Category: Issue Comments

I had lunch today with an industry professional who expressed doubt over my fears (expressed by my according a high weight to Yield-to-Worst) that PWF.PR.J would be redeemed early. So, first, let’s look at the situation – the redemption schedule is:

Since the closing quote on January 25 was $26.82-88, this give rise to the following information from the pseudoPortfolioReportBox:

Perhaps the exact dates, third decimal places and probabilities are arguable, but I think the basic facts above can be agreed by all practitioners.

So, in the HIMIPref™ analysis, the Yield-to-Worst is given a large weight in the REWARD_CLASS_YIELD component of valuation. In addition, the issue is tagged with the tradeSizeCalculationNotes tag of “14” indicating

If pseudoModifiedDurationWorstBid is less than minWorstBidPseudoModifiedDurationBuy, buySize and sellSize are set to 0.

Therefor, HIMIPref™ will not, given current market conditions, recommend the purchase of PWF.PR.J, since the internal calculateTradeSize function will always return a buySize of zero.

My companion said this is a load of hooey. PWF.PR.J has the lowest annual dividend of any of the Power Financial issues and therefore, if they redeem anything, it will be something else. Therefor, investors may assess this issue on the basis that it will survive until its softMaturity … maybe with an allowance for the risk of an early call, to be sure, but the default case is a pre-tax yield of 3.45% until 2013.

Well – I disagree. By me, the worst case is the default case. Remember, a pessimist is an optimist with experience. But let’s look at some details:

First: I agree that PWF.PR.J has the lowest coupon of any PWF issue … readers may confirm this by looking at the appropriate rows of the table on prefInfo. But let’s have a look at some of the details of these issues:

Now … first off, let’s just guess that PWF could sell new retractibles with a $0.90 coupon. This looks kind of skimpy, but consider the CGI.PR.C, issued last March with a $0.975 coupon, and MFC.PR.A, trading to yield 3.13%, and of course the yield-to-retraction on the PWF.PR.J themselves of 3.45%. In light of this, a coupon of $0.90 on a new retractible doesn’t look all that much out of place.

New perpetuals … well, Sunlife, an equally rated credit (by DBRS) is in the process of selling (or has sold, depending on the terms of their underwriting!) perpetuals with a coupon of $1.125.

If, instead of looking at the gross coupon on PWF’s preferreds, one looks at the savings achievable by refinancing with an issue of similar characteristics, then a lot of the PWF.PR.J cheapness to the issuer disappears.

And that is still not the end of the story! When we look at the Annual Report for 2005, we find a note that regular readers will have got thoroughly sick and tired of by this time:

Effective for fiscal years beginning on or after November 1, 2004, CICA 3860, Financial Instruments — Disclosure and Presentation was amended to require obligations that an entity must or can settle by issuing a variable number of the issuer’s own equity instruments to be presented as liabilities rather than equity. On January 1, 2005, the Corporation adopted the amended standard retroactively with restatement of prior periods. Some of the Corporation’s ($300 million) preferred shares were reclassified from Shareholders’ equity and some of the subsidiaries’ ($1,366 million) preferred shares were reclassified from Non-controlling interests to Liabilities and the associated preferred dividends were reclassified to Financing charges in the Consolidated Statements of Earnings. The change does not have any impact on earnings per share or net earnings available to common share holders since preferred share dividends were previously deducted from net earnings in determining net earnings  available to common shareholders.

The moral of the story being: for reporting purposes, PWF.PR.J is a BOND, not a component of equity, and will get lumped in with bonds when reporting the all important coverage ratios.

So let’s have a look at some bonds. The Great-West Lifeco bond, 6.14% of March 2018, is currently indicated at a spread of 68bp over the Canada 4% 2016. The Canada 4% 2016 are (or were, anyway, at the time the indications I’m reading were prepared) trading to yield 4.208%, so we can say the GWL bonds of 2018 are trading to yield, oh, call it 4.90%.

Now … the important thing about preferred dividends is that the company has paid tax on them. This is an area in which I will not pretend to any great expertise, but let’s just assume, for the sake of an argument, that the company’s average tax rate (disclosed in Note 8 of their annual report) of 26% is equal to the marginal rate they would save if they magically converted the dividends on their retractibles to interest on bonds and were entitled to deduct it. We’ll make this a more familiar by calculating 1/(1-0.26) = 1.35 and calling this the equivalency factor … fortunately, this is pretty close to the Equivalency Factor for Fat Cat Ontarians of 1.4x that I normally use for illustrative purposes.

So … getting back to PWF.PR.J … commencing 2008, the company will, effectively, be borrowing $25.00 at an annual rate of a dividend of $1.175, less the annual saving (a reduction in deemedDividend to the shareholder, so I presume it costs the company extra as well) of $0.25, for a net cost of $0.925. This is an annual rate of 3.7. Multiply by the company’s equivalency factor of 1.35x and this is the equivalent, as far as their treasury department is concerned, of 5%.

AND, I’ll point out, the pref will only be 4-year money. They can issue 10-year paper for 4.90%. You are the treasurer. What do you do?

So, let me summarize my position on PWF.PR.J:

• Since it is retractable, the proper benchmark as far as the company concerned is a bond, not other prefs.
• 10-year bonds can be issued at a net cost lower than the 4-year pref
• At the very least, the chance that the PWF.PR.J holder will get called early and experience the worst-case scenario is significant.
• Analysis, schmanalysis. We’re fixed income investors here, and we hate risk, especially risks that depend on the day-to-day tone of the market and decisions made on the basis of an individual company’s capital structure
• There are issues out there that carry significantly more yield
• Even if we know, absolutely for certain, that the issue will hang on until the day before retraction, the yield is only 3.45%. At an equivalency factor of 1.4, this is an interest-equivalent of 4.83%.
• Why don’t we simply buy a bond that yields this much, that has better credit protection? Just because the PWF.PR.J is classed as a bond-equivalent for balance sheet purposes, we don’t move closer to the front of the line if the company runs into trouble.

Great-West has an issuer bid outstanding for

These are both retractible and, as stated in their 2005 Annual Report:

The adoption of the amendments to the CICA Handbook section on Financial Instruments – Disclosure and Presentation … resulted in the reclassification of the Series D, 4.70% Non-Cumulative First Preferred Shares [GWO.PR.E, jh] and the Series E, 4.80% Non-Cumulative First Preferred Shares to liabilities.

So, they’re being bought back because they’re bond-like for balance sheet purposes. Unlike many Issuer Bids, this one actually has some meaning: 368,200 of the GWO.PR.X disappeared from the balance sheet from 2004 to 2005, as did 21,100 of the GWO.PR.E.

Not the biggest news to roil the markets, but it’s interesting.

Update & Bump : In their third quarter, 2006, financials, Great-West stated:

During the nine months ended September 30, 2006, 1,077,700 Series E 4.80% Non-Cumulative First Preferred Shares [GWO.PR.X … JH] were purchased pursuant to the Company’s Normal Course Issuer Bid for a total cost of $30 [million … JH] or an average of $27.37 per share. The price in excess of stated value was charged to income.

It has been announced that the Bell Canada / BCE Inc. Preferred Share Exchange Offer has been approved:

The face of telecom company BCE Inc. took another step towards being wiped from the Canadian business landscape after Bell Canada shareholders approved exchanging its preferred stock for BCE shares and accepting a one-time payout of 20 cents per share.
…
Shareholders representing 61.5 per cent of Bell Canada preferred stock voted 95.26 per cent in favour of the moves.
A couple of investors who attended the meeting voted against the motion, but later refused to explain their decision or reveal whom they represented.

According to the DBRS report dated December 12:

DBRS notes that the Company’s announcement to simplify its corporate structure will not result in DBRS consolidating the Bell Canada and BCE Inc. ratings, as the holding company and operating company structure will still exist from a legal perspective. Thus, DBRS will maintain a one notch rating differential reflecting structural subordination. Although DBRS does acknowledge that debt levels at BCE Inc. have decreased substantially, BCE Inc. will continue to be dependent on Bell Canada for the financing of its interest, preferred and common dividends on a go forward basis. Although the level of dividends from Bell Canada that will need to be up streamed to BCE Inc. are expected to decrease, DBRS believes that this will be somewhat offset by Bell Canada absorbing the BCE Inc. corporate expenses.
…
Finally, DBRS notes that the Company has indicated that under a plan of arrangement, holders of Bell Canada preferred shares will be asked to exchange their shares for BCE Inc. preferred shares with the same series rights. The arrangement must be approved by the holders of common and preferred shares of Bell Canada, each voting as a separate class, at a special meeting to be held on January 23, 2007. As previously stated, BCE Inc.’s preferred shares are rated one notch lower, reflecting structural subordination.

So: the Bell Canada pref holders have given up a notch of credit rating for a one-time payment and no extra dividend. Let’s hope it works out well for them.

I mentioned one half of this pairing in the January 17, 2007 report and I’ll just make things more explicit here.

This pair is very similar to the BBD.PR.B / BBD.PR.D pair that has been discussed previously. They are convertable into each other on December 1, 2007, just over ten months off, but are trading at prices that are very different.

At the moment, for instance, BCE.PR.Z is quoted at 26.38-75, 1×9, and BCE.PR.Y is quoted at 24.86-90, 4×9. Between now and December 1, 2007, when they can convert into each other, the former issue may be expected to pay dividends at a fixed rate totalling $1.3298, while the Ys, paying a ratchetFloatingRate based on Canadian Prime will pay somewhere around 0.92125. Therefore, the price differential, in the absence of liquidity premia, should be in the neighborhood of $0.40 … but it ain’t. The differential bid/bid is more like $1.50, implying (in this very simple analysis) that there is $1.10 just lying around waiting to be scooped up.

I recognize that the BCE.PR.Y (the cheap ones) are relatively illiquid, due to their float of only 1.1-million-odd shares. The Zs (the expensive ones) have a float of 8.8-million-odd shares. So, there will be problems exploiting this inefficiency for those who care to try it.

But … what I can’t understand is: why would anybody hold the Zs? Even if they’re unable to buy sufficient Ys to replace them? BCE can put an extremely low rate on the “Z” dividend payout commencing December 1, 2007, all but forcing conversion. Surely nobody seriously believes that BCE will leave the rate as it was set five years ago, at 5.319%. In BCE’s last ratchet/reset, the BCE.PR.S / BCE.PR.T the rate was reset to 4.502% … and I certainly wouldn’t bet on this very generous payout being offered again on the Y/Z reset date.

Update: The Zs actually went up in price today (2007-1-18) (bid/bid), closing at 26.39-75, 3×8. Why? Surely there will be few who disagree that the projected price as of the Dec. 1 conversion date is $25.00 … and if you do disagree, please leave a comment, because I’m interest. So the projected capital loss (from the bid price) exceeds the intervening dividends – expected total return over the next 10 months is negative. I don’t understand ….

Today’s Globe & Mail contained an article by Rob Carrick that mentioned preferred shares.

Riccardo Palombi, a salesperson at the Manitoba-based McLean & Partners had a few words to say:

Mr. Palombi of Mclean & Partners suggests sticking to preferred shares issued by the big banks and other top-quality issuers. As an example, he mentioned the TD preferred series O shares, which pay $1.21 in dividends a year and currently yield about 4.6%.

So, I thought I’d write a bit about TD.PR.O today.

The option schedule for TD.PR.O is:

 A perpetual, paying $1.2125.

Firstly, the 4.6% Carrick mentions is currentYield and I’m saddened, but not surprised that Carrick mentioned it in his article. As readers of my article A Call too, Harms know, I’m not a big fan of Current Yield and greatly prefer yield-to-worst as a measure of preferred share value – assuming, of course, that I’m writing for general publication and am only allowed a single measure of value!

The pre-tax YTW of TD.PR.O is 4.14%, based on the January 12 closing bid of $26.15. So the first thing we want to know is: why accept 4.14% when there are new issues (new bank issues, what’s more, from Royal, Scotia and BMO that yield 4.50%?

One possibility is the implicit degree of interest rate protection afforded to investors by the higher coupon. The TD issue pays $1.2125, as mentioned above, which works out to 4.85% on the original issue price. If rates rise, then all fixed income issue will be hurt, but (for the first little while, at least) TD.PR.O will have some protection, because it will still make sense for the issuer to call the issue at the same price as it would have called them in the absence of a rise.

If, for instance, all perpetual preferreds are trading at 4.80% (pre-tax) in 2014, then we will expect TD.PR.O to be redeemed at $25.00 (or trading slightly above that price), whereas one of the current new issues, paying $1.125 p.a., will be trading at around $23.40, at which price they will be yielding the 4.8% imposed by these hypothetical market conditions. In other words, they will have lost about $1.60 in value, compared to only $1.15 in value for the TD.PR.O. Additionally, the TD.PR.O will have paid about $0.09 more p.a. as dividends.

When HIMIPref™ is used to analyze the cash flows of TD.PR.O for the YTW scenario, we get the the attached report from the cashFlowDiscountingAnalysisBox. This report can also be saved as a text file and uploaded to an Excel spreadsheet.

I hate using Excel spreadsheets to explain things. At some point I’ll write a little feature into HIMIPref that will do this automatically, but that’s way down the list. The purpose of HIMIPref™ is to analyze preferreds write blog posts! While this sort of analysis is implicit in HIMIPref™ it’s buried pretty deeply, in things like curvePrice!

On the tab “Initial Analysis” of the attached spreadsheet, the data provided above has been put into Excel format. Additionally, equivalent data for the RY.PR.? new issue has been approximated by multiplying the cash flows for TD.PR.O by a factor of (4.50 / 4.85) to account for the reduced coupon. The cells highlighted in yellow have been further changed, to reflect an estimated value of $25.00 for the RY.PR.? on 2014-11-30: that is, this analysis projects no change in market interest rates between now and the analysis end-date.

When we sum the values of the individual flows, we find that the net present value for TD.PR.O is, indeed, about $26.10 (there’s some rounding error. So sue me.) which of course it should be since the discounting factors are derived from the Yield that results if it is redeemed on the End Date.

We are amazed and astounded, however, to note that the cash flows of the RY.PR.? new issue sum to about $25.40, which is forty cents more than the price we have to pay for it now. Bonus! Using this analysis, we can say that the TD.PR.O is fairly priced (by definition) but the RY.PR.? is forty cents cheap! So why buy the TD.PR.O.

Some scenario analysis is done on the “Scenarios” tab of the spreadsheet. For each presumed market yield, we calculate the price of each issue, being careful to cap this value at the appropriate redemption price. Then we account for tax effects to derive an exit value. We use the discounting factor from the “Initial Analysis” tab to compute the present value of the exit value, add this to the present value of the dividends, and then come up with the present value of the whole package. In columns “Q” & “R”, we compare this discounted present value to the actual market price to see whether it’s cheap or expensive, given the scenario for market yields. Obviously, if our scenario is for rising yields, they’re both expensive. Any fixed income will be! But the degree of protection has been calculated.

I’ve prepared a chart:

So, if you want some protection from rising interest rates, you may well prefer TD.PR.O to the new bank issues, accepting the fact that this will probably be an underperforming choice if rates are unchanged from this time until the call-date.

I consider this analysis to be very approximate and do not explicitly use it in HIMIPref™. I’m more interested in curvePrice, the price at which an instrument should theoretically trade if all its features are valued the same way as similar features on similar issues, and at Yield-to-Worst, these being two major components of valuation:

A full HIMIPref™ analysis shows this issue roughly comparable to one of the new bank issues. But I like RY.PR.B & RY.PR.C better in that “bank perpetual” space. 

5Banc Split Inc. has announced (via CCN Matthews) that:

it has completed the issuance of an additional 100,000 Class B Capital Shares (the “Capital Shares”) and 100,000 Class B Preferred Shares (the “Preferred Shares”) at prices of $10.00 per Capital Share and $10.00 per Preferred Share pursuant to the exercise of the over-allotment option granted to the Company’s agents in its recently completed offering of Capital Shares and Preferred Shares. All together, the Company has raised total gross proceeds of $282 million under the offering.

The FBS.PR.B new issue closed on December 15. This is good news. Always nice to see these things get a little more liquid.

I was surprised at the hostile reception accorded BNA.PR.C, the new issue that settled today after being announced December 20.

It traded in a fairly narrow range, 24.70-85, on heavy volume of 798,550 shares. The closing quotation was 24.74-78, 10×10.

More later.

Much later, more: I’ve uploaded the Split-Share sub-Index Portfolio Evaluation. Interested readers should be able to tell with a glance at the “Yield-to-Worst” column (and a peek at the “Modified Duration – Yield to Worst” column) that I do have some basis for considering this issue undervalued!