Algonquin Power & Utilities Corp. has announced:

the applicable dividend rates for its Cumulative Rate Reset Preferred Shares, Series D (the “Series D Preferred Shares”) and Cumulative Floating Rate Preferred Shares, Series E (the “Series E Preferred Shares”).With respect to any Series D Preferred Shares that remain outstanding after April 1, 2019, holders thereof will be entitled to receive quarterly fixed cumulative preferential cash dividends, if, as and when declared by the board of directors of the Company (the “Board”). The dividend rate for the 5-year period from and including March 31, 2019 to but excluding March 31, 2024 will be 5.091%, being equal to the 5-year Government of Canada bond yield determined as of today plus 3.28%, in accordance with the terms of the Series D Preferred Shares.

With respect to any Series E Preferred Shares that may be issued on April 1, 2019, holders thereof will be entitled to receive quarterly floating rate cumulative preferential cash dividends, if, as and when declared by the Board. The dividend rate for the 3-month floating rate period from and including March 31, 2019 to but excluding June 30, 2019 will be 4.962%, being equal to the 3-month Government of Canada Treasury Bill yield determined as of today plus 3.28%, calculated on the basis of the actual number of days in such quarterly period divided by 365, in accordance with the terms of the Series E Preferred Shares.

Beneficial owners of Series D Preferred Shares who wish to exercise their conversion right should communicate with their broker or other nominee to ensure their instructions are followed so that the registered holder of the Series D Preferred Shares can meet the deadline to exercise such conversion right, which is 5:00 p.m. (EST) on March 15, 2019.

AQN.PR.D is a FixedReset, 5.00%+328, that commenced trading 2014-3-5 after being announced 2014-2-24. The extension was announced 2019-2-26. The issue is tracked by HIMIPref™ but relegated to the Scraps-FixedReset (Discount) subindex on credit concerns.

The most logical way to analyze the question of whether or not to convert is through the theory of Preferred Pairs, for which a calculator is available. Briefly, a Strong Pair is defined as a pair of securities that can be interconverted in the future (e.g., AQN.PR.D and the FloatingReset that will exist if enough holders convert). Since they will be interconvertible on this future date, it may be assumed that they will be priced identically on this date (if they aren’t then holders will simply convert en masse to the higher-priced issue). And since they will be priced identically on a given date in the future, any current difference in price must be offset by expectations of an equal and opposite value of dividends to be received in the interim. And since the dividend rate on one element of the pair is both fixed and known, the implied average rate of the other, floating rate, instrument can be determined. Finally, we say, we may compare these average rates and take a view regarding the actual future course of that rate relative to the implied rate, which will provide us with guidance on which element of the pair is likely to outperform the other until the next interconversion date, at which time the process will be repeated.

We can show the break-even rates for each FixedReset / FloatingReset Strong Pair graphically by plotting the implied average 3-month bill rate against the next Exchange Date (which is the date to which the average will be calculated).

The market has lost its fleeting enthusiasm for floating rate product; the implied rates until the next interconversion are below the current 3-month bill rate as the averages for investment-grade and junk issues are at +1.12% and +1.45%, respectively. Whatever might be the result of the next few Bank of Canada overnight rate decisions, I suggest that it is unlikely that the average rate over the next five years will be lower than current – but if you disagree, of course, you may interpret the data any way you like.

Since credit quality of each element of the pair is equal to the other element, it should not make any difference whether the pair examined is investment-grade or junk, although we might expect greater variation of implied rates between junk issues on grounds of lower liquidity, and this is just what we see.

If we plug in the current bid price of the AQN.PR.D FixedReset, we may construct the following table showing consistent prices for its soon-may-be-issued FloatingReset counterpart given a variety of Implied Breakeven yields consistent with issues currently trading:

Estimate of FloatingReset (received in exchange for AQN.PR.D) Trading Price In Current Conditions |
|||||

Assumed FloatingReset Price if Implied Bill is equal to |
|||||

FixedReset | Bid Price | Spread | 2.00% | 1.50% | 1.00% |

AQN.PR.D | 20.83 | 328bp | 21.01 | 20.53 | 20.04 |

Based on current market conditions, I suggest that the FloatingResets that will result from conversion are likely to trade below the price of their FixedReset counterparts, AQN.PR.D. Therefore, it seems likely that I will recommend that holders of AQN.PR.D continue to hold the issue and not to convert, but I will wait until it’s closer to the March 15 notification deadline before making a final pronouncement. I will note that once the FloatingResets commence trading (if, in fact, they do) it may be a good trade to swap the FixedReset for the FloatingReset in the market once both elements of each pair are trading and you can – presumably, according to this analysis – do it with a reasonably good take-out in price, rather than doing it through the company on a 1:1 basis. But that, of course, will depend on the prices at that time and your forecast for the path of policy rates over the next five years. There are no guarantees – my recommendation is based on the assumption that current market conditions with respect to the pairs will continue until the FloatingResets commence trading and that the relative pricing of the two new pairs will reflect these conditions.

I am an economist but don’t work in the investment industry. I am new to Pref investing and am surprised by what seems to be irrationality in the market (I am probably missing something). This pref. is a good example. AQN.PR.D is paying 1.27275 and went up today to 21.30 to pay 5.974%, AQN,PR.A is paying 1.29 per year (reset at end of 2018) and went down today to 20.17 to pay 6.4%. They are only 3 months different in reset dates and likely have the same ex-dividend dates. So why is D getting bid up and A bid down on the same day when A pays more?

Or, more simply, is it just normal randomness that explains the price (and thus return) differential?

One thing that’s a little imprecise in your calculations is that you are using the closing price as the “price”. Today’s

quoteswere AQN.PR.A 20.16-17 and AQN.PR.D 21.15-30.So while both issues closed at the offer price, there’s a difference from the bid price. Using the bid price has its problems – you can regularly read me railing against the poor quality of the quotes provided by the Exchange – but is generally better to use than the closing price.

The main problem, however, is that you are ignoring the effect of the next round of resets, which will occur for AQN.PR.A on 2023-12-31 at GOC5 + 294bp, while the reset for AQN.PR.D will be effective 2024-3-31 at GOC5 + 328bp.

Assuming that GOC5 is constant at, call it 1.80%, this implies that AQN.PR.A will reset to 4.74%, or 1.185 p.a., while AQN.PR.D will reset to 5.08%, or 1.270 p.a. That’s a more significant difference than is now the case; the fact that GOC5 dropped in the interval between the current resets by an amount that nearly cancelled the difference in increment makes this a bit more obscure than it usually is.

In fact, when you do the math with all the discounting, the yields are virtually identical: AQN.PR.A yields 6.15% at the bid price of 20.16, while AQN.PR.D yields 6.12% at its bid of 21.15 (and only 6.06% at the offer of 21.30, which illustrates how small the difference is!).

I have developed a FixedReset Yield Calculator (Excel Spreadsheet) that handles the mechanics of the calculation; it is permanently linked under the heading “Calculators” in the Right-Hand Navigation Panel of this blog. Its use is discussed in detail in the post What is the Yield of HSE.PR.A?.

Welcome to the blog!

Or, more simply, is it just normal randomness that explains the price (and thus return) differential?There’s plenty of randomness, for sure. The signal to noise ratio is much lower in the preferred share market than most others … I devote quite a lot of time to dissecting the prices of preferreds in order to take action for, or make recommendations to, my clients.