You can’t make a silk purse out of sow’s ear, but you can always flog investments by spouting meaningless figures.
The Claymore Preferred ETF started trading on the TSX today, and the TSX advises us that 300,000 shares of the “Common Class” (CPD) are outstanding, as are 200,000 of the “Advisor Class” (CPD.A). So, assets of about $10-million. It’s a fair start, and it’s a bigger fund than Malachite Aggressive Preferred!
What has raised my ire, however, is their reporting of yield, which was largely supported by the S&P press release and relayed in the Globe & Mail in Rob Carrick’s column:
The yield on preferred shares is unspectacular at 4 to 5 per cent in most cases, but you get preferential tax treatment through the newly enhanced dividend tax credit. In fact, a 4.5-per-cent dividend yield is equivalent to a bond yield of about 6 per cent on an after-tax basis….The yield for the preferred share index is about 4.66 per cent. Factor in the 0.45-per-cent management expense ratio of the Claymore preferred share ETF and you’re left with a real-world yield of about 4.21 per cent.
The Claymore website reports “Weighted Average Yield” as 4.86% and “Weighted Average Dividend (Coupon)” as 5.26%. OK, let’s get things a little straight, here. The last figure, “Weighted Average Dividend (Coupon)” is a fairly meaningless figure, used to give an idea of whether a particular index is trading at a premium or not. Claymore does not specifically define this term, but I don’t see that it can possibly be anything other than Dividend / Par Value.
“Weighted Average Yield”, beloved of Claymore, S&P and Rob Carrick, is the Current Yield – again, the various participants are far too ashamed of themselves to define the term, but it can’t be anything else than Dividend / Market Price. This is a thoroughly nonsensical value to use, as it does not account for amortization of the premium to the expected call date – it assumes that nothing will ever be called.
As I discussed in A Call, too, Harms (and applied to the analysis of other closed-end funds in Closed End Preferred Funds: Effects of Calls), Yield-to-Worst is a much better, conservative, analytical measure than either of the two measures given above. When discussing bonds, for instance, the Globe and Mail does not report “Average Coupon”, or “Current Yield” – they use yield to maturity – equivalent to Yield-To-Worst for bonds with a single maturity and no embedded options.
The calculation of Yield-to-Worst is discussed in my article Yield Ahead, which includes a reference to Keith Betty’s Yield Spreadsheet ((broken link redirected 2024-2-1) (which is linked on this blog as an “Online Resource”).
The holdings of the fund have been published (well, disclosed as percentages, anyway) and Claymore has done a very good job in making the list downloadable as an Excel Spreadsheet. I’ve taken that raw material and filled in Yield-to-Worst from the HIMIPref™ pre-tax bid-YTW calculations.
In a few cases, I had to guess which issues they really meant; in others I replaced a negative YTW with zero – if anything, the figures shown will overstate the actual mean average Yield-to-Worst of the portfolio.
The value is 3.84%.
That’s probably comparable with the value for the other ETFs on the market, but I haven’t updated my calculations for those funds. The holdings of the Claymore ETF itself are an entirely reasonable market index (which means easy to beat! An active manager doesn’t have to hold the stuff with bond-like interest-equivalent yields-to-worst, even ignoring the potential for trading!) BUT THE YIELD CANNOT BE CLAIMED TO BE OVER 4.5% BY ANY RESPONSIBLE PERSON, without an awful lot of caveats to explain to Granny Oakum that there is good reason to believe that such a yield (Current Yield, Coupon Yield) will not be realized as actual money-in-the-bank yield.
This should not be taken as a knock against the Claymore ETF, or as a reason, in and of itself, to avoid the issue. It’s an ETF, it reflects the overall market, full stop. If an investor wants a passive fund then (subject to more intensive analysis), I’m sure it’s basically as good as any other. And comparing the Index Returns disclosed on the Investor Fact Card doesn’t have me quaking in my boots about the potential for active management.
| Annual Returns | ||
| Year | S&P Index | MAPF |
| 2003 | 10.49% | 33.54% |
| 2004 | 5.74% | 13.42% |
| 2005 | 3.30% | 5.92% |
| 2006 | 4.56% | 6.89% |
See the main Malachite Aggressive Preferred Fund page for more information. Past performance is not indicative of future returns, and I most particularly do not expect to see returns such as 2003’s again! I don’t WANT to see them again … that was a result of Bombardier Preferreds going completely crazy and represented a recovery from a horrible 2002. MAPF returns are shown after expenses, but before fees.
But there will always be investors who want a brand name rather than performance, so the index funds will do just fine.
I’ll probably be writing a full analysis of the Claymore fund in the near future. Hat tip to Financial Webring Forum for a discussion of this issue that made me realize how much of the yield disinformation is accepted at face value even by knowledgable retail investors.
Update: For comparison purposes, one may find current yields and yields-to-worst reported daily with the HIMI Preferred Indices, reported in this blog in the Market Action category. The Weighted Average Current Yield for MAPF at the close today was 5.05%; the Weighted Average Yield-to-Worst was 4.42%.
RE: the Claymore Pref ETF
You didnt mention that the index is not static. That is, changes will
be made every year at least that refresh the portfolio and, presumably,
weed out shares that may be called. Thus I’m not sure that the yield to
worst analysis tells the whole story here, as it would on a finite
portfolio of preferred shares.
You’re quite right – any yield measurement is a static measure and defines only what might happen if the issue is held to maturity. In the case of “Yield to Worst”, maturity is taken to mean that call exercise which results in the lowest yield for the purchaser; in the case of “Current Yield”, maturity is taken to mean “Never”.
Note that all preferred shares are callable. I am not aware of any issue that will exist forever, even if the issuer doesn’t want it to.
However, even if all shares with imminent calls are removed from the index prior to call (so that shares leave the portfolio due solely to sales on the market, never by actually being called) the concept of yield-to-worst will still have an impact on fund performance.
Consider, for example, a hypothetical issue that has an annual dividend of $1.25, is callable in four years at $25.00 and is currently trading at $26.00. Then the current yield is $1.25 / 26.00, or 4.81%. Yield-to-worst is an uglier calculation, so we’ll just approximated it by saying you get $1.25 per year dividend and you lose $1.00 over four years (from the market price of $26 to the call price of $25): net income $1.00 annually, costing you $26, therefore YTW (in this very, very rough approximation) is $1.00 / $26.00 = 3.85%.
Now, let us assume that interest rates are static over the next three years and, one year prior to the call date, the issue is removed from the index. I suggest that, given such a scenario, it is very unlikely that the fund will received $26.00 for the stock. This would imply that the market will pay $26 for an issue callable in ONE year at $25, that pays $1.25. You’d be paying $26.00 for an issue for which there is every expectation that you’d get $1.25, lose $1.00 capital, net = $0.25, YTW (at that point) of less than 1%. It can happen – it does happen – I wouldn’t rely on it!
It is much more likely that the issue will be priced at something more like $25.25 in three year’s time.
YTW is not the be-all and end-all of preferred share analysis. I refer to it a lot because it’s a good measure (not a perfect measure, just a good measure) of expected future returns given a static future yield curve. It’s certainly a far, far better predictor than Current Yield! In my article A call, too, harms I compare the predictive power of YTW and Current Yield.
And note particularly that the above examples assume a static yield curve. If rates on new issues suddenly jump to 10% tomorrow, for instance, then we expect a perpetual paying $1.25 to trade at a price much more like $12.50 than $26.00, and at those levels (below the call price) YTW will be virtually the same as Current Yield (small differences related to the timing of dividends will exist).
I am not claiming YTW is perfect. I am claiming only that it is a far, far better predictor of future returns than current yield.
I thought of a much better way to put it!
Yield-to-Worst conveys a certain amount of information. Current Yield conveys virtually no information at all.
[…] I’ve discussed CPD in this blog in the past and I’ve also written a more formal article about it. Now that the July/August edition of Canadian Moneysaver has been published, I can now release this analysis published in their June edition. […]