If you read any of Mr. Hymas' writings (and/or subscribe to his newsletters), you'll see that he grosses up the dividends received in his analysis by a factor of 1.3x to get back to a pre-tax interest equivalent dividend (a second important factor to consider is the after tax equivalent of holding preferreds in non-registered accounts for Canadians as the dividends qualify for the dividend tax credit).
When I initially did my analysis, I used only the stated dividend yield without considering the 1.3x gross up. I've redone the analysis for IFC.PR.A using a gross up of 1.3x vs. no gross up (on the assumption that participants other than Canadians hold them, or on the assumption that some Canadian participants hold them in registered accounts and can't make use of the DTC as a result).
I've also done the analysis excluding the impact of deemed retraction in 2025 as Mr. Hymas has postulated just to see what the market is theoretically pricing in here at Friday's close on a worst case basis. For my discount rate, I've interpolated in order to solve for this based on Friday's close. The 2018 PV is calculated as straight perpetual dividends = 2018 div / R / (1+R) ^ 2 ).
Here are the results:
And here are the results interpolating the discount rate for the possibility of deemed retraction in 2025:
Which really doesn't make an ounce of sense, as I've interpolated a higher discount rate with a better result in 2025 (i.e., redemption)? <January 18th update, this actually may make some sense now that I think about it, I've solved for YTM buying at current market, maturing in 2025, and the reason YTM is 12% is due to the market not pricing in redemption. If it did, YTM would approximate YTM on comparable bank preferred's.>
For comparisons sake, here is the same exercise using BNS.PR.Q fixed resets (at almost the same reset spread) with a deemed maturity in 2022 (note that in all of these cases, I've modelled next reset at GC 5 yr = 0):
You can immediately see the difference in market price between BNS.PR.Q which is slated for redemption in 2022 and IFC.PR.A which is not. The difference is about $7 in differential. Part of this difference must be attributable to optionality, i.e., BNS.PR.Q is +$7 vs. IFC.PR.A because the market recognizes and is pricing in the redemption feature in BNS.PR.Q and no such feature is priced into IFC.PR.A (yet), and the remaining difference must be due to duration.
Here's IFC.PR.A modelled at BNS.PR.Q interpolated discount rates + deemed retraction:
I have no idea what may happen from here, and I'm certainly not enjoying watching the daily demolition of these issues. I suppose my saving grace is that these issues are an overall small %ge of total assets and I can only observe the implosion with a sense of humour (if this is at all possible), while investors flock to obscure ETF's like BMO US Put Write ETF under the guise of "safety". Situation normal.
I can say one thing, should OFSI issue a ruling sooner rather than later (which was supposed to occur in early 2016), the market will most likely adjust violently to price the missing optionality in. In the meantime, it's batten down the hatches time.