The February 5 Financial Post had an article titled Investors Prefer Preferreds, a short piece about the popularity of Fixed Resets. I was able to offer one reason why they are popular with investors:
“Fear levels have ratcheted up,” said James Hymas, president and portfolio manager at Hymas Investment Management. “They are flocking to this because the new structure is giving them some degree of comfort.”
… and one reason why the banks like them …
“One reason the banks like this new structure is because they have a call at par in five years,” Mr. Hymas said.
Well, I can tell you one thing – it’s getting to be a very strange market! There are issues trading at a discount for which the YTW scenario is a call at par as soon as possible – because the reset makes the perpetual yield higher. And other issues which have a “five year yield” (‘yield to next call’ would be a more precise way of expressing the scenario) so high that investors are either (a) trading them as perpetuals, or (b) stupid.
It makes it very difficult to fit the data into a unified theory!
This is a timely blog post for me, as I have been playing the last week with modifying my preferred spreadsheet to be able to analyze fixed resets.
They are very difficult to get a handle on. I have been comparing an older reset and a recent reset in TD.PR.S (5.00%+160) and TD.PR.G (6.25%+438) to see how they stand up against each other and a comparable straight TD.PR.R.
I had considered your weighted (5-10% YTP and 90-95% YT5), but ended up with a simple progressive list of six YTR (yield-to-reset), where my single variable was the GoC 5 year yield assumption at the first reset and then carried through as a constant. Modifying the GoC rate after the first reset seemed folly to try and predict.
It is as you say (I think), the low coupon, low reset rate TD.PR.S that is selling at a fair discount, offers a very generous first YTR as a result of the capital gain component.
If rates remain at the present 2.0% GoC rate, it seems unlikely that it would be called since the new coupon would be 3.6% (200+160). This would certainly cause the market price to drop quite a bit. The holder wouldn’t be too happy.
But if GoC rates rose to 5.0% as an example, then the new reset coupon would be 6.6% (500+160). Perhaps at that point it would be a benefit for the company to call the issue, offering the holder a very nice YTR of about 8.3% if purchased at todays $22.
So, buyers of this type of reset would appear to be making a heavy bet that rates will rise. Or, are my numbers crazy? I’m sure you’ll set me straight.
Also as you say, the high coupon, high reset rate TD.PR.G that is selling slightly below PAR, results in all the YTR’s to be above the current yield.
Even if rates remain at the present 2.0% GoC rate, it seems possible that the issue would be called since the new coupon would be 6.38% (200+438). I suppose that’s not a bad five year return if called.
If the GoC rate rose even a small amount, it seems probable that the issue would be called, because the reset coupon would not be in the companies interest.
So, buyers of this type of reset appear to be using this as a pure five year bet at its current yield.
Compared against the straight TD.PR.R which has a current yield of 6.67% perpetually, it appears that the deciding factor when deciding between it and TD.PR G is whether five year GoC rates will rise.
I continue to remember your original article on resets where you said, “when I lend money on a perpetual basis, I want to get perpetual rates of return”. Given the newer resets, I wonder if they’re not a better deal now?
The resets have really made investing long in preferreds a more difficult task to analyze. I know that you’re presently working on the bond series in MoneySaver, but a good article for the future might be something on analyzing fixed resets in the present environment. They make straights appear like child’s play to analyze.
ltr
They are very difficult to get a handle on.
Hoo boy, you can say that again!
I had considered your weighted (5-10% YTP and 90-95% YT5),
You have quoted me backwards! I said 5-10% YT5 and 90-95% YTP! These figures will vary – it is not a measure of what I think the market should be, it is was simply a measure of where the market is. At the time I wrote the article, that was where it was … that is no longer where it is.
One problem with that analysis is that the market is now bimodal after the massive January issuance … you have the “old ones” and the “new ones”, not a smooth distribution. This calls into question the theoretical soundness of using Standard Deviation as a measure of success in blending.
Given the newer resets, I wonder if they’re not a better deal now?
They certainly are! While I continue to feel that as a class they are rather expensive, there is now enough heterogeniety and enough average yield that every now and then one looks cheap. You will note that the fund had a position on Jan. 30.
a good article for the future might be something on analyzing fixed resets in the present environment.
This will indeed be a future article and it will also be a future seminar.
Oops sorry, I did mistakenly write the mis-quote, but I meant it the correct way.
My point was that when I used your (5-10% YT5 and 90-95% YTP) it seemed to put so much weight on the long term assumption that rates would remain at todays reset rate, it didn’t give me a feel of the breakout of 5,10,15, etc. years to look at old and new issues.
Anyway, thanks for the comments.
ltr
Thks to both of you as it helps me in my reflexion. I do verily fear that the balance will swing back with high inflation by the time the first generation of reset will become redeemable. With the resulting BOC’s resulting quick increase, straight perps will be hit hardly. My current feeling is, therefore, that straight perp should recover furhter within the next 18 months at which time swaping those for fixed resets might be the course of action to take. It all depends of their market price at the time but, if I may ask James, I woulld be curious if he could just illustrate at a given date (say Friday prices) which of the resets he would put in each of his illustrative categories:
1. The “issues trading at a discount for which the YTW scenario is a call at par as soon as possible – because the reset makes the perpetual yield higher”; and
2. “the other issues which have a “five year yield” so high that investors are either (a) trading them as perpetuals, or (b) stupid”.
I assume that by 1) James is referring to the most recent issues and by 2) to the oldest ones. However, I am curious to know where the line is drawn between the two categories as I suspect that those issues which are the closest to the line (say possible the November / early December issues) might be where the best compromise will be when the market correct itself a bit further striking a fair balance between the two scenarios: If the are called in 5 years I will at least do a small capital gain while having done a decent return in the meantime; if they are resetted I will at least keep on making a decent return while have their market value remaining fairly constant.
Sorry, Louis, but details will have to await
I regret this, but I have to have something for sale!
Hi James,
Great Blog! What is the normal response time in market price of existing pref shares to new pref share issues? It appears to take over a week. Is this possible?
Fred
Thanks for the kind words, and thanks for becoming an Assiduous Reader!
During the last round, the adjustment took about two minutes.
It should be noted that the current situation, in which the market price of extant issues adjusts to the new issue, is abnormal. New issues are normally priced with a selling concession, in order to give potential buyers a reason to put fresh money to work; once this has been accomplished, the new issue price should adjust to market, rewarding the buyers with a capital gain equal to the selling concession.
During this crunch, however, deals need to be priced with a massive selling concession in order to get done and the market (worth, say, $20-billion) has adjusted to the new price of a $200-million issue.
I would be most interested in learning your methodology for estimating a week-long adjustment period. There’s always something new to learn!
James, thank you for the response. The approach used was not very scientific. The specifics are
Pref1)
RY.PR.P has a LONG term yield of +419bp and a first dividend 0.55651
Pref2)
RY.PR.R has a LONG term yield of +450bp and a first dividend 0.499229
The majority of other conditions are similar.
Condition for price adjustment.
When does the price in the market approximately reflect the the difference in the first dividend AND when does the volume of trading in Pref1 reduce significantly and Pref2 volume remains greater then Pref1
Since I am new to pref shares I hope I did not miss anything in the two prospectii!!
Thanks! This analysis is certainly reasonable, but I would prefer to refer to a basket of extant comparables than a single issue. The preferred share market can be very, very sloppy, which has the potential for skewing the results of single-issue comparisons.