Much is made of the importance of asset allocation here at ITA Wealth Management. Long-time readers are familiar with the general layout of the portfolios and new readers will find more information by doing a search for “Dashboard.” Not all portfolios tracked on this blog use the seventeen asset allocation classes, but we want to list the stable of assets we draw upon depending on the size of the portfolio and goals of the investor. Cash is included as one of the 17 in the data table shown below. There are actually a few more than 17 asset classes in the following table as I split out Dividends even though we fit those ETFs into other asset classes when actually managing a portfolio. For example, IDV goes into the Developed International asset class and VIG becomes Large-Cap Blend.
Below the title row we have the three cap sizes ranging from large to small and value through growth. Those nine asset classes we refer to as the “Big Nine.” If we eliminate the Blend column we end up with the “Big Six.” I no longer use VO and VB since the other mid- and small-cap asset classes are highly correlated. Some investors may want to tilt the portfolio toward value and one option is to replace VOT with VO and VBK with VB. There is nothing wrong with making this shift toward value. Holding VO and VB still provides some exposure to companies classified as growth companies.
The reason for including VTI instead of VV in Large-Cap Blend is that for some reason, TDAmeritrade still applies a commission charge to VV whereas VTI is commission free. Yes, I realize VTI holds stocks found in all the “Big Nine” U.S. Equity asset classes, but the majority of stocks fit the Large-Cap Blend asset class.
The ETFs coded with the olive green background are my preferred securities. VSS and SCZ are actually small-cap international ETFs, and including them in emerging markets is not completely accurate. Since they include obscure companies I decided to place them in the emerging market asset class.
The ETFs found in the various boxes represent basic holdings and we might substitute iShares or ETFs from other companies. One reason for sticking with Vanguard is cost as most of these ETFs have low expense ratios.
When we move into Commodities, International REITs, etc., we find we need to branch outside the Vanguard family to find ETFs to cover those asset classes.
If you are just starting a portfolio, I highly recommend you purchase a few shares of each of the above ETFs (olive green coded) so you have the means to eventually develop a customized benchmark such as the ITA Index for your portfolio. Stick with commission free ETFs where possible. For example, I will not buy GLD until it is performing above SHY (remember our cutoff ETF) and I am able to add sufficient shares to keep commissions to something below 0.5% of the purchase cost.
Hello Lowell,
When calculating the ITA Index, I noticed on the sample ss you use VTSMX & VGTSX instead of the representative ETFs. Is this the recommended method?
Vern,
Yes, I use the VTSMX index as the benchmark for all U.S. Equity asset classes as that index fund is a good representative (and investable) of the U.S. stock market. There are times when I may not hold all of the “Big Nine” or “Big Six” ETFs and when this is the case it will throw off the ITA Index.
The same holds true for VGTSX where I use it to represent all international investments, be they developed and/or emerging markets. I don’t have a similar index for some of the other asset classes so I need to rely on a primary ETF. RWX is an example I use when invested in International REITs.
This is why the ITA Index is not a perfect customized index, but it sure beats using nothing but the S&P 500, which many mutual funds use as their benchmark.
Lowell
Hello Lowell
One quick question, why you are using GLD, with an expense ration of .40%, while you can use IAU, from iShares, with an expense ration cheaper of .25%, is it the same?
thanks
David,
Good point. GLD and IAU, when compared on Yahoo!, are no different. May as well go with the one with the lower expense ratio. Thanks for the tip.
Lowell
David,
Good comment – http://goldetf.com/2011/gld-vs-iau-which-gold-etf-is-best/ provides a good description of the pros and cons of the 2 ETFs. I have tended to use GLD because of the higher liquidity (although IAU also has excellent liquidity, so not that important a difference) and the lower commission costs due to only having to buy/sell 1/10 of quantity. The 100% allocation to bullion issue may be more important – so this would favor IAU.
David
Dear Lowell
Something called my attention in the asset list, you have selected 4 ETFs for US equity + large-cap + Value, those are VTV, VIG, VYM, DVY. The correlation among those assets is 0.87, so very correlated. What is the rational behind your selection of this 3 ETFs? Would it have sense to you to keep only one? which one would you recommend?
thanks very much
David,
If I were limited to one from VTV, VIG, VYM, and DVY I would stick with VTV as it tracks the Large-Cap Value index. If one is using the Cluster Weighting Momentum model, then it does not matter all that much as the software will extract the top ETF. The advantage of using fewer ETFs right out of the starting gate is to reduce portfolio turnover. That is likely to be an advantage with taxable accounts.
VTV, VIG, and VYM are commission free ETFs with TD Ameritrade. DVY is not so I would eliminate it first.
Lowell
Lowell
For the classification I used information from ETF Database web.
In that web they have VYM – High Dividend Yield ETF, as value ETF http://etfdb.com/etf/VYM
In your classification it is as growth ETF
Thanks
David,
I use Morningstar for my asset class designation and VYM should be classified as Large-Cap Value. Either I made an error or the ETF shifted. In the Hoadley Optimizer, I classify VYM under Dividend.
David – I see my error in the above table. I was thinking horizontally in those last groupings. VIG, DVY, and VYM were all to be Dividend rather than Growth, Blend, and Value.
Lowell
David,
I need to create a new table for this blog entry so as to clear up the value-growth errors.
Lowell
David,
See if the new data table makes more sense. I tried to clean up the misconceptions.
Lowell
Lowell
That is perfect, thanks very much
David