"Smart beta strategies assume a stock market in which a few chosen factors produce persistent returns," Jacobs and Levy write in a forthcoming piece in the . "This assumption is not a good approximation of what is observed in reality."

While a classic index fund attempts to capture beta, or market return, by holding all shares in proportion to their market-capitalised weight, smart beta funds try to beat the market by making adjustments. Those adjustments are based on "factors", anomalies like, for example, the outperformance of stocks with certain characteristics such as small size, which managers hope will persist.

James Montier of fund manager GMO has criticised smart beta for, in what he says is the vast majority of cases, simply giving investors extra exposure to value and small-capitalisation shares, but at a higher price.

Jacobs and Levy, who more than 25 years ago were early to catalogue and study "factors", go a bit beyond - and to the side - of Montier in their critique. They argue that not only are these factors, on which a smart beta fund must depend, not stable, but also that managers would do better to spread bets across more types of shares and be willing to adjust these over time as needed.

It may well be, contrary to some of the marketing of smart beta funds, that there is no such thing as being "a little bit active" in investment management, just as it is impossible to be a little bit pregnant.

While most smart beta funds overweight stocks with a few criteria, such as small capitalisation, low volatility, momentum and value, these are far from the only characteristics that can lead to outperformance. A recent study by Jeremiah Green, John Hand and Frank Zhang found 24 factors with statistically significant outperformance, but further that size and a stock's price-book ratio, both popular in smart beta, were not among the leading group.

One of the key problems is that a factor that is an outperformer today may not be one tomorrow, or quite the opposite. Take price momentum, the tendency often seen that shares will carry on trading in the same direction, be it up or down. When the market bottomed in 2009, momentum strategies were hit hard.

Another problem, especially with smart beta exchange-traded funds alone accounting for US$350 billion, is too much money crowding into too few factors. Get a forced deleveraging, as happened with a sudden market move in August 2007, and stocks with commonly used factors can be particularly vulnerable.

Similarly, given the fairly static nature of smart beta, frontrunning can be a problem as outside investors try to get in ahead of a planned rebalancing of an index.

Now to be sure, the advantage of a fairly simple and static smart beta approach is that costs are kept down, and perhaps over time will fall further.

In contrast, a proprietary and less transparent approach to capturing and moving among factors will cost more, though it obviously brings with it more opportunity to outperform if the manager gets it right. Diversification, too, will come with less reliance on a few factors like small cap or value.

The other thing to remember, and this is true for all forms of index investing, is that ultimately someone needs to decide when to get in or out of a given fund or asset class.

"Smart beta strategies shift the decisions about the selection of factors and the timing of factor exposures from the investment manager to the asset owner," said Jacobs and Levy.

With more active switching and managing of factors, fund managers take responsibility for investment decisions.

All of this is nothing more than acknowledgement that the world is much more complex than smart beta would seem to imply.

Things change, and an investment strategy that fails to take this into account will inevitably come to grief.