Every investment decision is unique and every performance outcome is different. As bottom-up stockpickers, we believe our investment approach is best illustrated through the discussion of historical examples. The two stocks discussed here – SanDisk and AIG – tell very different stories. Both provide a useful introduction to our thought process. Click on the key milestones in each chart below to learn more about our experience with each stock.
Two months prior to our investment, SanDisk rejected a hostile takeover bid from Samsung Electronics at $26 per share. With its shares trading at $15, SanDisk looked foolish for not jumping at the opportunity. The global financial crisis would soon begin to worsen and Samsung looked highly unlikely to submit another bid. Although SanDisk's balance sheet was solid – we estimated the company had net cash of about $2.50 per share after certain adjustments – it wasn't bulletproof. The company had about $1.8 billion of off-balance sheet operating leases – equivalent to more than 140% of shareholders' equity – and had fallen out of compliance with certain covenants attached to these after an S&P downgrade in July 2008. The big risk, particularly amid rapidly worsening credit markets, was that further downgrades or non-compliance might trigger an acceleration clause, allowing SanDisk's lenders to call the full amount of debt at a time when raising fresh capital was nearly impossible.
SanDisk's share price had fallen almost 90% from its August 2007 highs to less than $6 – a level that had not been seen since the aftermath of the technology sector's collapse in 2002. Flash memory pricing was suffering from a severe capacity glut and rapidly slowing consumer demand. Of the 21 sell-side analysts covering SanDisk, only two rated the company a "Buy". We saw things differently. Although the short-term picture was undoubtedly bleak, we believed that all of the bad news appeared to be priced into SanDisk shares. At the time of our purchase, SanDisk shares were trading at about a third of tangible net asset value, compared to a typical historical multiple of two to three times. We began purchasing shares of SanDisk for the Orbis Funds in November 2008.
Shares of SanDisk continued to fall after our initial purchase, but we kept building the position. Using a conservative sum-of-the-parts calculation, we believed that SanDisk was worth more than twice the prevailing share price at the time. Based on the history of past cycles in the flash memory market, we thought it was reasonable to expect normal supply and demand conditions to return over a three-to-five year horizon. We could envision SanDisk shares having intrinsic value of at least $15, but perhaps as high as the mid-$20s depending on the path of the recovery in the industry and the economy. As for the potential downside, even after assuming massive losses and further cash drain, we believed that SanDisk would still have enough cash to control its destiny until at least March 2011 without needing to raise fresh capital. Obviously it was impossible to tell how long the global financial crisis would last – or how bad it would ultimately get – but we thought the risks to SanDisk's capital structure were overstated.
With an average sale price of over $22 a share – about 2.5 times our average purchase price – SanDisk was one of our top performers in 2009. We started to unwind the position in April 2009 when the price was about $13 per share. With the benefit of perfect hindsight, this was too early. We would have produced a much greater relative return in both 2009 and 2010 by maintaining a sizeable position as SanDisk continued its steady march toward $50. But the reality is always more complicated. Although we were confident that SanDisk could weather the storm and return to a more normal valuation in the future, there were plenty of risks. Consumers were under enormous pressure and it was clear that demand would take a long time to recover. If conditions got worse in the economy and credit markets, SanDisk could easily have started burning cash at a faster rate than we anticipated. The prudent course was to maintain our discipline and shift our focus to other ideas where the risk-reward proposition was more favourable.
We sold our last shares of SanDisk by the end of April 2010 when the stock was trading above $40. Our experience with SanDisk shows the power of contrarian thinking at its best. At the outset, there were a lot of good reasons to be sceptical and it was hard to envision any catalysts for improvement in the short term. Indeed, there was a chance – albeit remote in our view – that SanDisk might not survive the financial crisis. Looking beyond the pessimism, however, we were able to see a long-term opportunity. We believed that the flash memory market's horrific supply-demand dynamics – and SanDisk's profitability – would eventually normalise as it had in the past. In fact, we believed that SanDisk stood to benefit from the crisis as weaker players in the industry would be forced to consolidate or go out of business. And by spending just as much time thinking about the downside risks and questioning our assumptions, we were able to maintain our conviction and add to the position in the face of massive short-term volatility.
AIG was one of our initial holdings when the Orbis strategies were launched in 1990 and we owned the stock for much of the 1990s. We took a fresh look at AIG again in March 2005, just after the company had been forced to restate its results and long-time chief executive Hank Greenberg had been ousted by the board. AIG shares had fallen more than 25% since the accounting issues surfaced and had lost their premium valuation. AIG shares were trading in line with the rest of the US property and casualty insurance sector – then about 13 times earnings and 2.2 times tangible net asset value – despite significantly better long-term fundamentals. AIG had several dominant franchises, most notably in US property and casualty (P&C) insurance and life insurance in Asia, that could drive long-term earnings and book value growth at double-digit rates, significantly higher than the stockmarket average. Our estimate of AIG's intrinsic value implied about 25% to 50% upside from prevailing share prices.
Initially, our investment thesis proved to be correct. As of May 2007 – a little more than two years after our initial investment – shares of AIG had made a modest positive contribution to Orbis Global's relative performance. We locked in some of these gains with tactical share sales, but AIG remained a 2% position in the portfolio. In July 2007, US subprime mortgage debt went into free fall. This was unwelcome news for many financial firms, but AIG management was notably proactive in providing details about their exposure to the subprime market. Although we were surprised at the scale of housing market exposure, we were comfortable that it was well-contained at AIG, with only 4% of its investment portfolio in securities with subprime mortgages as collateral. Still, we proceeded to undertake a considerable analysis of AIG's mortgage exposure in effort to better understand the risks. We took comfort from AIG's stated $15-20 billion of excess capital which, along with the company's strong level of retained earnings, formed a buffer against our worst-case assumption for likely investment losses.
In May 2008, AIG surprisingly announced plans to raise capital. This was a shock because AIG planned to raise $20 billion, even though we believed it still had a modest amount of excess capital despite much of its previous cushion having been eroded by mark-to-market losses that we considered temporary. We remained willing to add to our position in part because AIG was now trading at just 7.5 times normalised earnings and in line with our calculation of "clean" tangible net asset value after making conservative assumptions. These were valuation levels not seen since the late 1980s, yet we believed that most of AIG's long-term earnings power remained intact.
In August 2008, AIG announced more losses. It became clear that more capital would eventually be needed. Still, we took some comfort in the $10-12 billion of normalised earnings from AIG's core insurance businesses. Worst case, we expected management to sell businesses to raise cash in lieu of raising equity. With AIG's market cap now at $50-60 billion, it was difficult for us to envision intrinsic value being that low even if we made unusually bearish assumptions about future losses. Raising more capital at lower prices would mean even more dilution and an even sharper fall in intrinsic value. While we trimmed our estimate of intrinsic value, we still believed the long-term risk/reward trade-off favoured owning shares in AIG. We bought more to maintain its 2% weighting in the Orbis Global Strategy.
Ultimately it was a shortage of cash, not capital, that brought down AIG. We were aware of this prospect, but thought that resources at the parent company were sufficient. Liquidity is sometimes difficult to analyse, but until very close to the end we thought AIG could sell businesses to placate the rating agencies. Instead, the plunging stock price caused the rating agencies to downgrade AIG, prompting collateral calls and a fatal loss of confidence. That left AIG short of cash and unable to raise equity because the share price had fallen so much. By then, AIG needed to raise far more than its market cap and it had no time to sell any of its businesses. The endgame was a US government takeover that took 80% of the company and immediately knocked our assessment of intrinsic value from $35 to $7 per share. With upside that was capped at $6 to $8 per share at this level and downside to zero if the government didn’t run the company in the best interests of equity holders, we sold our last shares of AIG in October 2008. As the firm's worst performance detractor over the past 10 years, our second experience owning AIG was clearly a tremendous disappointment. Put simply, we forgot to heed Keynes' famous warning that "the market can stay irrational a lot longer than you can stay solvent". It is in many ways a cautionary tale about the risks that are inherent in our contrarian investment approach. Despite careful analysis and a disciplined investment process, we failed to avoid a permanent loss of client capital. Fortunately, over the long run, our investment process has uncovered enough winning stocks to more than compensate for the inevitable detractors.
The upper part of the chart shows how the size of the Fund’s holding has varied in relation to movements in the share price, while in the lower section, the dark red line shows the percentage of the fund’s assets invested in the company over the holding period.
The green line is the most difficult data series to grasp. It represents the share price performance relative to the fund’s benchmark. At any given point in time, the green line shows the relative performance from that date to the chart's ending date.
Click on A,B,C etc. to learn more about our decision making during critical points in the investment thesis.
To open an investment account please log in to your online account. If you have not registered an online account you can do so here. Registering on the website will better protect your investment account details.
Your username will usually be your primary email address.
If you are still unsure about your username, please contact the Orbis Client Services Team on:
Please enter your username and answer the security question you chose during sign up and then we will email you further instructions for resetting your password.
If you have any problems please do not hesitate to get in touch with the Orbis Client Services Team:
We have sent you an email that contains a password reset link. After clicking this link we will require you to enter your username which will allow you to create a new password.
Please be aware that for security purposes this link will expire in 20 minutes. You may now close this box.
Unfortunately your browser requires an updated version of Adobe Flash Player to view our interactive charts. You can download it at: http://get.adobe.com/flashplayer/
Thank you for visiting the Orbis website. Please note that some functionality on tablets and smartphones may be limited as the site was developed for desktop use.