Now that the situation regarding RY.PR.O has been clarified, let’s take a look at another constituent of the PerpetualPreferred index with a negative YTW.
POW.PR.A hasn’t been mentioned much in this blog, but made the volume charts on October 11, 2006. At the close of business yesterday, October 18, it was quoted at $25.81-87, pays $1.40, and the next ex-date is (somewhere around) December 20.
The embedded options for this issue are:
- Redemption 2004-06-11 2005-06-10 26.000000
- Redemption 2005-06-11 2006-06-10 25.750000
- Redemption 2006-06-11 2007-06-10 25.500000
- Redemption 2007-06-11 2008-06-10 25.250000
- Redemption 2008-06-11 INFINITE DATE 25.000000
So it is currently redeemable at $25.50.
HIMIPref™ calculates the call probabilities as:
- Call 2006-11-17 YTM: -8.50 % [Restricted: -0.70 %] (Prob: 29.45 %)
- Call 2007-01-16 YTM: 0.69 % [Restricted: 0.17 %] (Prob: 5.99 %)
- Call 2007-07-11 YTM: 2.55 % [Restricted: 1.85 %] (Prob: 9.69 %)
- Call 2008-07-11 YTM: 3.70 % [Restricted: 3.70 %] (Prob: 5.09 %)
- Option Certainty 2036-09-12 YTM: 5.42 % [Restricted: 5.42 %] (Prob: 49.77 %)
Note that all these scenarios are combined at their probabilities to derive portfolioYield, but the worst result is used as Yield-To-Worst.
The various yields calculated for this issue are:
|Yields calculated for POW.PR.A|
|The normalization factor for the yield weightings is 0.245, resulting in a sum of yield components of valuation of -2.6618.|
The sum of yield components of valuation of -2.6618 is the lowest value in the perpetualPremium index, the average value for the this calculation for this index is just a hair over 4.00.
Some may wonder at the calculations done here – why not just use YTW? As it turns out, in some analytical environments the other measures of yield are more discriminatory, but there is some variation that may be observed even in the current environment. In the graph, the following data points have been removed in order to increase the resolution of the display: AL.PR.E, TOC.PR.B, AL.PR.F. The graph is here, the regression calculation here.
HIMIPref™ will not recommend this issue for purchase to clients, giving the Eligible For Purchase (Code) as ’14’, which a quick look at the glossary defines as ‘pseudoModifiedDuration (Worst) of buy side less than minimum setting’. In other words, there’s a very good chance (YTW!) the issue will be called in the near future. Not only does HIMIPref™ have trouble discriminating between issues with such small pseudoModifiedDurations, but even if it could do this well then there wouldn’t be much money in it (since a small price change for a short-term instrument can result in a very large change of yield.
Bottom line: POW.PR.A looks overpriced, to the extent it can be analyzed. There are many alternatives available which can be analyzed effectively and quantitatively … so why go to huge extremes to justify holding it? Research is continuing, as ever, to extend the reach of measurement in which the HIMIPref™ analysis can result in superior performance, but this range of accuracy will not be extended at the expense of confidence in what HIMIPref already does well.