Category: Reader Initiated Comments

I have recently received an eMail from an Assiduous Reader who, having read the post Shorting Prefs, writes in and says:

I read the “Shorting Prefs” discussion dated February 19th, 2009 under the archive for ‘Reader Initiated Comments’ in your PrefBlog. I find this subject very interesting as it appears to be a very low risk way of profiting from inefficiencies in the pricing of pref shares.

As an example, I recently found two issues of Royal Bank perpetual prefs with almost identical annual dividends, but one yielding 6.5% and the other 6.2%. A study of historic trading revealed the two issues often traded at the same yield in the past year or two. Subsequently, I bought the higher yielding pref and shorted the lower yielding pref. If and when the two prefs trade with the same yield, my maximum profit will be about 4.6% (=0.3/6.5) on one side of the trade on a before tax basis.

I have found shorting pref shares online to be extremely easy with TD Waterhouse. There appears to be no problem in selling any pref short. I have sold five different prefs short to date. The uptick rule does apply, so sometimes you have to wait a while for your order to be executed, but this is a minor problem. Also the short interest positions on pref shares are readily available from Stockwatch.com, which is free for the first 30 days. The short interest positions on most pref shares appear to be very low (less than 1% of the shares outstanding).

My question is on holding a short position through the ex-dividend date. I realize I will have to pay my broker (TD Waterhouse) for any dividends that are paid out on my short position. But, am I liable for the extra dividend tax credit on the short position as well? Don’t short positions create extra dividends over and above what the issuer is paying out? This means extra dividend tax credits will be created. Does the Canadian government get stuck with paying out extra dividend tax credits because of short selling? These are questions for which I cannot find any answers on the internet relating to Canada, although I did find some information relating to the U.S. In the U.S., it appears that the extra dividends paid out by short sellers may be ineligible for tax purposes, according to one website. Brokers could then assign these ineligible dividends to tax neutral accounts. I am sure a tax paying investor would be extremely upset if he found out his dividends were ineligible because his shares were sold short without his knowledge.

I would appreciate any insights you have on the topic of short selling and dividend
tax credits.

The tax question came up in the post BBD.PR.B / BBD.PR.D, where, to my chagrin, I was punished for stepping outside my specialty with respect to the application of tax laws.

All I’m willing to do with respect to the tax laws now is refer to Section 260(6) of the Income Tax Act:

(6) In computing a taxpayer’s income under Part I from a business or property
(a) where the taxpayer is not a registered securities dealer, no deduction shall be made in respect of an amount that, if paid, would be deemed by subsection 260(5) to have been received by another person as a taxable dividend; and

(b) where the taxpayer is a registered securities dealer, no deduction shall be made in respect of more than 2/3 of that amount.

However, with respect to the inefficiencies of the marketplace, I will agree that a long-only account can occasionally pick up a little extra yield by swapping between similar issues. An example of this was given in the post MAPF Performance: August 2008.

An Assiduous Reader writes in and says:

You are very bullish on Perpetual Discount preferred shares.

No I’m not. I’m neither bullish nor bearish. I will go so far as to say that at this moment in time and speaking very generally, I prefer PerpetualDiscounts to other preferred share classes as I believe their net present value of future cash flows and market price exceeds the net present value of future cash flows and market price of the other classes.

Are you not concerned with the reduction of liquidity that is caused by the issuer buying the shares at market prices versus purchasing them at the Call price? Some of those shares also have American style Call options that make them even uglier.

It is possible that issuers could buy up their extant PerpetualDiscount issues to the extent that trading volume would suffer; but I am not aware of this ever having happened.

However, just because something has never happened before doesn’t mean it will never happen – just ask a sub-prime paper mogul contemplating housing prices 25% below peak! So it’s always worthwhile to consider.

The issue of issuer repurchases was last discussed on PrefBlog last fall, in the post Repurchase of Preferred Shares by Issuer, in which I mentioned one of the rare non-split-share issuer bids, Great-West bidding for GWO.PR.E & GWO.PR.X, large issues that will retract in the relatively near future. That particular repurchase was a fizzle.

If the majority of the issued shares have been purchased back by the issuer then you are left stranded and have to hope that they purchase the rest to get rid of the nuisance.

The following example displays the benefit to the issuer:

If I am interpeting this table correctly, my interlocutor is saying that a new issue of 16-million shares of Fixed-Resets could be issued with an initial coupon of 6.70% for a cash receipt after underwriting costs of $388-million. These funds could then be used to buy up an extant issue of 16-million shares of PerpetualDiscounts with an original coupon of 4.7% that is trading at 15.88 to yield 7.4%.

The net effect of this action would be $134-million net new cash to the issuer with a net increase in dividend expenses of $1.8-million quarterly, or 5.38% p.a., which would be a nice way to finance.

Well, all I can say is that it’s not happening yet! I can think of several reasons for this:

• Inability to purchase significant stock at a 7.4% yield. If the price of the PerpetualDiscount in the example went up to $17.54, it would have a current yield of 6.7%, the same as the putative new issue. All that would happen, I think, is that you’d see a pop in the market price for the duration of the buying programme and people like me would say ‘thank you very much’ and swap into other issues.
• There would be no change in Tier 1 Regulatory Capital, except to the extent that a profit on cancellation was recorded. While the market price of the PerpetualDiscounts may only be about $16, it’s still on the books as $25.
• Even at 7.4%, the interest-equivalent yield is only about 10.4%. The banks are targetting a ROE in excess of this figure; therefore they would rather repurchase common equity than the PerpetualDiscount

I am not discounting the notion that liquidity might eventually dry up in the PerpetualDiscount market. In a recent post I highlighted the downward trend in PerpetualDiscount Average Trading Value and Assiduous Reader prefhound commented that It sure looks like the fixed reset pref has stolen some of the trading volume from discount prefs.. This may or may not be a factor in the recent elevated spreads against corporates. I suspect not, but it’s something of a chicken-and-egg problem and I’ll reserve judgement until the credit crunch is over!

Volumes are – to date! – sufficient to allow active trading, but we’ll see. It is possible – unlikely, I think, but nevertheless possible – that PerpetualDiscounts could go down the same road travelled by banks’ 100-year floating rate bonds, that were so popular in the eighties and now trade by appointment only at an enormous spread.

Obviously, the fixed resets can suffer the same fate! The only saving grace is that they have a floor on the yield that makes the probability of the option being called higher and saving you from being orphaned.

It would be interesting to know what the median lifetime of a Canadian Bank preferred share before total purchase/recall.

In my essay Are Floating Prefs Money Market Vehicles?, I reported that the average life of called straight perpetual issue was 10.2 years; i.e., just a little over the normal 9-year period before an issue with standard terms can be called at par (a call at $26 is normal after five years, declining by $0.25 annually).

So there you have it, such as it is! Forecasting future prices is chancy enough; forecasting liquidity is worse. All you can do is stay alert for changes and stay diversified.

An Assiduous Reader writes in and says:

I am a private investor and have been unable to get a satisfactory answer from any investment advisors/bank personnel on a question about the new flurry of rate re-set preferreds. They all refer to the rate being re-set in 5 years at a level of x% above government of canada bonds (let’s use a 4.5% re-set as an example). If in 2014 the appropriate GoC bond yield is 5%, and assuming the pfd is not called at that time, then everyone I have spoken to says the pfd rate will be re-set to 9.5%. However, a literal application of the description in the prospectus would be 5% GoC plus 4.5%, which would be a re-set rate of 5.225% (5+[5x.045]). The prospectus refers to a percent, not basis points. Could the banks be this misleading? What am I missing? I would greatly appreciate it if you are able to respond to me.

Well … I’m not sure what is meant by “literal application of the descriptions in the prospectus – no quotations were supplied. However, I looked at the TD prospectus supplement for 2009-2-27 and found the following language (emphasis added):

“Annual Fixed Dividend Rate” means, for any Subsequent Fixed Rate Period, the rate of interest (expressed as a percentage rate rounded down to the nearest one hundred–thousandth of one percent (with 0.000005% being rounded up)) equal to the Government of Canada Yield on the applicable Fixed Rate Calculation Date plus 4.15%.

Additionally, the front page of the Supplement states (emphasis added):

The Annual Fixed Dividend Rate for the ensuing Subsequent Fixed Rate Period will be determined by the Bank on the Fixed Rate Calculation Date (as defined herein) and will be equal to the sum of the Government of Canada Yield (as defined herein) on the Fixed Rate Calculation Date plus 4.15%. See “Details of the Offering”.

… and …

During each Subsequent Fixed Rate Period, the holders of the Series AI Shares will be entitled to receive fixed quarterly non-cumulative preferential cash dividends, as and when declared by the Board of Directors, subject to the provisions of the Bank Act, payable on the last day of January, April, July and October in each year, in an amount per share per annum determined by multiplying the Annual Fixed Dividend Rate applicable to such Subsequent Fixed Rate Period by $25.00.

Prospectuses may always be found on SEDAR; usually a few days after the new issue announcement.

I do not think that an argument that “plus 4.15% [full stop]” means “plus 4.15% of the 5-Year GOC rate” can be successful. On balance of probabilities – which would be the test in a civil action – if they wanted to say such a thing, they would have said “Annual Fixed Dividend Rate is … 104.15% of the 5-Year GOC rate”, or similar. This would follow examples such as the BCE fixed floaters, which have such language as:

“Selected Percentage Rate” for each Fixed Dividend Rate Period means the rate of interest, expressed as a percentage of the Government of Canada Yield, determined by the board of directors of BCE as set forth in the notice to the holders of the Series R Preferred Shares, which rate of interest shall be not less than 80% of the Government of Canada Yield.

It is also unclear to me as to whether OSFI would allow an issue paying a multiple of an index yield rather than a spread over the index yield to be a Tier 1 issue. It’s an interesting question, though!

It seems pretty clear to me that in the reader’s example, the applicable reset rate is 9.5%. If the prospectus was, in fact, sloppy enough to say “4.5% above…”, then there would in fact be ambiguity: but I suspect that this language comes from press releases and sales material, not directly from the prospectus.

But if anybody can find a counter-example, let’s hear about it! I’ve certainly seen some sloppy language in prospectuses that doesn’t seem to make a lot of difference … until one day, quite suddenly, it does.

What’s going on?

The other day I received an eMail from an Assiduous Reader who wrote in and said:

I follow your blog with great interest. I am a U.S. citizen who actively trades preferred securities, mainly of U.S. banks. I am interested in trading Canadian preferred stocks, but I am not sure of the implications of trading cross-country. Do you know of any resources (articles, online resources, etc) that could inform me of the issues of trading Canadian preferreds as a U.S. citizen?

… and now the thread on the BoC Review, of all things, has gone wild with discussion:

Assiduous Reader GAndreone:

Since I am familiar with the FX market I found the article very interesting. As a FX client if you trade above certain levels, order book and some limited flow information is available from your dealer.

I believe some of the same flow information is being used in the Pref market to set pricing even though volume information is available.
Short interest in pref shares appears to be the only information that is not readily available!

… to which I replied …

I believe some of the same flow information is being used in the Pref market to set pricing even though volume information is available.

Definitely. This will often happen when a trader is working an order … if a client wants to sell 100,000 shares ‘sometime this week’ and there are no takers you’ll often see the market get taken down in stages until the dealer can find a buyer.

Short interest in pref shares appears to be the only information that is not readily available!

Shorting prefs is expensive since the dividend effects are large. Shorts exist, to be sure, but to nowhere near the same extent as in the equity and bond markets.

… and Assiduous Reader cowboylutrell:

Since we’re on the subject of shorting preferred shares, I’d like to know if other brokers are as restrictive as mine on that matter. Since about the beginning of October 2008, it’s been impossible for me to short any preferred share at all. This unavailability of prefs for shorting doesn’t apply only to my own account, but to all customers’ accounts at my broker.

For instance, of the roughly 4,700 stocks listed on the TSX, only 61 were shortable today at my broker’s. None of them are preferreds (they were all common stocks).

As I mentioned above, I’d be curious to find out if shorting preferred shares is still feasable at other brokers in Canada these days.

Many thanks.

… and back to A.R. GAndreone:

Cowboylutrell Says:
As I mentioned above, I’d be curious to find out if shorting preferred shares is still feasable at other brokers in Canada these days.

RBC Direct indicated they would be willing to short prefs if they held shares to short.

jiHymas Says:
Shorting prefs is expensive since the dividend effects are large.

Should the dividend charge only be applicable if you run passed the next ex-date?

And I’ll answer that one here: Yes, the expense (due to tax effects) directly applies only if you’re short over the ex-Date. However, there could be indirect effects if you were to be a forced (or strongly encouraged) buyer in the period immediately preceeding the ex-Date.

Also, I will note that I have been trying to find sponsors for a market-neutral preferred share hedge fund for years. Tax effects won’t apply if the beneficiaries are non-taxable … pension funds, for instance. All inquiries welcomed!

Assiduous Reader Louis writes in and says:

Thank you very much for the very interesting link to Bronte.

On a totally other note, I first thought that the following news concerning NA was a good sign:

“(Marketwire – Feb. 16, 2009) – National Bank of Canada
(TSX:NA) announced today its intention to redeem, for the purpose of
cancellation, all of its 5.70% debentures due April 16, 2014, on Thursday,
April 16, 2009, for 100% of the principal amount of the debentures.
The regular interest due in respect of the debentures on April 16, 2009 will
be paid in the normal course, leaving no accrued and unpaid interest on the
debentures at the time of their redemption.”

But could not then refrain being (again) cynical about it linking this news with the only 15 days earlier news of the closing of NA’s last fixed reset issuance:

“Jan 30, 2009: The Series 26 Preferred Shares will yield 6.60% annually, payable quarterly, for the initial period ending February 15, 2014…Therafter, the dividend rate will reset every five years at a level of 479 basis points over the then 5-year Government of Canada bond yield.”

I fear you are gonna think that I am just not bright enough to understand but, here I am again:

I understand that the above “moves” by the NA may make “capitalisation” sense and that people will say that I am mixing apples with oranges. However, money is money whether you call capital or a debenture and it “moneywise” does not make sense to me issuing prefs paying a non-tax deductible 6.6% dividend while using an equivalent amount of money (for the sake of the discussion here) to buy back debentures now which only cost a tax deductible 5.7% interest which does not mature until 2014.

Bearing in mind that NA was more than meeting its regulatory capital requirements anyway beore its last issuance of resetables, wouldn’t it have been wiser not to issue the costly resetable at this particular time (when they are so expensive to issue due to the prevailing market conditions) and wait a bit longer before redeeming the debentures?

I suspect you are going to say that capital is key since it allows the bank to lend “10″ times that figure while borrowed money by way of the debenture doesn’t have that effect. However, does the NA have so much quality borrowers cuing up at their counters to borrow highly profitable (for the bank) loans? Not that I know.

I also take it that it looks good to redeem debentures at the first opportunity and I recall the DB having been highly criticised for not having done so but its shareholders now appear to accept that it is what had to be done. I submit that the market is weel beyond the “if it looks good, it must be good” stage? We are going through extremely difficult times such that I don’t understand why should our bankers even feel guilty of having to get closer to the minimum required tier 1 capital ratio if adding a cushion doesn’t impress anyone anyway (unless, obviously, more write-offs are to come). We are in a confidence crisis and the only way to fix this is, in my opinion, by stopping playing games.

One should increase his capital cushion when things go well, not in times of troubles when you most need it. Doing the opposite makes longer, costier and more painful the recovery. You cannot fixing the under-capitalisation now, it is when time will get better that I hope we will remember.

Buying back all these resetable in five years or letting them reset is going to come up quite expensive (I doubt very much the 5years Canadas will be anywhere near what it is now). Inflation will be back making high returns expectations even higher.

So, what is wrong with my thinking here?

I don’t think anything is wrong with your thinking here.

You know about the Deutsche Bank situation. It’s possible you’re not drawing a fine enough distinction between Tier 1 Capital (the prefs), which counts towards NA’s Tier 1 Capital Ratio, and between Tier 2 Capital (the sub-debt) which counts towards its Total Capital Ratio, but I’m basically fine with your reasoning.

One thing you did not take explicit account of was the penalty rate that sub-debt usually has, set at issuance to convince the market that they will be called five years before maturity … but according to page 132 of the Annual Report PDF:

Bearing interest at a rate of 5.70% until April 16, 2009, and thereafter at an annual rate equal (1) to the 90-day bankers’ acceptance rate plus 1%

Not much of a penalty in this environment!

All I can suggest to you is that it’s a game of chicken. It makes all kinds of financial sense to let the issue extend past the pretend-maturity (although five years prior to actual maturity, the Tier 2 effect starts amortizing on a straight line basis; after April 19 they would only be able to claim 80% of the face value as Tier 2), but … the market might freak out on them.

NBC unable to redeem debentures! the headlines scream, and all of a sudden their stock price is halved and nobody wants to buy their debt. Banking – as this crisis amply demonstrates – is all about confidence.

The market should not freak out. The market should accept that they will do whatever makes financial sense for them. But there are no guarantees that the market will behave in a rational manner and I can’t find it in my heart much to blame them for not wanting to find out.

Maybe TD or Royal could get away with it…

An Assiduous Reader writes in and says:

A question: How does a retail investor hedge out interest rate risk on perpetual preferreds?

Is there any simple, or reasonably efficient, way to do this?

Well … they can’t, really, which is one reason why I recommend that no more than 50% of a fixed income portfolio be in preferreds.

However, investors should be aware that the Modified Duration of a PerpetualDiscount is simply the inverse of its yield (see http://www.prefblog.com/?p=2582). With yields at about 7%, this means a MD of about 14 years. Many investors will blithely purchase 30-year strips while fussing about the interest rate sensitivity of perps.

I’ve written an essay on Perpetual Misperceptions (see http://www.prefblog.com/?p=1308) … there will be many more!

Perpetuals do have interest rate risk and – more importantly – inflation risk, but I suggest this be addressed in the rest of their portfolios; common stock in resource companies, for instance, or a shorter-than-otherwise-indicated duration in their bond portfolios. At one point – I haven’t done the calculation recently, it’s probably even better now – a taxable investor could swap his Universe iShares into perps and Short-Term iShares on a duration neutral basis and pick up a point in yield. (see http://www.prefblog.com/?p=2399)

A newly Assiduous Reader who is also new to the market writes in and says:

Today, RY.PR.P did not trade AT ALL until after 12:00! Is the market for prefs that illiquid? The price also swung from being below the value of RY.PR.R to ending above the value for R. In my opinion ONLY, the price for R should be higher then the price for P if you are a long term investor. Therefore I question my understanding of pricing in this marketplace!

You and me both, brother, you and me both.

This month’s edition of PrefLetter (currently at the Graphic Artist’s Spa, having its hair done and nails manicured) will contain a section with a new pricing model for Fixed-Resets … so I won’t discuss it here. Instead, I will refer to my last post on Sloppy, Sloppy Markets and take another look at a not-entirely-randomly chosen example of market inefficiency in the Pereptual Discount sector.

This table presents a difficult question to Efficient Market zealots – who implicitly presume infinite liquidity as part of their efficient market. How on earth is it possible to rationalize the quotation on BMO.PR.H?

We’ll review a little … I estimated in my 2007 essay on convexity that being 15% or more away from the call price was worth about 15bp in yield; that is, a PerpetualDiscount trading at around $21.75 should yield about 15bp less than a similar issue from the same issuer trading at par; the higher coupon / higher price issue should yield more since any gains from a decline in yields should be expected to be called away, while the lower priced issue has a higher potential for capital gains.

We can argue for as long as we like regarding details such as:

• 15bp yield difference?
• 15% price range of effect?
• straight line or curved effect?

but there definitely should be an effect and this effect should be positive. In June of 2008 this relationship went negative … while the curve returned to normal after a little while, it certainly resulted in a poor month for the fund I manage. It is these episodes in which the market defies common sense that make leverage such a dangerous game!

Note also that the ModifiedDuration of PerpetualDiscounts (which is a measure of price sensitivity to yield changes) is – to a first approximation – dependent solely upon the yield of the instrument. Any PerpetualDiscount with a given yield has the same yield risk as any other PerpetualDiscount with the same yield, except as distorted by the potential for calls taking away your winnings. So we can’t use yield-sensitivity as an argument.

In sum, I have to advise my newly Assiduous Reader to relax and enjoy the market inefficiency. Once you have a decent model for prices, you can make good money by exploiting transient anomalies and waiting for them to correct. This will increase your turnover and therefore your commission cost (which concerns a lot of people who are inspired by regulatory emphasis on the Trading Expense Ratio), but all moneymaking endeavors have some kind of cost.

Further examples of inefficiency and pricing models for PerpetualDiscounts will be presented at the seminar on February 26. Or, if you don’t want to do it yourself, you can always consider an investment in my fund, which uses many pricing models to check each other and is always on the prowl for anomalies.