My original intention for this Substack was to write up each of Berkshire Hathaway’s current equity holdings. I’ve now revised that - I will write up all their major holdings, as well as smaller ones I find interesting (like their Brazilian and Indian fintech plays). I will also do post-mortems on major positions that have been closed out, like IBM.
Berkshire’s began buying IBM in 2011, and it was no small purchase: they bought nearly 63.9 million shares for $10.9 billion. They added shares every year through 2016, at which point they owned 81.2 million shares for a total cost of $13.8 billion ($170 per share). In 2017, they had a change of heart and sold nearly the entire position (the rest was sold in 2018).
The quarterly breakdown of the 2017 sales are as follows:
Fortunately for Berkshire, IBM paid meaningful dividends along the way, so by my estimates they actually eeked out a small profit overall. However, the IRR comes out to just under 1% and the MOIC comes out to 1.05. Considering the size of the position, there were very substantial opportunity costs.
So what went wrong? In 2011, Berkshire bought 5.5% of IBM for $10.9 billion, implying a ~$200 billion market cap. IBM generated net income and free cash flow of ~$15 billion in 2010, so Berkshire bought in at around a 13x trailing earnings multiple (with >100% of earnings going into dividends and buybacks). The annual dividend increased from $2.90 in 2011 to $5.90 in 2017, and share count declined from 1.23 billion at 12/31/2010 to 922 million at 12/31/2017 (4% annual decline). So far so good - the issue is that IBM’s earnings power deteriorated during the period. This was partially due to some divestitures over the period but was primarily due to declines in core businesses.
Thus in retrospect, IBM was a value trap in 2011. It’s not difficult to see why Buffett was attracted to it though - it was a large business trading at 13x earnings with a 100% payout ratio while (at the time) growing revenue and profits at a decent clip. Buffett’s failure was simply in assessing IBM’s competitive position and growth prospects. Nobody bats 1.000. Yet, in the end Berkshire broke even. Not bad for one of Buffett’s “worst investments ever.”