You might think that index investing is mega-cheap, but smart beta guru, Rob Arnott, believes that index investing is more expensive than many realise.

(For this article, index investing means investing in well known indices such as the FTSE 100 where the companies in an index are weighted by market cap.)

Arnott, the founder of Research Affliliates, outlined his case at the Inside ETFs Europe conference in London. The biggest hidden cost arises when new entrants come into an index. Let's look at the US flagship index, the S&P 500

Whenever the index is rejigged, S&P announces who the new entrants will be before the changes are implemented. This means that the well-known index investors such as Vanguard or iShares have a 'grace period' of a few days to invest in a new entrant before it officially joins the index. This is important because when a new entrant is announced, the share price normally rises immediately.

Here's a hypothetical example: S&P announces on October 12 that XYZ Inc will be joining the S&P on October 17. Before the announcement, XYZ's share price was $10. Immediately after the announcement, investors (including big index investors) start buying shares. So by October 17th, the share price of XYZ has risen to $10.15, which means the stock enters the S&P at $10.15.

If XYZ officially joined the index on October 12, when the price was $10, the index funds would perform less well than their underlying index. That's because their purchases would have driven the price above $10, and most of the purchases wouldn't have been at $10. So the funds would be 'beaten' by the index and would look bad.

But because the official admission of the stock is delayed for a few days - at $10.15 in this example - it's easier for the passive funds to mimic the performance of the index.

As a result, ordinary investors face a hidden cost. They're hit by a 15 cent cost and they don't know about it. (We should stress, this is just a hypothetical example. The size of the rise will vary and there will be times when a company's share price falls during the grace period, perhaps due to wider market moves. But the basic point is sound.)

What's more, the stocks that come in are often trading on a pretty rich valuation. Stocks that all out may be cheaper. According to Arnott, new 'addition stocks' outperform the market by 36% in the year before they enter the S&P 500. With index investing, there's a 'flavour' of buying shares in the expensive companies and selling shares in the cheap ones.

Lazy portfolios

Arnott that one way round this issue would be run 'lazy portfolios' where the fund doesn't invest in the new entrants for 3 or 12 months. That's because, more often than not, the share prices of the new entrants fall back a little once the index investments have bought their shares and the initial excitement has worn off. Arnott's figures suggest a significant performance boost if you wait for a year to invest in the newcomer's shares.

Top dogs

Arnott also highlighted how the biggest stocks in an index may well not stay that way. So concentrating your portfolio in the biggest stocks in an index may be more risky than you realise.

Arnott produced lists of the top ten largest stocks worldwide since 1990.

Bank of Tokyo Mitsibushi
Industrial Bank of Japan
Sumitomo Mitsui Banking
Fuji Bank
Dai-Ichi Kangkyo Bank
UFJ Bank
The above 1990 list is dominated by Japanese companies just as the Japanese stock market boom was coming off its astonishing peak in 1989.

We see massive changes ten years later as the biggest stars of the tech boom get into the global top ten.

General Electric
NTT Docomo
Exxon Mobil
Lucent Technologies
Deutsche Telekom
By 2008 the top ten is more resource-heavy.
Exxon Mobil
General Electric
China Mobile
Royal Dutch Shell
And in 2018, we see several big tech plays and just one remaining resource play in Exxon Mobil.
Exxon Mobil
Berkshire Hathaway
Johnson & Johnson
JP Morgan Chase
History suggests we'll see a lot of changes by 2028 in the global top ten so don't assume that holdings in Apple or Alphabet will definitely serve the index funds well over the next decade.

These tables also provide support to the argument that as index funds invest most in the largest companies, they often end up putting more money into the more expensive stocks. And as these top dogs change, index funds are always having to buy and sell shares in the portfolio which incurs more costs. Arnott says that the average annual turnover for the S&P is 4.4%.

What to do

So in the light of all this, what should investors do?

If you want to avoid the hidden costs of indexing, sadly there's no ETF following the 'lazy portfolio' strategy that Arnott talked about. One option is to go for one of the smart beta strategies on the basis that you're not going to see such big share price movements when an individual stocks comes in or out of an index. That's because markets aren't as focused on any of the smart beta indices as they are for the big flagships like the FTSE 100 or S&P 500.

Or you could continue to invest in traditional index funds, and just be aware that there are these hidden costs. Given that some passive ETFs are now charging as little as 0.04% a year, we can perhaps cope with costs being incurred elsewhere.