Thursday, September 06, 2012

New Momentum Fund

Moment is a strange factor. Historically its magnitude is comparable to value and size in Ken French's data, yet clearly less popular. Yet unlike size, value and more recently low-vol, hardly any products out there attempt to specifically harvest the momentum premium in an efficient manner. The problem is that momentum peaked in November 2008, and has not recovered (graph below using French's data).  Many think it may never recover, which perhaps is why the main momentum fund, HMTM, trades only ten thousand shares a day.

Nonetheless, my friends at Robeco launched a fund designed to capture the momentum premium (led by Willem Jellema, right). They think it will be attractive because momentum tends to do particularly well when low volatility and value fails (e.g. late nineties was horrible for value and low-vol, but great for momentum). More importantly, their fund has some secret sauce.  I can't say what that is, but I can vouch for the fact that there is a way to play momentum that works much better than the standard approach used in academic papers.

Update: I should add that AQR runs momentum funds, and Cliff Asness has been actively researching  momentum for well over a decade (see Value and Momentum Everywhere), so he probably has some special sauce too.


mo said...

You can't say directly, but you can offer some areas of study for us to look into how to play momentum?

Anonymous said...

Here is a clue from Robeco:

We believe that we are able to avoid the main pitfalls which are often associated with momentum investing: high downside risk and high turnover. Robeco has a long experience applying the momentum factor in quantitative portfolios with limited turnover. Robeco's proprietary Residual Momentum technique delivers a superior risk-adjusted net performance by limiting unrewarded risk.

Not sure what "Residual Momentum" is in this context . . .

Anonymous said...

Instead of teasing us, you should have just provided the following link:

vonjd said...

@Anonymous: Good work! Thank you

I think Eric just wanted to check if his ecosystem is up to speed ;-)

Anyway, this paper supports my hypothesis: You could publish the golden grail (which this is not) - and nobody would even care but continue to hunt the hot stock tips.

See also here:

Most people in this area who try to hide their ├╝bersecret sauce are just posers who try to impress suckers (and get their money). Only invest in vehicles where the creators are able and willing to explain everything to you! (At least this is my take on it).

Will have a look at the paper - cheers again

Anonymous said...

It doesn't appear that Robeco has a version of this fund for the individual investor unless I'm not looking in the right places. Shame....