Simple is best, but not always.


Like with ETF construction

The largest holding in the Russell 1000 & 1000 Growth indices is Apple

The largest in the Russell 1000 Value index is Berkshire Hathaway

Whose largest portfolio holding is…


$108bn of their $447bn equity value

How does that make sense?

“Russell style indexes are built using three highly representative growth & value characteristics. Our style indexes use one value characteristic, book-to-price ratio (B/P) & two growth characteristics, medium-term forecast earnings growth rate based on I/B/E/S two-year forecasts & sales-per-share growth rate based on five-year historical sales.”

Berkshire is at 1.5x book value, in the lower 50% of the 1000 shares so you’d expect it to be included

on this simple basis

Except that

Apple itself is trading at 31x price to book

And other holdings

Coke is at 12x book

Moody’s is 32x

And a Bloomberg analysis shows that 45% of Berkshire’s portfolio is technology

So “book” is the MARKET VALUE of these holdings, not the underlying book value of these holdings

Ie. This methodology doesn’t work with investment holding companies

But how is Disney the 5th largest holding in the Russell Value at 48x forward earnings & a 0.4% FCF yield?

Well, it’s 3.7x book value which makes the cut

Except in 2019, Disney bought 21st Century Fox for $70bn

$50bn of which is goodwill and included in “book”

ie. it’s only book, not tangible book..

Yet excludes HPQ, presumably because they have bought back so many shares that their “book” (equity) is now negative

despite being on only 9x earnings

As long as you don’t think these Value ETFs represent Value