ETFs take this basic principle to its logical conclusion and simply invest in markets as a whole; they don't even attempt to identify which bits of the market are likely to do best because their holdings exactly mirror the entirety of the market they are tracking. Thus a Hang Seng Index (HSI) ETF simply mirrors all the stocks in the index in proportion to their weight in the index.

Naturally, all ETFs are not equal, and their differences are marked by which parts of the market they choose to mirror. ETFs often outperform so-called actively managed funds run by highly paid stock pickers who invest in the shares they like, and sell the ones they think will underperform.

It all sounds too good to be true, and a slew of studies suggests that it is. Kenneth French, a finance professor at Dartmouth College in the United States, produced one of the studies that sum this up more vividly than most. It is called 'The Cost of Active Investing' and calculates how much the average investor is likely to lose by trying to beat the market instead of simply picking up an ETF that tracks the US market. He looked both at the average cost of buying and selling stocks and their performance compared with the costs and benefits of holding a US market ETF. In 2006, the year for which he had the most comprehensive data, these difference came to a staggering US$99.2 billion.

In other words, if investors just passively tracked the wider market through a low-cost instrument such as an ETF, instead of paying someone to actively pick favoured stocks, they would have been better off to the tune of US$99.2 billion.

French is concerned about the costs of trading itself, but what of the more fundamental matter of overall performance of the underlying stocks? A US-based investment firm, Dalbar, published a 20-year study last year which calculated that the average annual return for investors holding stock funds in the period 1991 to 2010 came out as 3.83 per cent. If they hadn't bothered with these actively managed funds and simply bought an ETF tracking the S&P 500 they would have achieved an annual average gain of 9.14 per cent. Investors in the collective Hang Seng Index in this period would have done far better, registering an average annual gain of 14 per cent, excluding income from dividends.

How many managed investment funds can show returns of this kind over a 20-year period? The answer is very few indeed, yet investors in mutual funds pay far higher fees for worse performance. They are typically charged 5 per cent of their total investment on subscription, and annual management fees averaging 1.4 per cent.

By comparison, ETFs are sold like shares. The subscription fee is a brokerage fee, which tends to be small, perhaps 0.25 per cent. Annual fees and expenses are moderate, too. For example, the HSBC MSCI Hong Kong ETF levies a total annual expense (including management fee) of 0.50 per cent.

Many investors still believe their fund managers have that extra something that will beat the boring old indices tracked by ETFs.

Lamentably, this belief has crept into ETFs, where an element of active-investment strategy has been introduced. Some new ETFs are also more narrowly defined, reducing the universe of the stocks they cover.

But mainland investors needn't worry about this as long as their ETFs have broad coverage. The only people complaining are fund managers faced with a loss of fees.