RioCan Real Estate Investment Trust has announced:
that it will exercise its right to redeem all of its 5 million outstanding Cumulative Rate Reset Preferred Trust Units, Series A (the “Series A Units”) on March 31, 2016 at the cash redemption price of $25.00 per Series A Unit, for total redemption proceeds of $125 million.
The regular quarterly distribution will be paid in the usual manner on March 31, 2016 to unitholders of record on March 31, 2016.
From and after March 31, 2016, the Series A Units will cease to be entitled to distributions and the only remaining rights of holders of such units will be to receive payment of the cash redemption price.
Beneficial holders who are not directly the registered holder of Series A Units should contact the financial institution, broker or other intermediary through which they hold these units to confirm how they will receive their redemption proceeds. Instructions with respect to receipt of the redemption amount will be set out in the redemption notice to be mailed to the registered holder of the Series A Units shortly. Inquiries should be directed to our Registrar and Transfer Agent, CST Trust Company, at 1-800-387-0825 (or in Toronto 416-682-3860).
REI.PR.A is a FixedReset, 5.25%+262, which commenced trading 2011-1-26 after being announced 2011-1-17.
This redemption is really, really weird. A spread over Canadas of +262bp is not really considered all that much nowadays, not for a junk-rated company when investment-grade issuers are paying close to +500 for new money. I will also point out the following from their 15Q3 Report:
As at September 30, 2015, the weighted average contractual interest rate of RioCan’s debt portfolio is 3.87% (4.12% as at December 31, 2014), a decrease of 27 basis points from the weighted average contractual rate of 4.14% as at September 30, 2014.
So extending this issue would have continued to decrease their average funding cost and at the same time would be permanent capital. Against that, the estimated reset rate of 3.20% is probably a little more than they’re paying at the moment, there is – by definition – recourse to the company, and the rate will be reset in five years just like a regular mortgage. Referring to the 15Q3 Report again for the word “recourse”, we find:
As at September 30, 2015, the Trust’s mortgages payable and drawn lines of credit, was $4.7 billion ($4.6 billion as at December 31, 2014). The vast majority of the Trust’s Canadian mortgage indebtedness provides recourse to the assets of the Trust, as opposed to only having recourse to the specific property charged. RioCan follows this policy as it generally results in lower interest costs than would otherwise be obtained. In the United States, mortgage debt is generally non-recourse financing, with no U.S. secured debt having recourse to the assets of the Canadian operations of the Trust.
We also look back to their 11Q1 Report to see what they had to say about their issue of preferred units:
RioCan was the first Canadian real estate investment trust to issue preferred units, indicative of the Trust moving closer to its objective of becoming “best in class” from a capital markets perspective. The ability to issue preferred units allows RioCan greater flexibility in accessing capital markets and developing a desired capital structure.
RioCan is relatively unscathed by the Sears withdrawal:
Still, RioCan Real Estate Investment Trust, which was Target’s largest landlord, has so far been unscathed by Sears’s latest move to exit more stores. While Sears is leaving a home store in a RioCan-owned mall in British Columbia, the retailer has a deal to sublet the space to Leon’s Furniture Ltd., said Edward Sonshine, CEO of RioCan, which has five other Sears home stores and two of its full-line stores.
Even so, RioCan is already looking for an alternative tenant for one of its other Sears home stores because its lease expires in about a year and “we assume they won’t be renewing,” he said.
However, they recently sold their US portfolio:
RioCan Real Estate Investment Trust is ending its six-year foray into the U.S. with a deal to sell its 49 shopping centers in the country to Blackstone Group LP for $1.9 billion.
The sale to the Blackstone Real Estate Partners VIII fund will provide capital for RioCan’s recently announced acquisition of 23 properties from Kimco Realty Corp. and to cut debt, Canada’s largest retail landlord said in a statement Friday. The U.S. shopping centers are located in the Northeast and Texas.
…
RioCan entered the U.S. market following the financial crisis, purchasing grocery-anchored retail sites at a discount. In July, when the company announced its strategic review of the properties, Sonshine said a weakening Canadian dollar made it costly to expand in the U.S. and RioCan was looking to get more value from the assets.
The total price RioCan paid for the 49 retail sites was C$1.7 billion. The sale at C$2.7 billion, 59 percent more, will provide an internal rate of return of about 16 percent, according to the statement. The Canadian dollar has dropped about 25 percent in the past six years to about 71 cents per U.S. dollar, giving RioCan a sizable currency gain on its investment. The sale is expected to be completed on April 30.
So, I can’t really figure this one out. The best I can come up with is the idea that they don’t really see many new investment opportunities in Canada at the moment and assume that those that do come up can be financed with good old Canadian non-recourse, cut-rate mortgages.
But the price action on the day for the issue was pretty interesting:
Click for BigYou don’t see intra-day changes of 50%+ very often!
REI.PR.C was also up on the day, +12.8% to close at 21.15, presumably on speculation that the same thing will happen to it when its Exchange Date comes on 2017-6-30. It resets at +318, so it’s easy to follow the reasoning!
Update, 2016-2-11: Barry Critchley of the Financial Post has written a piece titled Behind RioCan’s decision to redeem its first-of-its-kind rate reset preferreds. In addition to the sale of the US portfolio noted above, there is the fact that the ratings agencies are not giving any equity credit to preferreds issued by REITS and:
2.The refinancing rate of 262 basis points is not that attractive. “Both of those spreads [on the new fixed and floating rate preferreds] are higher than spreads we can borrow at,” she said. Aside from its operating lines [which are priced at “125 basis points over” the U.S. or Canadian prime rate] RioCan can borrow five-year money on an unsecured basis about 50 basis points lower than the 262 basis points it would be required to pay if it extended the maturing preferreds.
Well, sure, but it’s still not all that convincing. The preferreds are not five-year money; they’re permanent money priced at a spread over five-years. So Cynthia Devine, RioCan’s chief financial officer, is saying she’s not willing to pay a 50bp term premium for this money, and:
“We will have a lot of proceeds and we have made it clear that one of the primary uses is to pay down debt. This is easy to execute because it’s coming due in March,” she said.
… which doesn’t sound like a very good reason to me. At the very least, I would have been sorely tempted to have extended the issue and put in a big Normal Course Issuer Bid; perhaps preceded by a tender offer; or even with a distribution of ‘put rights’ (by which I mean rights distributed to common shareholders that would allow X rights and 1 share of REI.PR.A to be sold to the company for Y dollars).
Ms. Devine may have had very good reasons for doing what she did. But she left a lot of money on the table and that’s not what CFOs are supposed to do.
Update, 2016-2-22: Barry Critchley has written a follow-up piece titled Two tales of preferred redemption, Rona and RioCan REIT, in which Cynthia Devine offers up a new rationale for her decision to redeem:
On Friday, Devine said RioCan “did look at that [a tender offer below $25] but there was a chance that not all of them would be taken out of circulation. Some [investors] may not tender so you have a series outstanding.”
Well, I don’t want to get too vituperative here, because this is only a quote in the press … there may be considerably more nuance and rationale behind this statement than was published.
But the excuse doesn’t cut any ice. The shares closed at $16.00 the day before the announcement, and there are 5-million of them outstanding; so Ms. Devine’s decision awarded windfall gains of $45-million to the holders of the preferred shares.
So the first question is: how did the cost-benefit work? A tender at $20 would – I feel quite certain – have captured all but a handful of the shares, given the preferred shareholders a windfall gain of $20-million and given the ordinary unitholders a gain of $25-million. So, maybe the series would have remained outstanding. What’s the cost of that, compared to $25-million in hand?
And the second question is: if it was deemed absolutely necessary to eliminate the series completely (which requires rather a leap of the imagination), why wasn’t another method tried? A Plan of Arrangement would serve the purpose nicely – just get the preferred shareholders to vote on whether to eliminate the series at $20 / share. Dissenters might get right of dissent, but it is very difficult to see any judge awarding them more than $20, when the prior market value was $16.
The fact of the matter is that any rational holder of the preferred shares would be very happy to have gotten $20 (which could be redeployed into comparables) and this would have given the ordinary unitholders a quick gain of $25-million. This is not an insignificant number: RioCan’s profit for all of 2015 was $142-million. Ms. Devine had the opportunity to realize one-sixth of that, while leaving the preferred shareholders very happy … and chose not to do so.
