《金融时报》蓝色巨人投资者手中未必是好牌

In 2011, the legendary investor Warren Buffett caused a stir by announcing Berkshire Hathaway’s first major investment in a technology company, an area of the market he had always avoided, claiming that he didn’t understand it.

This approach accorded well with his advice that investors should always stick to investing within their so-called “circle of competence”. Most sensible investors would agree with this idea; after all, who would embrace a strategy of investing in companies or assets you don’t understand? In my experience, however, many investors still fail to realise how narrow their circle of competence is.

Buffett’s purchase was over $10bn of stock in technology company IBM, making it Berkshire’s second-largest portfolio investment after Coca-Colaand turning Berkshire into IBM’s largest shareholder.

Since his purchase, IBM has reported falling revenues. The fourth quarter of 2013 was the seventh in a row in which revenues were lower than they had been in the equivalent quarter the year before. The chief executive and other senior executives have volunteered to forego their bonuses for 2013 as a result of this poor performance. As the Lex column of the FT said recently, “IBM quarterly results practically write themselves now: dreadful revenue growth accompanied by cost controls, share buybacks and dividends that combine to make the technology legend look somewhat less bad than the top line would suggest”.

It so happens that I was looking at IBM at about the same time as Warren Buffett was accumulating his stake, but I decided to avoid it. Why?

A couple of features about IBM unsettled me. One was that they had announced a “road map” to generate growth in earnings per share (EPS) to $20 by 2015, up from $11.50 in 2010. I do not like management that uses terms such as “road map” unless they are discussing driving cars. “Plan” is a perfectly good word.

I also dislike the focus on EPS. Not all earnings are created equal. Some require greater or lesser amounts of capital to generate them and not all are delivered in cash. This should come as no surprise to Mr Buffett, who identified return on capital as the primary test of company performance in his 1979 annual chairman’s letter. Much EPS growth is generated at the expense of return on capital and so destroys value.

But if the focus on EPS growth was worrying, the planned means of achieving this identified in the “road map” was even more troubling: roughly 40 per cent revenue growth, including acquisitions, 30 per cent operating leverage and 30 per cent share buybacks.

Revenue growth is a higher quality source of value creation than share buybacks or cost control (which is what I take it IBM means by “operating leverage”), provided it is not achieved at the expense of returns. I am wary of revenue growth achieved through acquisitions, the majority of which do not create value. However, there is at least no limit to revenue growth, while cost-cutting and buybacks are both finite.

Moreover, although some buybacks create value for shareholders, many do not and are executed seemingly irrespective of the valuation of the shares. A company cannot create value for remaining shareholders if it pays more for the shares it buys than their intrinsic value, but many companies (and investors) are fooled by the familiar boast that buybacks are accretive to EPS. In an age where the alternative use of cash often generates little income because interest rates are close to zero, almost any alternative is accretive to EPS, but it does not necessarily create value.

I found a five-year plan for significant share buybacks particularly disturbing. How can the management possibly know whether the shares will trade sufficiently below intrinsic value to create value enhancing buyback opportunities on such a scale and so far ahead?

One other feature of IBM struck me in 2010 when my Fundsmith colleagues and I were reading the company’s 2009 annual report. We noticed that it had a $1.9bn error in its cash flow statement. We rang IBM to check that we were not misinterpreting it, and they confirmed we were correct and that we were the only people who had asked about it.

Maybe others had discovered it and didn’t bother to call, but I suspect the reality is that very few investors or analysts read annual reports and 10k filings any more. The error was corrected in the 2010 annual report and did not materially affect my view of IBM, but it certainly affected my view of those who were analysing it.

Terry Smith

Financial Times