Is there a risk of crowding in smart beta?

Alan Pickering: If there are concentrations in particular market sectors then there are greater opportunities in other market sectors, some have argued, which almost leads you to think that soon someone will invent the dartboard approach to investment to take advantage of random gaps that might appear in the market.

It is impossible to know when risk factors are crowded, or when they are going to go up or down

Eric Shirbini, ERI Scientific Beta

James Price: Smart beta is quantitative in its nature. Post-2000, there was a big run-up in quant assets, until we had the quant crunch in 2007, which was a crowding issue. There was then a period of time after that where quant investing was a dirty word around many trustee tables and they did not want to do it.

Smart beta has, in many ways, emerged from that. It was a bit more palatable at the beginning, and it was simpler. It is now becoming more complicated and is converging with the previous quant strategies in many respects. So there is definitely a cycle that we have gone through. Will that crowding effect happen again? Undoubtedly it will.

Eric Shirbini: I do not know what people mean by crowding with regard to smart beta. I know what it means when you talk about hedge funds, but as far as risk factors are concerned, there is no such thing as crowding. The biggest risk factor of all that you have is the equity market risk factor, the FTSE 100 itself. Can you turn around and tell me when the FTSE 100 is crowded? It is impossible to know when risk factors are crowded, or when they are going to go up or down.

All you have to do is look at the low volatility factor very recently. People have been talking about low volatility being crowded for the past two years and nothing ever happened. But what did happen during the past six months is that there was talk of interest rates going up.

When interest rates go up more quickly than the market had originally anticipated, then low-volatility stocks are going to suffer. Low-volatility stocks have suffered, but I do not believe that they have suffered because low volatility got crowded. They have suffered because of exactly the reason it should have happened: that low-volatility stocks, utilities, are more sensitive to interest rates. And you can talk about that for any factor; there is a reason associated with every single factor.

You should only stick to factors that have been very, very well documented that we know a lot about, which is a handful of factors, four or five. Factors will go up and down, but the word ‘crowding’ is not the right terminology.

Julien Barral: We have seen crowding, or concentration, in certain sections. Some smart beta strategies would focus on certain stocks and that would make valuations higher than they should be.

The smart beta ethos for some of these strategies was not applied because it was not necessarily diversified as much as it should have been. So there is a risk that you are not really balanced, you are not looking at your exposures in the right way.

Shirbini: Yes, but that is a different argument. It is that you can get people investing in lots of stocks, but that is not the right way to invest in factors. You should be well diversified. To know when a factor is going to turn around is like looking into a crystal ball.

Barral: If you could time factors, then obviously you would potentially be considered an absolute return strategy or hedge fund. We see strategies trying to incorporate factor timing and we think it is a very difficult thing to do.

The strategies that no one is worried about are actually where the crowding probably is

James Price, Willis Towers Watson

Price: ‘Crowding’ is a nice catch-all term for lots of different things that people are worried about. Let’s say it all starts by buying cheap companies, and those companies become less expensive. The future returns from those cheap companies versus the market are then lessened over time. That is not the factor ‘breaking’, that is just part of a natural cycle of people allocating and deallocating from areas of the market.

Returns from strategies are not stable, and if everyone starts investing in something then returns probably will go down over some period of time.

The question then becomes one of product design. A product might be very sensitive to certain inputs, which means that when it goes wrong it underperforms by an unexpectedly large amount, and that is the kind of product that catches people unawares and is not one that, if you are trying to harvest a long-term premium, you necessarily want exposure to.

Shirbini: Which is why you need to create a well-diversified factor portfolio.

Paul Black: Value investing has worked in the past because investors generally don’t like unexciting stocks, so these have a price to earnings ratio lower than might be justifiable compared with the market as a whole.

However, as more investors mechanistically search out this value premium, the P/E ratios of these stocks increase towards, and possibly past, a reasonable level. If this situation occurs, then the value premium will not be there.

But all it takes is for some investors to stop searching for this value premium for the pricing to revert to the previous levels, and once again there should be a premium going forward for value investors. So if a factor does become crowded, it might only be for a short period.

Price: Predicting how these things will do is very difficult. The question of whether strategies are crowded or not is the right one to be asking to understand the risks. As long as everyone is asking whether strategies are crowded, it is probably a reassuring sign that maybe they are not too crowded. The strategies no one is worried about are actually where the crowding probably is.