Andrew Dickson is the founder of Albert Bridge Capital, a fund based in London. In this post, he counters a recent FT opinion piece by Lawrence Burns of Baillie Gifford on the value of never-selling shares in superstar firms.

On Monday, the FT published an opinion piece written by Lawrence Burns, an investment manager at Scottish fund house Baillie Gifford.

The piece was titled “Why it is usually a mistake for investors to take profits” with its stand first suggesting “a tiny number of superstar companies account for returns from equity markets.”

Baillie Gifford, among other things, is well known for its prescient investment in Tesla. The author also referred to an investment Goldman Sachs made in Alibaba, and how it invested both too little and sold too early. Using these examples, the author then makes the claim that “the upside of not selling is nearly unlimited.”

In an after-the-fact diagnosis, Burns is implying that a biased sample of self-selected winners like Tesla and Alibaba suggest that it is a mistake to ever sell any shares in any companies that you think are “winners” either historically or prospectively. Yes, it is true, if you sold Tesla too early, you didn’t make as much money as those who didn’t. But what about those investments that weren’t Tesla? And what about former consensus winners like Eastman Kodak, or Cisco, or Nokia?

Justifying his view, Burns cites research by professor Hendrik Bessembinder which suggests that a) a majority of stocks do not outperform 1 month US treasury bills b) 1 per cent of companies accounted for all global net wealth creation and c) 99% of companies were a “distraction to the task of managing money”.

Burns suggests that Bessembinder’s research should “shake the foundations” of the investment industry and that we need a “vastly different mentality” to “focus on the possibility of extreme upside”.

That sure would be nice, wouldn’t it? If we all could just sit back and take our money out of 99 per cent of our investments and all plough it in the same 1 per cent that everyone else already knows about, and then expect to outperform.

As preposterous as it sounds, the conclusions drawn from Bessembinder’s paper were also misinformed in the first place. It isn’t entirely Burns’ fault, Bessembinder himself presented his observations and conclusions in a sensational way.

His original paper was based on US stock market data going back to 1926. In it, he claimed that “the best four per cent of listed companies explain the net gain for the entire US stock market since 1926”. But dig deep and you’ll find that these were not the top four per cent of a broad index like the S&P 500, but instead the 25,300 companies that have appeared in the Center for Research in Security Prices over the past 90 years.

So that was 1,092 companies, or over twice as many “big winners” as there are in the S&P 500.

Bessembinder then states “simply put, large positive returns to a few stocks offset the modest or negative returns to more typical stocks.” In other words, he is effectively asserting, as if it is an epiphany, that the portfolio names with the highest excess returns drive the excess returns of the portfolio. Well, duh.

But back to the 1,092 companies that companies that “accounted for all the net wealth creation”. Well, that was a bit misleading as well. It turns out that another 9,579 stocks also outperformed US treasury bills.

Similarly, Bessembinder claimed that “less than half the names outperformed US Treasuries.” Well, less than half the names underperformed US Treasuries too. About five per cent of the universe was in line, and the balance was split between the relative winners and the relative losers.

Bessembinder chose to exclude this finding from the abstract and conclusion, although he admitted this point in a footnote; yet still concluded that “poorly diversified portfolios will underperform because they omit the relatively few stocks that generate large positive returns.”

Very frankly, the data just doesn’t support his argument, and conclusions drawn from it are misinformed.

Subsequent extensions of this paper (here and here) utilise a global dataset, but the issues persist; both with the presentation and conclusions. But Baillie Gifford is still running with it, because Bessembinder tells the story that confirms their recent success, presumably abetted by their still-large stake in Tesla.

Good on Baillie Gifford for generating such excellent returns. Yet I sincerely believe Burns’ advice is the worst possible going forward.

For the rest of us, I suggest following more tried and tested advice from a different BG:

My guess going forward is that this advice is going to help people make or save a lot more money than the never-sell-anything mantra proposed by the fund managers at Baillie Gifford.