A Model ETF Portfolio With Recognizable Flaws

Roger Nusbaum submits: I took a stab at creating an ETF portfolio last night (it's listed below). The idea was not to have only a few ETFs and hope it is diversified, but more along the lines of trying to recreate the sector weightings, and some of the other things I have built into the accounts I manage.

According to Morningstar I came very close on the sector weighting. That is not a surprise, given the way that ETFs are structured -- you know what you are getting:

The portfolio misfired on almost every other aspect of trying to mimic my model portfolio.

The most disappointing was the dividends. The portfolio only yields 0.9%, which is way below the S&P 500. The cap size of the portfolio missed as well, coming in at only $18 billion vs. closer to $35 billion but I might be able tweak that with a broad based market cap ETF.

Foreign was only 21% of the portfolio. With a little bit more tweaking I could work in more foreign, but that might upset the sector applecart.

The performance of the allocation clearly beat the market over the last 12 months, which is encouraging but it would have been better with a decent yield:

All of that said, the portfolio is listed below with the weightings. Perhaps this can be a collaborative effort to make the improvements that are so desperately needed. To be 100% clear -- this is not a portfolio I would implement for anyone, as it is too flawed. This was nothing but an academic exercise.

Financials

StateStreet Bank (KBE) 8%
iShares Australia (EWA) 3%
iShares UK (EWU) 3%
StateStreet Capital Market (KCE) 2%

Tech

iShares Global Tech (IXN) 8%
PowerShares Semiconductor (PSI) 2%
iShares Taiwan (EWT) 2%

Health

iShares Global Health (IXJ) 10%
iShares Medical Device (IHI) 2%
StateStreet Biotech (XBI) 2%

Staples

iShares Consumer (IYK) 8%
PowerShares Food (PBJ) 5%

Discretionary

PowerShares Leisure (PEJ) 5%

Industrials

Industrial Sector SPDR (XLI) 5%
PowerShares Water (PHO) 2%
iShares Defense (ITA) 3%
iShares Transportation (IYT) 1%

Energy

iShares Global Energy (IXC) 5%
PowerShares Alt Energy (PBW) 1%
PowerShares E&P (PXE) 2%
PowerShares China (PGJ) 2%

Materials

Gold (GLD) 3%
iShares Brazil (EWZ) 2%
StateStreet Miners (XME) 2%

Utilities

Vanguard Utilities (VPU) 4%

Telecom

Vanguard Telecom (VOX) 3%
Emerging Market Telecom (ETF) 2% -- this is a closed end fund

REITs

StateStreet REIT (RWR) 3%

Comments

[GeoffConsidine]

Roger:

Nice article. Once again, in a rare anomaly, we agree :)

Just looking at style analysis in ETF's will not be a great idea. There are so many dimensions of asset allocation that cannot be captured by a 'style' breakdown. I think that the points that you are making highlight a major flaw in asset allocation schemes advocated by many advisors. It is not enough to simply say X% in a given sector. Further, the way the total portfolio works--the way that asset returns are correlated or not--is another major issue that you did not even discuss.

If the market cap is a major issue for you, you could certainly change your ETF allocations to get a market cap much closer to what you want, so I assume your point is simply that style analysis by itself is bad, right? You are not saying that you could not recreate a portfolio more similar to your policy portfolio with ETF's--or are you?

[RogerNusbaum]

As I understand <em>style</em... in this context it refers to growth vs. value which I did not even get to for space sake.

In assembling a portfolio, I care about (in no particular order) sectors, countries, style, cap size, volatility, yield and a couple of things I may not come to mind after a long day.

These are all things I have written about in the past. I have never been a fan of all ETF portfolios because they do not allow for easy management of all of the above, at least I find it harder this way as opposed to utilizng all available products.

[GeoffConsidine]

Style analysis is a pretty broad term. In academic circles, Style analysis means attributing the performance of a mutual fund or ETF to contributions from a series of indices (large cap, Euro, etc). Morningstar also calls its size/value/growth boxes 'style boxes' which is what you are talking about. I am simply saying that a generic portfolio in which you say X% to sector A, Y% to sector B, etc. is too blunt an instrument to use in building your best portfolio--even if you believe is fairly passive investing styles. Portfolio performance has more dimensions than simply what shows up using Morningstar's sector analysis. While this sounds obvious, many advisory firms put a lot of weight on a simple 'pie chart' allocation that does just this.