Bank of Montreal has announced:
that it has entered into an agreement to privately place its Non-Cumulative 5-Year Rate Reset Class B Preferred Shares Series 36 (Non-Viability Contingent Capital (NVCC)) (the “Preferred Shares Series 36”). BMO Capital Markets is acting as the sole agent on the transaction. Bank of Montreal will issue 600,000 Preferred Shares Series 36 at a price of $1,000 per share to raise gross proceeds of $600 million. The closing of the offering is scheduled to occur on October 16, 2015, subject to the satisfaction of certain closing conditions. The net proceeds will be used by the Bank for general corporate purposes.
Holders will be entitled to receive non-cumulative preferential fixed quarterly dividends as and when declared by the board of directors of the Bank, payable in the amount of $14.625 per share, to yield 5.85 per cent annually. Subject to regulatory approval, on or after November 25, 2020, the Bank may redeem the Preferred Shares in whole or in part for an amount equal to $1,000 per Preferred Share Series 36 together with declared and unpaid dividends to the date fixed for redemption.
So this is a strange one on a great many levels and we probably won’t really be able to understand this issue until we get the BMO 2015 Annual Report – and perhaps not even then!
So first off, it’s a private placement. The only private placement of preferred shares – from an investment-grade, major public issuer – that I can recall is BNS.PR.S, FixedReset, 6.25%+384, which closed December 12, 2008 after having been announced December 3, 2008; $250-million issued to Sun Life Financial as part payment for CI Financial Income Fund. In that case, details were later available on SEDAR, but there was a difference in that BNS.PR.S was the private placement of a public issue, while there is nothing in the current announcement to indicate that this issue will be – at least theoretically – tradable by the public. At $1,000 per share par value, my guess is “no”!
Second, there’s the size of the thing. $600-million, done without a whisper, so one can well presume that it was to an individual client or, possibly, a small consortium of clients. Now my question is: who’s got that kind of money? The size represents roughly 1% of the entire Canadian preferred share market; while there are a fee issues that are in that ballpark (RY.PR.J, $600-million; FTS.PR.M, $600-million; TRP.PR.D, $600-million; TRP.PR.A / TRP.PR.F weighs in at $550-million; BMO.PR.S, $500-million; RY.PR.H, $500-million; ) they were all issued at a time when the preferred share market was, shall we say, a little more robust than it is now.
Who’s got that kind of money? I suggest that there are two logical places to look for people who can throw down amounts like this: pension funds and foreigners. But the problem is … pension funds and foreigners don’t get the benefit of the Dividend Tax Credit and Gross-up (although foreigners could do it through a Canadian subsidiary). So why would they care about preferred share dividends. Which leads us to the next question …
Thirdly, does it pay dividends or interest? On the one hand the word “dividends” is used twice in the press release; on the other hand, so what? I don’t think anybody will go to jail if they refer to the payments as dividends in a press release, but then call it interest when preparing the tax slips – of course, I could be wrong on that! But 5.85% is a whacking great huge rate for a dividend; it will be recalled that BAM did a recent issue at 5% after CU did one at 4.50%. BMO is still Pfd-2 by DBRS although only P-3(high) from S&P. Do they really need to pay 5.85%? Are they really that short of capital?
I suspect they aren’t; and I note that when you divide 5.85% by the standard equivalency factor of 1.3, you get 4.5% (exactly!) which is at least in the ballpark of where they would be willing to do a public issue (whether they actually could do it in size in the present environment is another question!). So, from two perspectives (three, if you include the $1,000 par value) it makes sense that this 5.85% is an interest rate, not a dividend rate; but whether or not this is true will have to await confirmation.
And fourthly, what’s the Issue Reset Spread? We are told that this issue represents “(Non-Viability Contingent Capital (NVCC))” which suggests that OSFI has blessed the issue and OSFI won’t (quite rightly) allow step-ups, so the spread won’t be much more than +500bp over five-year Canadas; but it could, conceivably, be less. Another mystery! And we’re not even sure if the touted “Rate Reset” bears any relation to the standard terms of public FixedReset issues. The underlying rate could be just about anything and the reset frequency is equally obscure.
Hat tip to Assiduous Readers JB, GB and LM, who ensured I was informed of this issue!
I am a US investor in Canadian rate-reset preferreds. I believe that for US investors CA preferred dividends are treated as qualified dividends, at least that is how my broker (Interactive Brokers) has always reported them to the IRS. Qualified dividends *are* tax advantaged for me in that they are taxed at significantly lower rate than non-qualified corporate bond dividends. Furthermore, dividends from Canadian preferreds in US retirement accounts do not suffer CA withholding tax due to a tax treaty between the US and CA.
CA floaters and rate-reset preferreds now have a *much* higher yield, over the long run, than similarly structured/rated instruments traded on US markets. That is not the case for CA fixed preferreds.
Qualified dividends *are* tax advantaged for me in that they are taxed at significantly lower rate than non-qualified corporate bond dividends.
I hadn’t known that … but I’ve just read up on Qualified Dividends on Wikipedia. Very interesting, thank you.
CA floaters and rate-reset preferreds now have a *much* higher yield, over the long run, than similarly structured/rated instruments traded on US markets. That is not the case for CA fixed preferreds.
That’s also very interesting! Can you give any examples?
Here are two examples. Libor3Mo is currently 0.32%.
UBS-D (Perpetual, non-cumulative)
Libor3Mo + 0.70%.
At current market price of $18.34 YTM is a paltry 1.39%.
Rated BB by Moodys.
GS-J (Perpetual, non-cumulative)
Fixed at 5.50% until 5/10/23 and then floats at Libor3Mo + 3.64%.
Current market price of $24.95.
Rated BB by Moodys.
My database tracks > 100 US floaters and fixed-to-float preferreds and many such corporate bonds too. There was a market panic in these preferreds during the “Great Recession”. There was another market panic in early 2012 when it looked like the Euro might crumble. There has been no “interest rates are going to be 0% forever” market panic in the US this year. Barely a light sweat.
Not much *again* on the details but some NVCC discussion…
http://www.theglobeandmail.com/globe-investor/investment-ideas/how-will-canadas-banks-boost-capital-ratios/article27518651/
As discussed in the post, the 5.85%+500(?) looks a lot more like an interest rate than an eligible dividend rate, but there is still no definitive description I can find. We will just have to hope that disclosure in the Annual Report is adequate.
I haven’t found any detail as to whom the prefs were issued to (on a cursory overview of their financials), but BMO has announced retirement of $450MM in Tier 1 securities (series e boats) – as prefs qualify for Tier 1 capital it is likely that this was the purpose. As suggested here before, the high rate to was probably to facilitate such a sizeable issue (to likely a tax advantaged party, so probably interest, not dividend) and possibly to compensate for NVCC fixed-resets not being allowed to put in a floor (was suggested on another blog). It sounds like these were placed to 1 buyer and no one who tried was able to get any of the issue.
I’ve read recently that RBC has $1200MM of tier 1 capital securities to retire in the near future – wonder if there’s another major pref issuance from them coming soon.
I concur – I can’t find a definitive statement regarding the tax identity of the distributions either. Unless the actual Annual Report has something about it [as opposed to the bits that have been released to date] I guess our next opportunity will be with the 16Q1 report, which should break out the dividend payments in the income statement.
and possibly to compensate for NVCC fixed-resets not being allowed to put in a floor (was suggested on another blog).
I discussed floors for NVCC issues when the first minimum-reset issue came out, in the post New Issue: CU FixedReset, 4.50%+369M450. I concluded there was no problem – do you believe there is a problem? What was the other blog?
Hi James,
The floor comment isn’t based on my own research – it was from – http://www.nexgenfinancial.ca/blog/finally-some-relief-for-preferred-share-investors/
I don’t think it really matters whether NVCC issues can set floors or not – we just have to note which issues do so we can adjust our models and call probabilities.
So Jeff Herold at Nexgen says:
So he’s considering the headline 5.85% to be taxable as a dividend; which I suggest is way more than 1.00% too high. And he says that OSFI is “not yet” permitting banks to issue rate reset issues with floor dividend rates; it may well be that the lack of bank issuance with this feature is because it is hung up in the review process, but I have no information about that one way or the other.
I don’t think it really matters whether NVCC issues can set floors or not – we just have to note which issues do so we can adjust our models and call probabilities.
I’ll suggest that given current yields, the floor rate is basically meaningless and that the feature is, like cumulativity, more of a marketing tool than an investment attribute. It does not seem too likely to me that GOC-5 will decline substantially from here – but critics will gleefully point out that I’ve said that before!