Most stocks make money, probably

I was distracted this morning by the first sentence in Chapter one of 'What's Behind the Numbers', a new book recommended by an investor on Twitter. The line is:

Most stocks lose money

But then the book confused me. 

It invites us to consider research [pdf] from 1983 to 2007, "the longest secular bull market most people today have ever experienced". The Russell 3,000, a US index, rose nearly 900%, yet 39% of 8,000 stocks in had a negative total return and 64% did less well than the index.

Contrary to the authors' assertion, the statistics seem to show most companies, 61%, made money, although many of them did not beat the index. The book then quotes another statistic: the worst performing 6,000 stocks in the study, 75% of them, had a collective total return of 0%.

But that can't be the origin of the "most stocks lose money" claim either. If the collective return was 0%, some of the bottom 75% would have lost money and some would have made it. They won't all have been losers.

Assuming the period of the study is representative of all periods, the second statistic might tell us "a minority of stocks are responsible for the majority of the market’s gains" as the study says but not as the book also says "a few stocks are responsible for all the market's gains".

Maybe it was just a mistake, and the authors meant to say, "Most stocks lose money relative to the index," but since 'What's Behind the Numbers' is about spotting chicanery in financial statements and the authors warn us of self-serving executive spiel, this odd chapter is making me wonder whether I should persist with the book.

It seems to be an example of what the authors warn against. How words can obscure the truth in numbers.