach to the lenders and borrowers who piled debt on debt until the edifice sways even in a dead calm.'"

What a call!

Notes on 2007 Letter

The credit default swaps that Watsa bought as he was watching the housing and collateralized debt bubble build finally took off in 2007, appreciating from $198M to $1.3B (almost 6 times Fairfax's money in merely 2-3 years).

Net debt to equity ratio of 20.7%.

Notes on 2008 Letter

Continued to sell credit default swaps to the tune of $1.5B in gains shorting the credit bubble. Total gains of over $3B by shorting the credit bubble through credit default swaps.

Net debt to equity ratio of just 6.5%. This represents significant opportunity to lever up and expand in harder insurance markets, whereas, if your balance sheet is very debt laden, your opportunities are limited beyond paying down debt.

Fairfax loaded up on US equities, holding $2.9B by year end (total of $4.4B in common stocks).

Notes on 2009 Letter

This quote says it all regarding the perils of the insurance industry:

"Speaking of the long term and why there is no place for complacency in business, AIG’s history is quite instructive. It took AIG 89 years to accumulate almost $100 billion in shareholders’ capital and one year (2008) to lose it all. Frightening!"

Levering up at net debt to equity ratio of 13% - up from 6.5% in 2008. This seems to be a great way to take advantage of mispriced opportunities - deleveraging in heady, frothy years, and levering up once you have the capacity and opportunities in the down years like 2008-2009.

Continued to load up on stock to the tune of $5.1B at year end in 2009.

Notes on 2010 Letter

Just as I stated above, with low net debt to equity ratio in times where opportunities abound, a company has the capacity to take advantage of these wonderful opportunities, and Fairfax did just that by buying 5 companies in 2010.

Continued to lever up - net debt to equity of 23.8%.

Great discussion of the commodities bubble that deflated after 2010 - mostly driven by Chinese demand and huge amounts of construction going on in China (construction was 40% of GDP!).

Fairfax got burned somewhat on its equity hedges in 2010 when it increased its hedge to 100% of equity exposure and markets continued to rebound from 2008-2009 lows. This resulted in muted investment returns for 2010.

Watsa was concerned about deflationary effects in the economy from deleveraging and the fiscal burdens on the economy, and bought a significant notional amount of CPI-linked derivatives that would increase in value in deflationary environments. For a small cost, they provide big protection in the event of deflation, which can decimate investment based companies.

Notes on 2011 Letter

Major catastrophes cost company 19.4 points on its combined ratio (over $1B in losses) including: Japanese earthquakes and tsunamis, Thailand floods, U.S. tornadoes, New Zealand earthquake, Hurricane Irene, Australian floods. Tough year!

This year marks an increased interest in investing directly in operating businesses - Fairfax invested in SandRidge Oil, William Ashley (retail), and Sporting Life (retail), Kennedy Wilson (real estate services) and Bank of Ireland (which, interestingly, was owned partly by the Irish government after it nationalized a lot of the banking assets due to the financial recession).

Continuing to lever up at 24.5% net debt to equity ratio.

Interestingly enough, Watsa called the Chinese real estate bubble in 2010, and it began in 2011. He also warned about commodity prices going parabolic and that he wouldn't be surprised if they reverted to longer term average prices if the Chinese economy slowed - this began in 2011 as well.

Total stock portfolio only amounted to $3.8B in 2011 compared to $5.1B in 2009.

Notes on 2012 Letter

Purchased 77% of the largest Indian foreign exchange and travel company, Thomas Cook UK.

Purchased an 82% interest in Prime Restaurants, a franchisee of various restaurants, at about 10 times free cash flow.

Extremely interesting piece of writing by Watsa in the 2012 letter on their large investment in Blackberry (Research in Motion):

"Markets fluctuate – and very often in extreme directions. Remember the tech boom, when companies with no sales were valued at tens of billions of dollars? In 2000, Northern Telecom accounted for 36.5% of the Toronto Stock Exchange index and was worth almost Cdn$400 billion; by 2009, it was bankrupt! Well, last year the opposite happened to Research in Motion (now known as BlackBerry). At its low of approximately $61⁄2 per share, it sold at 1⁄3 of book value per share and a little above cash per share (it has no debt). The stock price had declined 95% from its high! The company produces the BlackBerry which for years was synonymous with the smart phone. The BlackBerry brand name is perhaps one of the more recognizable brand names in the world and the company has 79 million subscribers worldwide. Revenues went from essentially zero to $20 billion in about 15 years – and then it hit an air pocket! The company got complacent, perhaps overconfident, and did not respond quickly enough to Apple and Android. Mike Lazaridis, the founder and a technological genius – and a good friend – asked me to join the Board, which I did after meeting Thorsten Heins, whom Mike recommended as the next CEO of the firm. Thorsten’s 27 years of experience in all types of leadership jobs in small and large divisions at Siemens, combined with his five years at BlackBerry, were exactly what was needed. Thorsten hired a very capable management team and then focused on producing a high quality BB10 – the next generation of BlackBerries. The brand name, a security system second to none, a distribution network across 650 telecom carriers worldwide, a 79 million subscriber base, enterprise customers accounting for 90% of the Fortune 500, almost exclusive usage by governments in Canada, the U.S. and the U.K., a huge original patent portfolio, an outstanding new operating system developed by QNX and $2.9 billion in cash with no debt, are all formidable strengths as BlackBerry makes its comeback! The stock price recently moved as high as $18 per share, a far cry from the $140 per share it sold at a few years ago. And please note, 1.8 billion cell phones are sold worldwide annually, and of the 6 billion cell phones in the world, only 1 billion are smart phones. Lots of opportunity for Canada’s greatest technology company! What is striking, even for a person like me who has seen many bull and bear markets, is that at $61⁄2 per share, all the Wall Street and Bay Street analysts were uniformly negative – just as they were uniformly positive only a few years ago at prices north of $100 per share. John Templeton’s advice to us: “Buy at the point of maximum pessimism”, still rings in our ears!! We own approximately 10% of the company at an average cost of $17 per share and we are excited about its prospects under Thorsten’s leadership and Mike’s technical genius."

Selling for just above cash per share? What is there to not like? Superior downside protection, with zero debt, and a call option on the underlying business operations. Interestingly, this position is still Watsa's biggest holding in 2018 as of 6/5/18.

Net debt to equity ratio of 21.6%.

Common stock holdings of $4.5B total.

Notes on 2013 Letter

Unrealized hedging losses caused a net loss in 2013 on a mark-to-market basis.

Acquired Hartville (pet insurance) and American Safety (environmental liability).

Invested an additional $100M in CARA, another restaurant business that was merged with Prim Restaurants. Also purchased a 55% interest in Kitchen Stuff Plus, a specialty kitchen and household supply and giftware retailer.

Fantastic stat here - in the 27 year period Watsa managed funds for Fairfax, they have only had 3 negative return years! And their 2013 loss was eliminated in 2014 by unrealized gains. Incredible.

Purchased convertible debt in Blackberry to the tune of $1.25B (6% 7 year convertible debentures convertible at $10/share).

Net debt to equity of 20.7%.

Continue to hedge their equity holdings, despite recent unrealized losses that have caused investment returns to drop significantly. Watsa continued to be wary of leverage and excess, especially in China:

"In the last few years we have discussed the huge real estate bubble in China. In case you continue to be a skeptic, here are a few observations from Anne Stevenson Yang, an American who has been in China for over 20 years and is the founder of JCapital Research in Beijing:

1. China added 5.9 billion square metres of commercial buildings between 2008 and 2012 – the equivalent of more than 50 Manhattans – in just five years!

2. In 2012, China completed about 2 billion square metres of residential floor space – approximately 20 million units. For perspective, the U.S. at its peak built 2 million homes in a year.

3. At the end of 2013, China had about 6.6 billion square metres of new residential space under construction, around 60 million units.

4. Yinchuan, a city of 1.2 million people including the suburbs, has 30 million square metres of available apartments – roughly 300,000 units that could house 900,000 people. This is in addition to the delivered but unoccupied units. The city of Guiyang, capital of Guizhou Province, has roughly 5.5 million extra units for a city of 5 million.

5. In almost every city Anne has visited, pretty much the whole existing housing stock has been replicated and is empty.

6. Home ownership rates in China are estimated to be over 100% versus 65% in the U.S. Many cities report ownership over 200%. Tangshan, near Beijing, is one.

7. This real estate boom could only be financed through unrestrained credit growth. Since 2009, the Chinese banks have grown by the equivalent of the entire U.S. banking system or 15% of world GDP.

8. The real estate bubble has resulted in companies extensively borrowing and investing in real estate or lending on real estate in the shadow banking system. This is exactly what happened in Japan in the late 1980s.

9. And one observation of our own: Since 2009, the easing by the Federal Reserve combined with the explosive growth in China, backed by higher interest rates, has resulted in huge inflows (‘‘hot money’’) into China. The near unanimous view that the renminbi would strengthen has resulted in a massive carry trade where speculators have borrowed at low rates across the world and invested in China, almost always backed by real estate. The shadow banking system in China – i.e., assets not on the books of the major Chinese banks – is estimated by Bank of America Merrill Lynch to be approximately $4.7 trillion or 51% of Chinese GDP. Oddly enough, prior to the credit crisis, the U.S. had $4.5 trillion in asset-backed securities outstanding or approximately 31% of U.S. GDP. You know what happened then. When the flows reverse in China, watch out!"

Notes on 2014 Letter

Continued to expand insurance operations under Fairfax Asia as well as under Hartville with several bolt-on acquisitions. Also acquired QBE insurance in Hungary, Czech Republic, and Slovakia. Purchased Brit PLC, a market-leading global specialty insurer and reinsurer.

Formed a new entity, Fairfax India, to invest specifically in Indian businesses. Raised $1.1B in an equity offering, mostly subscribed by current institutional Fairfax shareholders. Fairfax Financial provided $300M of the capital, giving them a 28% stake.

Fairfax made several investments in Greek companies due to a more stable government outlook and a pro-business government. Fairfax is clearly looking to expand more globally in its operating (non-insurance) businesses since about 2010.

Net debt to equity ratio of 20.2%.

$4.9B in equities at year end in the investment portfolio, or roughly 30%.

Notes on 2015 Letter

Acquired Eurolife, a life and property and casualty insurance company. Upped their stake in ICICI Lombard. Purchased a 7.2% stake in Africa Re. Purchased a 10 year 7% convertible bond in FBD, a farm insurance company in Ireland. One can truly see how much of a global operation Fairfax is when reading this letter: Investments in Canada, US, Greece, Ireland, Africa, India, Brazil, Ukraine, the Middle East, Vietnam, Thailand, and their insurance operations touching over 100 countries.

Net debt to equity ratio of 17.3%. Low leverage for quite a few years, perhaps waiting for new opportunities again, as in 2005-6-7 era. Not unlike the large cash buildup on Berkshire Hathaway's balance sheet over the last 5 years.

$4.4B in equities across their investment portfolio.

Notes on 2016 Letter

Acquired Allied World for $4.9B, the largest acquisition to date in the company's history. Acquired Zurich Insurance Group's South African Operations. Bought AIG's insurance operations in Latin America and Europe.

Completely removed equity and deflation hedges when President Trump was elected, which is interesting. This effectively was a directional bet on the US and North American economies when Trump was elected. However, upon removing these hedges, they realized significant losses which effectively eliminated their profits for the year. Fairfax also reduced exposure to longer maturity bonds and reduced their average bond duration to about a year and a half.

Invested in Chorus Aviation by buying debt and warrants. Bought preferred securities and debentures in Mosaic Capital. Purchased 5% preferred shares and warrants in Altius Materials.

Net debt to equity of 29.1% - levering up on their purchase of Eurolife, Allied World, and ICICI Lombard.

Notes on 2017 Letter

Tough year for catastrophes with Hurricanes Irma, Maria, and Harvey along with California wildfires and Mexico earthquakes. Cost the company 13.1% of net premiums earned.

Reduced stake in ICICI Lombard to below 10% and used proceeds to start Fairfax's own insurance operation in India, Digit.

Fairfax listed another division, Fairfax Africa, on the TSX. This company consists of various investments solely based in Africa.

Acquired a significant amount of debt and warrants in Seaspan (shipping), Chorus Aviation, AGT Foods, Mosaic Capital, Altius Minerals, and Westaim - a total of $786M invested across this group of companies.

Net debt to equity ratio of 22.0%.

Interesting commentary on Bitcoin and the Cryptocurrency craze:

"Can I finish this section on investments without making a comment on Bitcoin and cryptocurrencies? I am always amazed at how rank speculation comes into markets once every decade or so! Reminds me of the outstanding book ‘‘Extraordinary Popular Delusions and the Madness of Crowds’’. Bitcoin, a digital currency operating independently of central banks that uses blockchain technology to regulate the creation of new coins, went from $500 in 2015 to $19,000 in 2017 – that’s a parabolic curve if you have ever seen one! And it currently trades at under $10,000. It has no assets, no revenue and no profit. Madness of crowds all right! Will any of us be surprised to see this go back to $500 – down 97% from the top!! Having said that, I must say I felt the same about Amazon in 1999. It did fall 92% from the top but has had huge success in the last ten years. Perhaps Bitcoin and the blockchain technology will experience the same result – you have to survive the first 92% drop of course!!"

Closing Stats:

Average combined ratio of 97.2% for insurance companies they have owned since 2008 (at least 10 years), which represented $53.2B in cumulative premiums written over that period.

Weighted average cost of float was -.5%, or in other words, they were paid to invest the float! Outstanding record.

Investment float grew from $13M in 1985 to $22.7B in 2017 (~1,746 times increase)!

Book value compounded at 19.5% annually over the past 32 years and the stock price followed at 18.1% (not including dividends). That record is Buffett-esque.