I want to follow up on a post about portfolio construction at the sector level with a look at some ideas on how to construct each sector -- the context being that a portfolio is built at the sector level and benchmarked to the the S&P 500 with decisions about overweighting, underweighting or equalweighting each of that index's 10 sectors.
Making decisions about how to weight a sector can come from a combo of knowing a little about market history and assessing the current environment to make a forward-looking analysis (at least this is how I try to do this).
The obvious way to execute this would be to take a sector -- like industrials, which is currently 11.2% of the SPX -- and allocate 13-15% to it for an overweight or 8-9% to underweight it ... or any other numbers you prefer.
Another way to come at this is by increasing or decreasing the volatility. The energy sector provides an easy example. The Energy Sector SPDR (XLE) owns the energy stocks in the S&P 500 and so could be thought of as the benchmark of the sector. In the context we are talking about, an equalweight to energy would be buying XLE in exact proportion with the S&P 500.
One way to underweight the sector would be to own stocks and/or funds in the same proportion as the SPX but with less volatility. A combo of Exxon Mobil (XOM) and a partnership name like Energy Transfer Partners (ETP) will very likely (no guarantee) be far less volatile than XLE. This might be the right type of exposure after some sort of crazy spike in oil prices that brings out many calls for $200 or $300 oil, or if the economy appears to be rolling over into a slowdown.
The other side of the spectrum could be a combo along the lines of Global X Uranium (URA) and the PowerShares Small Cap Energy ETF (PSCE). It doesn't take a lot of imagination to see where these two are going to be much more volatile (no guarantee) than XLE. Putting on this type of a combo might make sense after a large drop in the market or a drop in oil prices.
A third possibility is a combo from the two groups for some desired purpose. Something like 75% of the energy allocation into a low volume name like ETP and the rest in something high volume like URA could generally dampen the volatility, but allow for growth the next time the volatile holding doubles. At some point things like uranium or some other niche will double, leaving XLE in the dust ... but until then, someone could capture some yield with tilt to the low volume name. The above are just examples; we don't own any of the stocks or funds mentioned.
The bigger idea here is to think about how each sector is comprised. It may make sense to go low volume in one sector but high volume in another. Blending sector volatilities can allow for creating a portfolio that is more or less volatile (depending on preference) than the broad index that serves as the benchmark.
This is the type of stuff I have in mind when I talk about creating very specific effects in the portfolio. The narrower you go, the more precise you can be.