July 28, 2006

Index Current Yield (at bid) YTW Average Trading Value Mod Dur (YTW) Issues Day’s Perf. Index Value
Ratchet 4.21% 4.22% 92,290 17.01 2 0.3822% 997.9
Fixed-Floater 5.22 3.92% 99,784 5.17 5 0.1523% 1,004.8
Floater 4.54% -15.02% 59,223 6.43 5 0.1680% 1,003.4
Op. Retract 4.74% 3.19% 78,823 2.76 18 0.0154% 995.6
Split-Share 5.17% 3.93% 47,897 2.85 14 0.1909% 1,000.2
Interest Bearing 6.86% 5.15% 66,114 2.19 7 -0.1801% 1,009.4
Perpetual-Premium 5.32% 4.33% 124,166 3.86 40 0.0947% 1,005.8
Perpetual-Discount 4.77% 4.89% 388,850 15.84 14 0.0881% 1,005.8
Major Price Changes
Issue Index Change Notes
MST.PR.A InterestBearing -1.43% 3,832 shares traded. It was up 1.55% yesterday on 6,088 shares … easy come, easy go.
TDS.PR.B SplitShare +1.91% 900 shares traded. Somebody didn’t want many shares, but they REALLY wanted them. This issue is callable at 28.10 in November and is now quoted at 28.50-65. YTW is -0.03%
Volume Highlights
Issue Index Volume Notes
CM.PR.C PerpetualPremium 204,166 BMO crossed 200,000 @26.84 for delayed delivery.
SLF.PR.B PerpetualPremium 105,942 Gained 0.48% on day.
GWO.PR.G PerpetualPremium 35,570 Scotia bought 20,000 from CIBC @26.00
RY.PR.B PerpetualDiscount 14,000 Is the “Inventory Reduction Sale” over?
MFC.PR.C PerpetualDiscount 12,600  
SLF.PR.C PerpetualDiscount 10,900  

There were no other index-included issues with volume in excess of 10,000 shares.

My more observant readers will have noticed that the “Perpetual” index has been split into “Premium” and “Discount”. With a bi-modal distribution of index characteristics, the averages didn’t make a lot of sense. I don’t really know if this will be a permanent change – I’ll have to see what happens over time, when I get around to writing and testing automated index preparation software. If there aren’t any discount perpetuals in my database, for instance, I’m not sure what I’ll do or how meaningfully I’ll be able to present the results. Stay tuned!

4 Responses to “July 28, 2006”

  1. Drew says:

    I presume the difference between premium and discount perpetuals lies in the relationship of the first redemption price to the present trading price. Is this correct? In any event, how in theory would you expect each category to perform in a rising, falling and flat interest rate environment – similar to premium and discount bonds?

  2. jiHymas says:

    That’s worth a long article in itself! In my essay How Long is Forever? I make the point that a high-premium perpetual can be very attractive relative to a retractible, since rates will have to increase substantially before a call becomes unlikely.

    Premium issues should out-perform discount issues in a rising rate environment, because a lot of the loss (between market and call prices) has already been ‘lost’ in the calculation of YTW. Also, the simple fact that their dividend rate is higher means, even after they have become discount issues, that a similar change in market yield will have a lesser price effect.

    In a falling rate environment, the opposite is true: the market price of a discount issue can rise to par ‘for free’, whereas the price change of a premium issue will be constrained from the start by the fact that a redemption is likely on a particular call date.

  3. Drew says:

    Why you are not yet convinced that separating the index as you have makes sense? Further, why not also separate the index into long and short-dated perpetuals based on first call date or some combination of first call date and likelihood of call?

  4. jiHymas says:

    The premium instruments have a much shorter modified duration (of the YTW scenario) than do the discount instruments.

    The first call date certainly could be used as a method to partition the perpetuals, but the premium/discount partition is much more important. And I don’t want to get too many sub-indices going, or the value of the indices will be lost in a chaos of detail.

    Likelihood of call is a somewhat subjective measure, and it’s the reason why the yield measures “Portfolio Yield”, “Cost Yield” and “Curve Yield” were developed. For public (which is to say, free) purposes, I’d prefer to use an objective measure.

Leave a Reply

You must be logged in to post a comment.