I received a comment by Josh Betts, an employee of Powershares, regarding my views on PGF, which I thought was worth repeating here and addressing:
An important factor that needs to be considered when evaluating securities is their tax consequences. Roughly ½ of fund investments reside in a qualified account, however, this still leaves the other ½ vulnerable to inefficient tax managed funds. It should be noted that when comparing the yields of closed-end funds to the PowerShares Financial Preferred Portfolio (ticker PGF), one of the main filters in determining which securities are eligible for inclusion in the fund is that the dividends paid by the preferred stocks must be 100% qualified dividend income eligible. This significantly improves the after-tax yield of PGF in a taxable account, which is what investors are ultimately concerned with, as this is how much ends up in their pocket. Also, while closed-end funds can offer some upside appreciation potential when purchased at a discount, they are also subject to the risk of trading at an even greater discount to their NAV. Exchange-traded funds however, tend to trade at or near their NAV prices as specialist are able to create and redeem shares with the fund in larger blocks of 100,000 shares. This helps to address the supply/demand of shares and helps ETFs to trade much tighter to their NAVs.
First of all, thanks to Josh for his input. The two main points that he brings up are the issue of the qualified dividend tax break, and the fundamental difference between ETFs, which have virtually zero divergence from NAV, and CEFs which can diverge from NAV a good deal.
My 2 cents on the second of these issues: Yes, of course ETFs trade at (or very close to) NAV. But the comparison I was making was aimed at showing the poor yield of PGF. PowerShares could, if they so chose, pick a different index, with higher-yielding preferred shares, and achieve a higher yield within their ETF format. In this respect, the question of discount or premium is irrelevant - if there were other preferred ETFs I would have used them as a benchmark, but there are none that I know of.
The first issue, of tax efficiency, is only relevant to retail investors who are natives of the USA, which I am not, and from the statistics provided by Statcounter, a lot of my readers are not either. As a foreign national who pays his taxes to a whole other government, qualified dividends are irrelevant to me. They are also irrelevant, or nearly so, for American investors using retirement tax deferred accounts.
As a rule, I wouldn't comment on a tax system that is irrelevant to me and I know very little of. This is also why my blog will not deal with municipal bonds in the foreseeable future. But of course Josh has a point, that qualified dividends are very relevant to many investors, and should be factored in on an individual basis. I am sure, however, that this will not change the fundamental result of my analysis, namely that PGF doesn't yield as much as I'd like to see from a preferred shares ETF.