AIG Deal Death Blow To ETF Industry?

September 17, 2008


Any thoughts regarding the other email that I sent "Sectorally equally-weighted ETFs???" --- ETFs equally (or otherwise) weighting sectors?  Isn't that a pretty simple concept? 

I suspect you guys have a whole lot more pull with fund sponsors and index providers than I (actually, I've got practically zero, based on my inquiries/experiences over the last 3+ years). Is it just because I'm communicating through wholesalers and wholesalers, in general, know squat and are, for whatever reasons, disinclined to pass along "egghead" questions like mine? Don't other advisors look beyond the name of an index to arrive at an understanding of what they're getting their clients in to?  I got to believe that I'm not the only one out there throwing questions at these folks. Pretty damned frustrating (and one of the reasons, I suppose, that I'm not so quick to pat sponsors and index providers on the back).


Matt and I discussed this a bit, and his initial reaction was similar to mine ... that the general direction of thought (not overweighting sector bubbles) makes sense, but logistically, it probably doesn't. You're going to weight telecom or utilities equally to, say, energy? That sounds like a bit of a wild bet. I went into the data and looked at historical sector weighting (just for the S&P 500, attached) to see just how much of a wild bet. Take a look at the attached. Actually the crossover in weighting across pretty much ALL of the sectors is dramatic, which would help to make YOUR case.

And you do have a great point about the tremendous financials overweighting.

But my argument would be that the economy evolves, and while there are bubbles, lots of them and big ones, by and large these weightings are representative of genuine economic activity. Maybe you could run some metric that takes a look at valuations (current book, with allowances for growth, etc.?) but running a straight equal-weighted sector portfolio would make me nervous.

As a side note, there ARE RYDEX equal-weighted sector funds ... they equal weight the stocks IN each sector. I know that's not what we're talking about, but I thought I'd mention it.

What I REALLY think you're looking for is bubble protection, right?  So to me a more sensible approach would be a "fundamental" weighting like the FTSE RAFI (and maybe WisdomTree and maybe value) funds. The idea there is to track not market-weighted returns based on market valuations, but to track actual underlying economic activity. Ironically, a lot of those indexes got raked by the problems with financials - what it may be more difficult to account for than irrationality in market valuations, is irrationality or overreaching in economic activity. And there IS such a thing, of course.

My feeling is you need something more sophisticated than just straight equal weighting to deal with the issues you're talking about.

That is my quick take on this, but I'm happy to debate it/think about it some more.

In terms of contacts into the ETF product issuers, it's not like iShares will just launch a fund if we run it by them, that's for sure. But we do know all the people working on product development, and if we DO feel very strongly about a concept, we can likely get SOMEONE to launch it...


I'm not advocating an allocation equally-weighted or otherwise across sectors. Nor am I advocating any particular weighting scheme within each sector. But for discussion and illustration purposes, we've got to start somewhere and, accordingly, I rather generically have talked about equal weight across sectors.   

And while I believe that baseline allocation across sectors probably ought to be something other than straight equal weights (which quite arguably seems to over-weight telecom and, perhaps utilities and materials) I'd rather hold a portfolio of sectors equally-weighted than one weighted according to the migratory extremes that many indices like the S&P 500 engender.  US equities holdings in my portfolios are baseline weighted equally at 10.5% across nine sectors and 5.5% in telecom (excluding tactical over and under-weights).  I fully agree with you regarding the need for a sector weighting program more sophisticated than a straight equal-weighting approach but from what other point to you start? 

Unfortunately, I've not come across research that nails down baseline sector allocations from which any "tactical" over/under-weighting can be registered.  Both Morgan Stanley and UBS research, for example, quote allocation recommendations against the sector composition of the S&P 500 --- i.e., against a moving target --- meaning that they present no static baseline sector allocation levels.  As a practical matter, quite frankly, I could care less about the sector composition of the S&P 500, and how a current recommended allocation might differ from it, except as a curious aside that has no bearing regarding what constitutes a well-reasoned, and at least relatively static, baseline allocation schedule across sectors.

In order for an allocation to truly bring risk moderation to the portfolio it has to be rather agnostic - avoiding exposure overloads (and under-exposure) and, I'd argue, with respect to sectors in particular.

Moderation and less than full participation in bubble extremes (on the upward swing and the downward spiral) is what matters to me and, I think, to many investors and advisors.  And that, quite matter of factly, is not to be found in S&P-500 based, style-based and "broad" market approaches or others that are inclined to significantly participate in sector-weighting extremes (lacking in any or in any meaningful sector constraints). Research groups rather routinely set out base growth/value and capitalization allocation static baselines. Why not on the sector front? 

The reality is you simply cannot tame the sector over-exposure extremes in total market or style oriented alternatives without taking an extremely crude and broad shot-gun approach ...  Underweighting, for example, the whole value side (and value-oriented sectors) in order to simply trim financials and, then, overweighting the growth side in order to have the exposure in sectors other than financials --- is like using a chain saw to cut a bad, nickel-sized piece of cheese from a one-pound block. You can certainly get rid of the bad nugget that way but at what cost to the block?

Thank you for the feedback!


Find your next ETF

Reset All