PrefBlog

Malachite Aggressive Preferred Fund’s Net Asset Value per Unit as of the close June, 2020, was $6.3568 after a dividend distribution of 0.119146.

Performance for the month was hurt by the fund’s relatively low holdings of low-spread bank issues – see discussion below on the insights obtained by disaggregating the 1-month performance of the Pfd-2 Group.

Quote quality declined slightly this this month, with the difference in portfolio values when calculated with closing prices vs. calculation with bid prices increasing from about 0.6% to 0.8%.

MAPF returns assume reinvestment of dividends, and are shown after expenses but before fees. Past performance is not a guarantee of future performance. You can lose money investing in Malachite Aggressive Preferred Fund or any other fund. For more information, see the fund’s main page. The fund is available directly from Hymas Investment Management.

The preferred share market continues to be underpriced relative to other capital markets, leaving a lot of room for outsized gains. The Seniority Spread (the interest-equivalent yield on reasonably liquid, investment-grade PerpetualDiscounts less the yield on long term corporate bonds) is extremely elevated (chart end-date 2020-6-12):

Click for Big

Note that the Seniority Spread was recorded at 440bp shortly before month-end a slight (and possibly spurious) narrowing from the incredible 445bp near May month-end. As a good practical example of the spreads between markets, consider that CIU issued a long-term bond in early September yielding 2.963%, about 411bp cheaper than the interest-equivalent figure of 7.07% for CIU.PR.A, which was then yielding about 5.44% as a dividend. Shaw Communications issued 30-year notes at 4.25% interest on December 5, 2019, when their FixedResets, SJR.PR.A, were yielding 6.59% dividends; at the end of May, 2020, Pembina issued 30-year notes at 4.67% at a time when their FixedResets were yielding between 6.92% and 8.22% as dividends.

As has been noted, the increase in the Seniority Spread over the past one or two years has been due not to an increase in yield (drop in prices) of Straight Preferreds over the year, but largely because the yield of the Straight Preferreds has remained relatively constant while the yield of long-term corporate bonds has dropped dramatically.

… and the relationship between five-year Canada yields and yields on investment-grade FixedResets has gone even deeper into what I consider ‘decoupled panic’ territory (chart end-date 2020-6-12):

Click for Big

In addition, I feel that the yield on five-year Canadas is unsustainably low (it should be the inflation rate plus an increment of … 1%? 1.5%? 2.0%?),and a return to sustainable levels is likely over the medium term.

It seems clear that many market players are, wittingly or not, using FixedResets to speculate on future moves in the Canada 5-Year yield. This is excellent news for those who take market action based on fundamentals and the long term characteristics of the market because nobody can consistently time the markets. The speculators will, over the long run and in aggregate, lose money, handing it over to more sober investors.

It should be noted that I have been unable to explain the relatively strong performance of Floor issues during the 2018-19 downdraft relative to their non-Floor counterparts. See the discussions on PrefBlog at LINK, LINK and LINK.

I believe the bear-market outperformance by the Floor issues is a behavioural phenomenon with very little basis in fundamentals. When interest rates in general move, FixedReset prices should not change much (to a first approximation, for issues priced near par), since in Fixed Income investing it is spreads that are important, not absolute yields. There should be some effect on Floor issues, which should move up slightly in price as yields go down since the ‘option’ to receive the floor rate will become more valuable. Adjustments due to this effect should be fairly small, however – and over the past year issues with a floor, that started the period being expensive, have simply gotten even more expensive, relative to their non-floored counterparts.

And the tricky thing about behavioural models of investing is that they can lose their explanatory power very quickly when an investment fashion shifts, whereas fundamentals will always be effective – sometimes it just takes a little time! Just to give an example from the preferred share market – until the end of 2014, FixedResets were priced relative to each other according to their initial dividend; when the reset of TRP.PR.A shocked a lot of investors, relative pricing became much more dependent upon the Issue Reset Spread, a much more logical and fundamental property. This paradigm shift was discussed extensively in PrefLetter.

FixedReset (Discount) performance on the month was +3.24% vs. PerpetualDiscounts of +2.31% in May; the two classes finally decoupled in mid-November, 2018, after months of moving in lockstep, but it still appears to me that yields available on FixedResets are keeping the yields of PerpetualDiscounts up, even though a consistent valuation based on an expectation of declining interest rates would greatly increase the attractiveness of PerpetualDiscounts (in other words, PerpetualDiscounts are now priced off FixedResets rather than off Long-term Corporates):

Click for Big

Floaters underperformed FixedResets, returning +2.51% for June and the figure for the past twelve months remains awful at -25.00%. Look at the long-term performance:

Click for Big

Some Assiduous Readers will be interested to observe that the ‘Quantitative Easing’ decline was not initially as bad as the ‘Credit Crunch’ decline, which took the sector down to the point where the 15-year cumulative total return was negative. I wrote about that at the time. but it became worse in August, 2019! On August 30, 2019 the HIMI Floater Index (total return) value was calculated as 1906.6; the index first surpassed this value on 2003-8-13. Thus, cumulative total return (that is, including dividends) was negative over a period of slightly-over sixteen years. Worse, on March 31, 2020, the index level was 1454.8, a milestone first passed on 1997-7-30; a cumulative negative total return for 22 years and 8 months; at its low on March 18 the index level was 1253.7, first surpassed on 1996-1-4, a span of 24 years and over two months!

It seems clear that Floaters are used, wittingly or otherwise, as a vehicle for speculation on the policy rate and Canada Prime, while FixedResets are being used as a vehicle for speculation on the five-year Canada rate. In support of this idea, I present an Implied Volatility analysis of the TRP series of FixedResets as of May 29, which is comprised of six issues without a Minimum Rate Guarantee and two issues which do have this feature:

Click for Big

The two issues with floors, TRP.PR.J (+469, minimum 5.50%) and TRP.PR.K (+385, minimum 4.90%) are $4.39 and $4.18 rich, respectively. These figures are wider than the 4.10 and 3.38 calculated last month’s figures; however, it should be noted that their floors have become effective since five-year Canadas dipped below 0.81% and 1.05%, respectively. We expect something of an increase in fair value as noted above; but these levels seem elevated!

It will also be noted that the spread of a notional non-callable TRP FixedReset priced at par has declined slightly from 489bp last month to 481bp this month, while GOC-5 has been steady, moving from 0.36% to 0.37%.

I also show results for the BAM series of FixedResets, which includes three issues with dividend floors: BAM.PF.H (+417, Minimum 5.00%); BAM.PF.I (+386, Minimum 4.80%); and BAM.PF.J (+310, Minimum 4.75%); these issues are all rich compared to their non-floor siblings, being rich 1.99, 2.06 and 1.58 respectively, respectively, cheapening up a bit from last month’s figures of 2.17, 2.28 and 2.07

Click for Big

It will also be noted that the spread of a notional non-callable BAM FixedReset priced at par has remained fairly steady; 517bp last month to 526bp this month, while GOC-5 has declined from 0.45% to 0.36%.

Relative performance during the month was uncorrelated with Issue Reset Spreads for either the “Pfd-2 Group” or the “Pfd-3 Group” issues:

Click for Big

… and results over the quarter for the Pfd-2 Group were better correlated (26%) but uncorrelated for the Pfd-3 Group:

Click for Big

Disaggregating the one-month performance for the Pfd-2 group reveals an unusual pattern:

Click for Big

Bank issues with Issue Reset Spreads at the low end of their range performed unusually well and led to an unusual pattern in which the slope of the regression line for bank issues has a difference sign than that for insurance issues. I do not believe that this behaviour will continue.

As for the future, of course, it’s one thing to say that ‘spreads are unsustainable and so are government yields’ and it’s quite another to forecast just how and when a more economically sustainable environment will take effect. It could be years. The same caution applies for an end to the overpricing of issues with a minimum rate guarantee. There could be a reversal, particularly if either Trump’s international trade policies or the economic damage wreaked by the coronavirus approaches the gloomier extreme of current forecasts. And, of course, I could be just plain wrong about the sustainability of the current environment.

On the other hand, I will pass on my observation that international interest in the Canadian preferred share market is increasing, as other Floating Rate indices globally are doing much better. Consider, for example the Solactive Australian Bank Senior Floating Rate Bond Index, which “provides exposure to the largest and most liquid floating rate debt securities issued by selected Australian banks. The index is comprised of investment grade floating rate debt securities denominated in AUD and calculated as a Total Return Index” (LINK although the index constituents currently all have a remaining term of less than five years), and the S&P U.S. Floating Rate Preferred Stock Index.

Yields on preferred shares of all stripes are extremely high compared to those available from other investments of similar quality. As I told John Heinzl in an eMail interview in late November, 2018, the best advice I can offer investors remains Shut up and clip your coupons!

I think that a broad, sustainable rally in FixedResets will require higher five-year Canada yields (or a widespread expectation of them), since paradigm shifts generally require a trigger (a Wile E. Coyote moment, as they say!) … and although I’m sure this will happen eventually, it would be foolish to speculate on just when it will happen.

These calculations were performed assuming constant contemporary GOC-5 and 3-Month Bill rates, as follows:

I note that the sustainable yield calculated above is not directly comparable with any yield calculation currently reported by any other preferred share fund as far as I am aware. The Sustainable Yield depends on:
i) Calculating Yield-to-Worst for each instrument and using this yield for reporting purposes;
ii) Using the contemporary value of Five-Year Canadas to estimate dividends after reset for FixedResets. The assumption regarding the five-year Canada rate has become more important as the proportion of low-spread FixedResets in the portfolio has increased.
iii) Making the assumption that deeply discounted NVCC non-compliant issues from banks (and insurers, until November 2019), both Straight and FixedResets will be redeemed at par on their DeemedMaturity date as discussed above.

Turnover declined in June to 10%.

The fund’s trading will probably be higher in the future than has been normal for the past several years, since the extreme segmentation in the marketplace that I complained about for so long is now effectively ended. Low-Reset insurance issues were considered so cheap relative to their peers that a large portion of the fund’s holdings were effectively frozen. However, this differentiating factor is no longer considered applicable.

I am no longer making any adjustments for special qualities of insurance issues but note that this policy may change again in the future – a requirement for a Principal Loss Absorbency Mechanism (PLAM), whereby any security included in Tier 1 Capital will be wiped out prior to a government bail-out, even if technical bankruptcy is avoided, remains good public policy; it is a disgrace that the IAIS has rejected this principle and even worse that OSFI argued strenuously against it. I will continue to read notifications from these two entities with great interest, but while it is within the realm of possibility that ICS 2.0 will be revised following the expiry of the current five-year testing period, I can’t say I have any great confidence in the wisdom of the bureaucrats. However, it is a positive move that the increase in the limit for preferred share issuance was increased from 10% of the capital requirement to 15%; but this increase may only be met with issues having a PLAM.

Sectoral distribution of the MAPF portfolio on June 30 was as follows:

The “total” reflects the un-leveraged total portfolio (i.e., cash is included in the portfolio calculations and is deemed to have a duration and yield of 0.00.). MAPF will often have relatively large cash balances, both credit and debit, to facilitate trading. Figures presented in the table have been rounded to the indicated precision.

Credit distribution is:

Liquidity Distribution is:

The distribution of Issue Reset Spreads is:

Distribution of Floating Rate Start Dates is shown in the table below. This is the date of the next adjustment to the dividend rate, if the issue is currently paying a fixed rate for a limited time; which in practice is successive terms of 5 years. Issues that adjust quarterly are considered “Currently Floating”.

MAPF is, of course, Malachite Aggressive Preferred Fund, a “unit trust” managed by Hymas Investment Management Inc. Further information and links to performance, audited financials and subscription information are available the fund’s web page. The fund may be purchased directly from Hymas Investment Management. A “unit trust” is like a regular mutual fund, but are not sold with a prospectus. This is cheaper, but means subscription is restricted to “accredited investors” (as defined by the Ontario Securities Commission). Fund past performances are not a guarantee of future performance. You can lose money investing in MAPF or any other fund.

A similar portfolio composition analysis has been performed on the Claymore Preferred Share ETF (symbol CPD) (and other funds) as of July 31, 2017, and published in the August, 2017, PrefLetter. It is fair to say:

• MAPF credit quality is similar
• MAPF liquidity is lower
• MAPF Yield is higher
• Weightings
  • MAPF is less exposed to Straight Perpetuals
  • Neither portfolio is exposed to Operating Retractibles (there aren’t too many of those any more!)
  • MAPF is similarly exposed to SplitShares
  • MAPF is less exposed to FixFloat / Floater / Ratchet
  • MAPF is higher weighted in FixedResets, with a greater emphasis on lower-spread issues

Malachite Aggressive Preferred Fund’s Net Asset Value per Unit as of the close May 29, 2020, was $6.2721.

Quote quality is much better this this month, with the difference in portfolio values when calculated with closing prices vs. calculation with bid prices shrinking from about 1.9% to 0.6%.

Mind you, there are still puzzling differences in returns, as shown by the six series of TRP issues retained by the fund after it sold out of TRP.PR.E:

MAPF returns assume reinvestment of dividends, and are shown after expenses but before fees. Past performance is not a guarantee of future performance. You can lose money investing in Malachite Aggressive Preferred Fund or any other fund. For more information, see the fund’s main page. The fund is available directly from Hymas Investment Management.

The preferred share market continues to be underpriced relative to other capital markets, leaving a lot of room for outsized gains. The Seniority Spread (the interest-equivalent yield on reasonably liquid, investment-grade PerpetualDiscounts less the yield on long term corporate bonds) is extremely elevated (chart end-date 2020-5-8):

Click for Big

Note that the Seniority Spread was an incredible 445bp near month-end, a little wider than last month’s figure of 435bp. As a good practical example of the spreads between markets, consider that CIU issued a long-term bond in early September yielding 2.963%, about 411bp cheaper than the interest-equivalent figure of 7.07% for CIU.PR.A, which was then yielding about 5.44% as a dividend. Shaw Communications issued 30-year notes at 4.25% interest on December 5, 2019, when their FixedResets, SJR.PR.A, were yielding 6.59% dividends; at the end of May, 2020, Pembina issued 30-year notes at 4.67% at a time when their FixedResets were yielding between 6.92% and 8.22% as dividends.

As has been noted, the increase in the Seniority Spread over the past one or two years has been due not to an increase in yield (drop in prices) of Straight Preferreds over the year, but largely because the yield of the Straight Preferreds has remained relatively constant while the yield of long-term corporate bonds has dropped dramatically.

… and the relationship between five-year Canada yields and yields on investment-grade FixedResets has gone even deeper into what I consider ‘decoupled panic’ territory (chart end-date 2020-3-20):

Click for Big

In addition, I feel that the yield on five-year Canadas is unsustainably low (it should be the inflation rate plus an increment of … 1%? 1.5%? 2.0%?),and a return to sustainable levels is likely over the medium term.

It seems clear that many market players are, wittingly or not, using FixedResets to speculate on future moves in the Canada 5-Year yield. This is excellent news for those who take market action based on fundamentals and the long term characteristics of the market because nobody can consistently time the markets. The speculators will, over the long run and in aggregate, lose money, handing it over to more sober investors.

It should be noted that I have been unable to explain the relatively strong performance of Floor issues during the 2018-19 downdraft relative to their non-Floor counterparts. See the discussions on PrefBlog at LINK, LINK and LINK.

I believe the bear-market outperformance by the Floor issues is a behavioural phenomenon with very little basis in fundamentals. When interest rates in general move, FixedReset prices should not change much (to a first approximation, for issues priced near par), since in Fixed Income investing it is spreads that are important, not absolute yields. There should be some effect on Floor issues, which should move up slightly in price as yields go down since the ‘option’ to receive the floor rate will become more valuable. Adjustments due to this effect should be fairly small, however – and over the past year issues with a floor, that started the period being expensive, have simply gotten even more expensive, relative to their non-floored counterparts.

And the tricky thing about behavioural models of investing is that they can lose their explanatory power very quickly when an investment fashion shifts, whereas fundamentals will always be effective – sometimes it just takes a little time! Just to give an example from the preferred share market – until the end of 2014, FixedResets were priced relative to each other according to their initial dividend; when the reset of TRP.PR.A shocked a lot of investors, relative pricing became much more dependent upon the Issue Reset Spread, a much more logical and fundamental property. This paradigm shift was discussed extensively in PrefLetter.

FixedReset (Discount) performance on the month was -0.95% vs. PerpetualDiscounts of +1.41% in May; the two classes finally decoupled in mid-November, 2018, after months of moving in lockstep, but it still appears to me that yields available on FixedResets are keeping the yields of PerpetualDiscounts up, even though a consistent valuation based on an expectation of declining interest rates would greatly increase the attractiveness of PerpetualDiscounts (in other words, PerpetualDiscounts are now priced off FixedResets rather than off Long-term Corporates):

Click for Big

Floaters had yet another bad month, returning -4.22% for March and the figure for the past twelve months remains awful at -28.63%. Look at the long-term performance:

Click for Big

Some Assiduous Readers will be interested to observe that the ‘Quantitative Easing’ decline was not initially as bad as the ‘Credit Crunch’ decline, which took the sector down to the point where the 15-year cumulative total return was negative. I wrote about that at the time. but it became worse in August, 2019! On August 30, 2019 the HIMI Floater Index (total return) value was calculated as 1906.6; the index first surpassed this value on 2003-8-13. Thus, cumulative total return (that is, including dividends) was negative over a period of slightly-over sixteen years. Worse, on March 31 the index level was 1454.8, a milestone first passed on 1997-7-30; a cumulative negative total return for 22 years and 8 months; at its low on March 18 the index level was 1253.7, first surpassed on 1996-1-4, a span of 24 years and over two months!

It seems clear that Floaters are used, wittingly or otherwise, as a vehicle for speculation on the policy rate and Canada Prime, while FixedResets are being used as a vehicle for speculation on the five-year Canada rate. In support of this idea, I present an Implied Volatility analysis of the TRP series of FixedResets as of May 29, which is comprised of six issues without a Minimum Rate Guarantee and two issues which do have this feature:

Click for Big

The two issues with floors, TRP.PR.J (+469, minimum 5.50%) and TRP.PR.K (+385, minimum 4.90%) are $4.10 and $3.38 rich, respectively. These figures are not significantly different from the 4.42 and 3.38 calculated last month’s figures; however, it should be noted that their floors have become effective since five-year Canadas dipped below 0.81% and 1.05%, respectively. We expect something of an increase in fair value as noted above; but these levels seem elevated!

It will also be noted that the spread of a notional non-callable TRP FixedReset priced at par has increased from 470bp last month to 489bp this month, while GOC-5 has declined from 0.45% to 0.36%.

I also show results for the BAM series of FixedResets, which includes three issues with dividend floors: BAM.PF.H (+417, Minimum 5.00%); BAM.PF.I (+386, Minimum 4.80%); and BAM.PF.J (+310, Minimum 4.75%); these issues are all rich compared to their non-floor siblings, being rich 2.17, 2.28 and 2.07, respectively, much more expensive than last month’s figures of 1.63, 1.25 and 1.07.

Click for Big

It will also be noted that the spread of a notional non-callable BAM FixedReset priced at par has remained fairly steady; 517bp last month to 526bp this month, while GOC-5 has declined from 0.45% to 0.36%.

Relative performance during the month was uncorrelated with Issue Reset Spreads for either the “Pfd-2 Group” or the “Pfd-3 Group” issues:

Click for Big

… and results over the quarter for the Pfd-2 Group were better correlated (26%) but uncorrelated for the Pfd-3 Group:

Click for Big

In the three-month chart, there are four data points in the Pfd-2 Group that are well below the range of the remainder. These are the Husky Energy issues, HSE.PR.A, HSE.PR.C, HSE.PR.E and HSE.PR.G, which have also suffered from the Saudi-Russian oil price war.

As for the future, of course, it’s one thing to say that ‘spreads are unsustainable and so are government yields’ and it’s quite another to forecast just how and when a more economically sustainable environment will take effect. It could be years. The same caution applies for an end to the overpricing of issues with a minimum rate guarantee. There could be a reversal, particularly if either Trump’s international trade policies or the economic damage wreaked by the coronavirus approaches the gloomier extreme of current forecasts. And, of course, I could be just plain wrong about the sustainability of the current environment.

On the other hand, I will pass on my observation that international interest in the Canadian preferred share market is increasing, as other Floating Rate indices globally are doing much better. Consider, for example the Solactive Australian Bank Senior Floating Rate Bond Index, which “provides exposure to the largest and most liquid floating rate debt securities issued by selected Australian banks. The index is comprised of investment grade floating rate debt securities denominated in AUD and calculated as a Total Return Index” (LINK although the index constituents currently all have a remaining term of less than five years), and the S&P U.S. Floating Rate Preferred Stock Index.

Yields on preferred shares of all stripes are extremely high compared to those available from other investments of similar quality. As I told John Heinzl in an eMail interview in late November, 2018, the best advice I can offer investors remains Shut up and clip your coupons!

I think that a broad, sustainable rally in FixedResets will require higher five-year Canada yields (or a widespread expectation of them), since paradigm shifts generally require a trigger (a Wile E. Coyote moment, as they say!) … and although I’m sure this will happen eventually, it would be foolish to speculate on just when it will happen.

These calculations were performed assuming constant contemporary GOC-5 and 3-Month Bill rates, as follows:

I note that the sustainable yield calculated above is not directly comparable with any yield calculation currently reported by any other preferred share fund as far as I am aware. The Sustainable Yield depends on:
i) Calculating Yield-to-Worst for each instrument and using this yield for reporting purposes;
ii) Using the contemporary value of Five-Year Canadas to estimate dividends after reset for FixedResets. The assumption regarding the five-year Canada rate has become more important as the proportion of low-spread FixedResets in the portfolio has increased.
iii) Making the assumption that deeply discounted NVCC non-compliant issues from banks (and insurers, until November 2019), both Straight and FixedResets will be redeemed at par on their DeemedMaturity date as discussed above.

Turnover declined in May to 15%.

The fund’s trading will probably be higher in the future than has been normal for the past several years, since the extreme segmentation in the marketplace that I complained about for so long is now effectively ended. Low-Reset insurance issues were considered so cheap relative to their peers that a large portion of the fund’s holdings were effectively frozen. However, this differentiating factor is no longer considered applicable.

I am no longer making any adjustments for special qualities of insurance issues but note that this policy may change again in the future – a requirement for a Principal Loss Absorbency Mechanism (PLAM), whereby any security included in Tier 1 Capital will be wiped out prior to a government bail-out, even if technical bankruptcy is avoided, remains good public policy; it is a disgrace that the IAIS has rejected this principle and even worse that OSFI argued strenuously against it. I will continue to read notifications from these two entities with great interest, but while it is within the realm of possibility that ICS 2.0 will be revised following the expiry of the current five-year testing period, I can’t say I have any great confidence in the wisdom of the bureaucrats. However, it is a positive move that the increase in the limit for preferred share issuance was increased from 10% of the capital requirement to 15%; but this increase may only be met with issues having a PLAM.

Sectoral distribution of the MAPF portfolio on May 29 was as follows:

The “total” reflects the un-leveraged total portfolio (i.e., cash is included in the portfolio calculations and is deemed to have a duration and yield of 0.00.). MAPF will often have relatively large cash balances, both credit and debit, to facilitate trading. Figures presented in the table have been rounded to the indicated precision.

Credit distribution is:

Liquidity Distribution is:

The distribution of Issue Reset Spreads is:

Distribution of Floating Rate Start Dates is shown in the table below. This is the date of the next adjustment to the dividend rate, if the issue is currently paying a fixed rate for a limited time; which in practice is successive terms of 5 years. Issues that adjust quarterly are considered “Currently Floating”.

MAPF is, of course, Malachite Aggressive Preferred Fund, a “unit trust” managed by Hymas Investment Management Inc. Further information and links to performance, audited financials and subscription information are available the fund’s web page. The fund may be purchased directly from Hymas Investment Management. A “unit trust” is like a regular mutual fund, but are not sold with a prospectus. This is cheaper, but means subscription is restricted to “accredited investors” (as defined by the Ontario Securities Commission). Fund past performances are not a guarantee of future performance. You can lose money investing in MAPF or any other fund.

A similar portfolio composition analysis has been performed on the Claymore Preferred Share ETF (symbol CPD) (and other funds) as of July 31, 2017, and published in the August, 2017, PrefLetter. It is fair to say:

• MAPF credit quality is much better
• MAPF liquidity is lower
• MAPF Yield is higher
• Weightings
  • MAPF is less exposed to Straight Perpetuals
  • Neither portfolio is exposed to Operating Retractibles (there aren’t too many of those any more!)
  • MAPF is a little more exposed to SplitShares
  • MAPF is less exposed to FixFloat / Floater / Ratchet
  • MAPF is higher weighted in FixedResets, with a greater emphasis on lower-spread issues

The Financial Statements and related documents for Malachite Aggressive Preferred Fund are now available on the fund’s main page:

• Audited Financial Statements, 2019
• Transactions in 2019

Malachite Aggressive Preferred Fund’s Net Asset Value per Unit as of the close April 30, 2020, was $6.2894.

Attribution analysis, even of the most rudimentary kind, is very difficult this month, due largely to continued awful quotes provided by the Toronto Stock Exchange. For example, the fund holds seven different series of TRP preferreds, with the largest holdings in TRP.PR.D and TRP.PR.G:

And two series of TD:

MAPF returns assume reinvestment of dividends, and are shown after expenses but before fees. Past performance is not a guarantee of future performance. You can lose money investing in Malachite Aggressive Preferred Fund or any other fund. For more information, see the fund’s main page. The fund is available directly from Hymas Investment Management.

This has been the best month for the BMO-CM “50” index since the beginning of my data for this index, December, 1992. The index return of +12.42% is well ahead of that of the second-place March, 2016, which was a “mere” +9.86%. This all seems quite appropriate since March, 2020, was by far the worst month ever – but we have a long way to go before we even break even on a one-year basis!

The preferred share market continues to be underpriced relative to other capital markets, leaving a lot of room for outsized gains. The Seniority Spread (the interest-equivalent yield on reasonably liquid, investment-grade PerpetualDiscounts less the yield on long term corporate bonds) is extremely elevated (chart end-date 2020-4-9):

Click for Big

Note that the Seniority Spread was an incredible 435bp near month-end, much narrower than last month’s figure of 515bp. As a good practical example of the spreads between markets, consider that CIU issued a long-term bond in early September yielding 2.963%, about 411bp cheaper than the interest-equivalent figure of 7.07% for CIU.PR.A, which was then yielding about 5.44% as a dividend. Shaw Communications issued 30-year notes at 4.25% interest on December 5, 2019, when their FixedResets, SJR.PR.A, were yielding 6.59% dividends.

As has been noted, the increase in the Seniority Spread over the past one or two years has been due not to an increase in yield (drop in prices) of Straight Preferreds over the year, but largely because the yield of the Straight Preferreds has remained relatively constant while the yield of long-term corporate bonds has dropped dramatically. This month’s change breaks the pattern (as did last month’s mover upwards), as long-term corporate bond yields decreased by 49bp through the measured period, while PerpetualDiscount Interest-Equivalent yields decreased by 125bp.

… and the relationship between five-year Canada yields and yields on investment-grade FixedResets has gone even deeper into what I consider ‘decoupled panic’ territory (chart end-date 2020-3-20):

Click for Big

In addition, I feel that the yield on five-year Canadas is unsustainably low (it should be the inflation rate plus an increment of … 1%? 1.5%? 2.0%?),and a return to sustainable levels is likely over the medium term.

It seems clear that many market players are, wittingly or not, using FixedResets to speculate on future moves in the Canada 5-Year yield. This is excellent news for those who take market action based on fundamentals and the long term characteristics of the market because nobody can consistently time the markets. The speculators will, over the long run and in aggregate, lose money, handing it over to more sober investors.

It should be noted that I have been unable to explain the relatively strong performance of Floor issues during the 2018-19 downdraft relative to their non-Floor counterparts. See the discussions on PrefBlog at LINK, LINK and LINK.

I believe the bear-market outperformance by the Floor issues is a behavioural phenomenon with very little basis in fundamentals. When interest rates in general move, FixedReset prices should not change much (to a first approximation, for issues priced near par), since in Fixed Income investing it is spreads that are important, not absolute yields. There should be some effect on Floor issues, which should move up slightly in price as yields go down since the ‘option’ to receive the floor rate will become more valuable. Adjustments due to this effect should be fairly small, however – and over the past year issues with a floor, that started the period being expensive, have simply gotten even more expensive, relative to their non-floored counterparts.

And the tricky thing about behavioural models of investing is that they can lose their explanatory power very quickly when an investment fashion shifts, whereas fundamentals will always be effective – sometimes it just takes a little time! Just to give an example from the preferred share market – until the end of 2014, FixedResets were priced relative to each other according to their initial dividend; when the reset of TRP.PR.A shocked a lot of investors, relative pricing became much more dependent upon the Issue Reset Spread, a much more logical and fundamental property. This paradigm shift was discussed extensively in PrefLetter.

FixedReset (Discount) performance on the month was +13.10% vs. PerpetualDiscounts of +13.13% in April; the two classes finally decoupled in mid-November, 2018, after months of moving in lockstep, but it still appears to me that yields available on FixedResets are keeping the yields of PerpetualDiscounts up, even though a consistent valuation based on an expectation of declining interest rates would greatly increase the attractiveness of PerpetualDiscounts (in other words, PerpetualDiscounts are now priced off FixedResets rather than off Long-term Corporates):

Click for Big

Floaters recovered somewhat, returning +7.12% for March but the figure for the past twelve months remains awful at -29.43%. Look at the long-term performance:

Click for Big

Some Assiduous Readers will be interested to observe that the ‘Quantitative Easing’ decline was not initially as bad as the ‘Credit Crunch’ decline, which took the sector down to the point where the 15-year cumulative total return was negative. I wrote about that at the time. but it became worse in August, 2019! On August 30, 2019 the HIMI Floater Index (total return) value was calculated as 1906.6; the index first surpassed this value on 2003-8-13. Thus, cumulative total return (that is, including dividends) was negative over a period of slightly-over sixteen years. Worse, on March 31 the index level was 1454.8, a milestone first passed on 1997-7-30; a cumulative negative total return for 22 years and 8 months; at its low on March 18 the index level was 1253.7, first surpassed on 1996-1-4, a span of 24 years and over two months!

It seems clear that Floaters are used, wittingly or otherwise, as a vehicle for speculation on the policy rate and Canada Prime, while FixedResets are being used as a vehicle for speculation on the five-year Canada rate. In support of this idea, I present an Implied Volatility analysis of the TRP series of FixedResets as of April 30, which is comprised of six issues without a Minimum Rate Guarantee and two issues which do have this feature:

Click for Big

The two issues with floors, TRP.PR.J (+469, minimum 5.50%) and TRP.PR.K (+385, minimum 4.90%) are $4.42 and $3.38 rich, respectively. These figures are a little higher than the 3.45 and 2.64 calculated last month’s figures; however, it should be noted that their floors have become effective since five-year Canadas dipped below 0.81% and 1.05%, respectively. We expect something of an increase in fair value as noted above; but these levels seem elevated!

It will also be noted that the spread of a notional non-callable TRP FixedReset priced at par has increased from 541bp last month to 470bp this month, while GOC-5 has declined from 0.57% to 0.45%.

I also show results for the BAM series of FixedResets, which includes three issues with dividend floors: BAM.PF.H (+417, Minimum 5.00%); BAM.PF.I (+386, Minimum 4.80%); and BAM.PF.J (+310, Minimum 4.75%); these issues are all rich compared to their non-floor siblings, being rich 1.63, 1.25 and 1.07, respectively, much more expensive than last month’s figures of 0.45, 0.46 and 0.05.

Click for Big

It will also be noted that the spread of a notional non-callable BAM FixedReset priced at par has declined from 557bp last month to 517bp this month, while GOC-5 has declined from 0.57% to 0.45%. This is very similar to the effect seen for TRP. This is mercifully consistent with the TRP results.

Relative performance during the month was uncorrelated with Issue Reset Spreads for either the “Pfd-2 Group” or the “Pfd-3 Group” issues:

Click for Big

… and results over the quarter for the Pfd-2 Group were better correlated (30%) but uncorrelated for the Pfd-3 Group:

Click for Big

In the three-month chart, there are four data points in the Pfd-2 Group that are well below the range of the remainder. These are the Husky Energy issues, HSE.PR.A, HSE.PR.C, HSE.PR.E and HSE.PR.G, which have also suffered from the Saudi-Russian oil price war.

As for the future, of course, it’s one thing to say that ‘spreads are unsustainable and so are government yields’ and it’s quite another to forecast just how and when a more economically sustainable environment will take effect. It could be years. The same caution applies for an end to the overpricing of issues with a minimum rate guarantee. There could be a reversal, particularly if either Trump’s international trade policies or the economic damage wreaked by the coronavirus approaches the gloomier extreme of current forecasts. And, of course, I could be just plain wrong about the sustainability of the current environment.

On the other hand, I will pass on my observation that international interest in the Canadian preferred share market is increasing, as other Floating Rate indices globally are doing much better. Consider, for example the Solactive Australian Bank Senior Floating Rate Bond Index, which “provides exposure to the largest and most liquid floating rate debt securities issued by selected Australian banks. The index is comprised of investment grade floating rate debt securities denominated in AUD and calculated as a Total Return Index” (LINK although the index constituents currently all have a remaining term of less than five years), and the S&P U.S. Floating Rate Preferred Stock Index.

Yields on preferred shares of all stripes are extremely high compared to those available from other investments of similar quality. As I told John Heinzl in an eMail interview in late November, 2018, the best advice I can offer investors remains Shut up and clip your coupons!

I think that a broad, sustainable rally in FixedResets will require higher five-year Canada yields (or a widespread expectation of them), since paradigm shifts generally require a trigger (a Wile E. Coyote moment, as they say!) … and although I’m sure this will happen eventually, it would be foolish to speculate on just when it will happen.

These calculations were performed assuming constant contemporary GOC-5 and 3-Month Bill rates, as follows:

I note that the sustainable yield calculated above is not directly comparable with any yield calculation currently reported by any other preferred share fund as far as I am aware. The Sustainable Yield depends on:
i) Calculating Yield-to-Worst for each instrument and using this yield for reporting purposes;
ii) Using the contemporary value of Five-Year Canadas to estimate dividends after reset for FixedResets. The assumption regarding the five-year Canada rate has become more important as the proportion of low-spread FixedResets in the portfolio has increased.
iii) Making the assumption that deeply discounted NVCC non-compliant issues from banks (and insurers, until November 2019), both Straight and FixedResets will be redeemed at par on their DeemedMaturity date as discussed above.

Turnover increased sharply in April to 32%. However, very wide spreads largely prevented optimization trading, while there were enough knowledgeable participants in the market to prevent price differences between similar issues from getting far enough out of line to overcome that hurdle.

The fund’s trading will probably be higher in the future than has been normal for the past several years, since the extreme segmentation in the marketplace that I complained about for so long is now effectively ended. Low-Reset insurance issues were considered so cheap relative to their peers that a large portion of the fund’s holdings were effectively frozen. However, this differentiating factor is no longer considered applicable.

I am no longer making any adjustments for special qualities of insurance issues but note that this policy may change again in the future – a requirement for a Principal Loss Absorbency Mechanism (PLAM), whereby any security included in Tier 1 Capital will be wiped out prior to a government bail-out, even if technical bankruptcy is avoided, remains good public policy; it is a disgrace that the IAIS has rejected this principle and even worse that OSFI argued strenuously against it. I will continue to read notifications from these two entities with great interest, but while it is within the realm of possibility that ICS 2.0 will be revised following the expiry of the current five-year testing period, I can’t say I have any great confidence in the wisdom of the bureaucrats. However, it is a positive move that the increase in the limit for preferred share issuance was increased from 10% of the capital requirement to 15%; but this increase may only be met with issues having a PLAM.

Sectoral distribution of the MAPF portfolio on April 30 was as follows:

The “total” reflects the un-leveraged total portfolio (i.e., cash is included in the portfolio calculations and is deemed to have a duration and yield of 0.00.). MAPF will often have relatively large cash balances, both credit and debit, to facilitate trading. Figures presented in the table have been rounded to the indicated precision.

Credit distribution is:

Liquidity Distribution is:

The distribution of Issue Reset Spreads is:

Distribution of Floating Rate Start Dates is shown in the table below. This is the date of the next adjustment to the dividend rate, if the issue is currently paying a fixed rate for a limited time; which in practice is successive terms of 5 years. Issues that adjust quarterly are considered “Currently Floating”.

MAPF is, of course, Malachite Aggressive Preferred Fund, a “unit trust” managed by Hymas Investment Management Inc. Further information and links to performance, audited financials and subscription information are available the fund’s web page. The fund may be purchased directly from Hymas Investment Management. A “unit trust” is like a regular mutual fund, but are not sold with a prospectus. This is cheaper, but means subscription is restricted to “accredited investors” (as defined by the Ontario Securities Commission). Fund past performances are not a guarantee of future performance. You can lose money investing in MAPF or any other fund.

A similar portfolio composition analysis has been performed on the Claymore Preferred Share ETF (symbol CPD) (and other funds) as of July 31, 2017, and published in the August, 2017, PrefLetter. It is fair to say:

• MAPF credit quality is much better
• MAPF liquidity is lower
• MAPF Yield is higher
• Weightings
  • MAPF is less exposed to Straight Perpetuals
  • Neither portfolio is exposed to Operating Retractibles (there aren’t too many of those any more!)
  • MAPF is a little more exposed to SplitShares
  • MAPF is less exposed to FixFloat / Floater / Ratchet
  • MAPF is higher weighted in FixedResets, with a greater emphasis on lower-spread issues

Malachite Aggressive Preferred Fund’s Net Asset Value per Unit as of the close March 31, 2020, was $5.5596 after a distribution of 0.149833.

The fund’s performance was hurt in the month and the quarter by its holdings in Husky Energy preferreds, which have shockingly underperformed the market as discussed below. Husky, while under Review-Negative at DBRS remains classified as investment grade by that firm; a review by Standard & Poor’s resulted in a Negative Outlook assignment but the credit rating itself (P-3(high)) was unchanged, in contrast to other Canadian energy firms.

MAPF returns assume reinvestment of dividends, and are shown after expenses but before fees. Past performance is not a guarantee of future performance. You can lose money investing in Malachite Aggressive Preferred Fund or any other fund. For more information, see the fund’s main page. The fund is available directly from Hymas Investment Management.

This has been the worst month for the BMO-CM “50” index since the beginning of my data for this index, December, 1992. The index return of -19.92% is nearly double that of the second-place November, 2008, which was a “mere” -10.70% in the depths of the Credit Crunch, and more than double the worst of the “Quantitative Easing” bloodbath which managed “only” -8.46%. Surprisingly, the year to month-end was only the second-worst I have on record, as the return to March 31, 2020 was -21.65%, while the return for the year ending 2016-2-29 was -22.09%.

What really sets this decline apart from other, though, is that it comes with a ferocity undampened by prior declines. The total return index level of 345.20 was first exceeded in January, 2010, meaning that the preferred share market has now had a negative total return for a holding period of ten years and two months.

Click for Big

The prior worst 122-month rolling return was for the period ended 2016-2-29, which marked a minimum at +2.09%; this period included both the Credit Crunch and the “Quantitative Easing” market hits; prior to then, the low had been at the depth of the Credit Crunch, November 2008, with a 122-month return of +7.78%. However, the current situation is the worst in our experience, with a cumulative total return of -0.25% for the 122 month period.

The preferred share market continues to be underpriced relative to other capital markets, leaving a lot of room for outsized gains. The Seniority Spread (the interest-equivalent yield on reasonably liquid, investment-grade PerpetualDiscounts less the yield on long term corporate bonds) is extremely elevated (chart end-date 2020-3-20):

Click for Big

Note that the Seniority Spread was an incredible 515bp near month-end, ridiculously wider last month’s figure of 385bp. As a good practical example of the spreads between markets, consider that CIU issued a long-term bond in early September yielding 2.963%, about 411bp cheaper than the interest-equivalent figure of 7.07% for CIU.PR.A, which was then yielding about 5.44% as a dividend. Shaw Communications issued 30-year notes at 4.25% interest on December 5, 2019, when their FixedResets, SJR.PR.A, were yielding 6.59% dividends.

As has been noted, the increase in the Seniority Spread over the past one or two years has been due not to an increase in yield (drop in prices) of Straight Preferreds over the year, but largely because the yield of the Straight Preferreds has remained relatively constant while the yield of long-term corporate bonds has dropped dramatically. This month’s change breaks the pattern, as long-term corporate bond yields increased by 87bp through the measured period, while PerpetualDiscount Interest-Equivalent yields increased by an jaw-dropping 215bp.

… and the relationship between five-year Canada yields and yields on investment-grade FixedResets has gone even deeper into what I consider ‘decoupled panic’ territory (chart end-date 2020-3-20):

Click for Big

In addition, I feel that the yield on five-year Canadas is unsustainably low (it should be the inflation rate plus an increment of … 1%? 1.5%? 2.0%?),and a return to sustainable levels is likely over the medium term.

It seems clear that many market players are, wittingly or not, using FixedResets to speculate on future moves in the Canada 5-Year yield. This is excellent news for those who take market action based on fundamentals and the long term characteristics of the market because nobody can consistently time the markets. The speculators will, over the long run and in aggregate, lose money, handing it over to more sober investors.

It should be noted that I have been unable to explain the relatively strong performance of Floor issues during the 2018-19 downdraft relative to their non-Floor counterparts. See the discussions on PrefBlog at LINK, LINK and LINK.

I believe the bear-market outperformance by the Floor issues is a behavioural phenomenon with very little basis in fundamentals. When interest rates in general move, FixedReset prices should not change much (to a first approximation, for issues priced near par), since in Fixed Income investing it is spreads that are important, not absolute yields. There should be some effect on Floor issues, which should move up slightly in price as yields go down since the ‘option’ to receive the floor rate will become more valuable. Adjustments due to this effect should be fairly small, however – and over the past year issues with a floor, that started the period being expensive, have simply gotten even more expensive, relative to their non-floored counterparts.

And the tricky thing about behavioural models of investing is that they can lose their explanatory power very quickly when an investment fashion shifts, whereas fundamentals will always be effective. Just to give an example from the preferred share market – until the end of 2014, FixedResets were priced relative to each other according to their initial dividend; when the reset of TRP.PR.A shocked a lot of investors, relative pricing became much more dependent upon the Issue Reset Spread, a much more logical and fundamental property. This paradigm shift was discussed extensively in PrefLetter.

FixedReset (Discount) performance on the month was -22.62% vs. PerpetualDiscounts of -16.04% in March; the two classes finally decoupled in mid-November, 2018, after months of moving in lockstep, but it still appears to me that yields available on FixedResets are keeping the yields of PerpetualDiscounts up, even though a consistent valuation based on an expectation of declining interest rates would greatly increase the attractiveness of PerpetualDiscounts (in other words, PerpetualDiscounts are now priced off FixedResets rather than off Long-term Corporates):

Click for Big

Floaters had a poor month, returning -27.08% for March and the figure for the past twelve months has crashed to -33.31%. Look at the long-term performance:

Click for Big

Some Assiduous Readers will be interested to observe that the ‘Quantitative Easing’ decline was not initially as bad as the ‘Credit Crunch’ decline, which took the sector down to the point where the 15-year cumulative total return was negative. I wrote about that at the time. but it became worse in August, 2019! On August 30, 2019 the HIMI Floater Index (total return) value was calculated as 1906.6; the index first surpassed this value on 2003-8-13. Thus, cumulative total return (that is, including dividends) was negative over a period of slightly-over sixteen years. Now, however, it’s even worse. On March 31 the index level was 1454.8, a milestone first passed on 1997-7-30; a cumulative negative total return for 22 years and 8 months; at its low on March 18 the index level was 1253.7, first surpassed on 1996-1-4, a span of 24 years and over two months!

It seems clear that Floaters are used, wittingly or otherwise, as a vehicle for speculation on the policy rate and Canada Prime, while FixedResets are being used as a vehicle for speculation on the five-year Canada rate. In support of this idea, I present an Implied Volatility analysis of the TRP series of FixedResets as of February 28, which is comprised of six issues without a Minimum Rate Guarantee and two issues which do have this feature:

Click for Big

The two issues with floors, TRP.PR.J (+469, minimum 5.50%) and TRP.PR.K (+385, minimum 4.90%) are $3.45 and $2.64 rich, respectively. These figures are a little higher than the 2.48 and 4.12 calculated last month figures; however, it should be noted that their floors have become effective since five-year Canadas dipped below 0.81% and 1.05%, respectively. We expect something of an increase in fair value as noted above; but these levels seem elevated!

It will also be noted that the spread of a notional non-callable TRP FixedReset priced at par has increased from 462bp last month to 541bp this month, while GOC-5 has declined from 1.07% to 0.57%.

I also show results for the BAM series of FixedResets, which includes three issues with dividend floors: BAM.PF.H (+417, Minimum 5.00%); BAM.PF.I (+386, Minimum 4.80%); and BAM.PF.J (+310, Minimum 4.75%); these issues are all rich compared to their non-floor siblings, being rich 0.45, 0.46 and 0.05, respectively, much more expensive than last month’s figures of 2.15, 3.32 and 4.19. In this case the guarantee rate “option” is now in the money and yet the issues are only being valued as ‘normal’ FixedResets. While this can be explained by suggesting that PerpetualDiscounts and FixedResets are being priced at the same levels, this cannot be reconciled to the richness of the TRP issues discussed above. The preferred share market throws up a new puzzle every month!

Click for Big

It will also be noted that the spread of a notional non-callable BAM FixedReset priced at par has increased from 462bp last month to 557bp this month, while GOC-5 has declined from 1.07% to 0.57%. This is very similar to the effect seen for TRP. This is mercifully consistent with the TRP results.

Relative performance during the month was uncorrelated with Issue Reset Spreads for either the “Pfd-2 Group” or the “Pfd-3 Group” issues:

Click for Big

… and results over the quarter for the Pfd-2 Group were better correlated (22%) but uncorrelated for the Pfd-3 Group:

Click for Big

In both charts, there are four data points in the Pfd-2 Group that are well below the range of the remainder. These are the Husky Energy issues, HSE.PR.A, HSE.PR.C, HSE.PR.E and HSE.PR.G, which have also suffered from the Saudi-Russian oil price war.

As for the future, of course, it’s one thing to say that ‘spreads are unsustainable and so are government yields’ and it’s quite another to forecast just how and when a more economically sustainable environment will take effect. It could be years. The same caution applies for an end to the overpricing of issues with a minimum rate guarantee. There could be a reversal, particularly if either Trump’s international trade policies or the economic damage wreaked by the coronavirus approaches the gloomier extreme of current forecasts. And, of course, I could be just plain wrong about the sustainability of the current environment.

On the other hand, I will pass on my observation that international interest in the Canadian preferred share market is increasing, as other Floating Rate indices globally are doing much better. Consider, for example the Solactive Australian Bank Senior Floating Rate Bond Index, which “provides exposure to the largest and most liquid floating rate debt securities issued by selected Australian banks. The index is comprised of investment grade floating rate debt securities denominated in AUD and calculated as a Total Return Index” (LINK although the index constituents currently all have a remaining term of less than five years), and the S&P U.S. Floating Rate Preferred Stock Index.

Yields on preferred shares of all stripes are extremely high compared to those available from other investments of similar quality. As I told John Heinzl in an eMail interview in late November, 2018, the best advice I can offer investors remains Shut up and clip your coupons!

I think that a broad, sustainable rally in FixedResets will require higher five-year Canada yields (or a widespread expectation of them), since paradigm shifts generally require a trigger (a Wile E. Coyote moment, as they say!) … and although I’m sure this will happen eventually, it would be foolish to speculate on just when it will happen.

These calculations were performed assuming constant contemporary GOC-5 and 3-Month Bill rates, as follows:

Although it is disappointing to see the estimated sustainable income declining over the quarter, this is explained by the precipitous drop in Canada yields, with the five-year rate down well over a point from year-end levels. As yields recover, the estimated sustainable income should also bounce back nicely.

I will also note that the sustainable yield calculated above is not directly comparable with any yield calculation currently reported by any other preferred share fund as far as I am aware. The Sustainable Yield depends on:
i) Calculating Yield-to-Worst for each instrument and using this yield for reporting purposes;
ii) Using the contemporary value of Five-Year Canadas to estimate dividends after reset for FixedResets. The assumption regarding the five-year Canada rate has become more important as the proportion of low-spread FixedResets in the portfolio has increased.
iii) Making the assumption that deeply discounted NVCC non-compliant issues from banks (and insurers, until November 2019), both Straight and FixedResets will be redeemed at par on their DeemedMaturity date as discussed above.

Turnover declined sharply in March to 15%. Very wide spreads largely prevented optimization trading, while there were enough knowledgeable participants in the market to prevent price differences between similar issues from getting far enough out of line to overcome that hurdle.

The fund’s trading will probably be higher in the future than has been normal for the past several years, since the extreme segmentation in the marketplace that I complained about for so long is now effectively ended. Low-Reset insurance issues were considered so cheap relative to their peers that a large portion of the fund’s holdings were effectively frozen. However, this differentiating factor is no longer considered applicable.

I am no longer making any adjustments for special qualities of insurance issues but note that this policy may change again in the future – a requirement for a Principal Loss Absorbency Mechanism (PLAM), whereby any security included in Tier 1 Capital will be wiped out prior to a government bail-out, even if technical bankruptcy is avoided, remains good public policy; it is a disgrace that the IAIS has rejected this principle and even worse that OSFI argued strenuously against it. I will continue to read notifications from these two entities with great interest, but while it is within the realm of possibility that ICS 2.0 will be revised following the expiry of the current five-year testing period, I can’t say I have any great confidence in the wisdom of the bureaucrats. However, it is a positive move that the increase in the limit for preferred share issuance was increased from 10% of the capital requirement to 15%; but this increase may only be met with issues having a PLAM.

Sectoral distribution of the MAPF portfolio on March 31 was as follows:

The “total” reflects the un-leveraged total portfolio (i.e., cash is included in the portfolio calculations and is deemed to have a duration and yield of 0.00.). MAPF will often have relatively large cash balances, both credit and debit, to facilitate trading. Figures presented in the table have been rounded to the indicated precision.

Credit distribution is:

Liquidity Distribution is:

The distribution of Issue Reset Spreads is:

Distribution of Floating Rate Start Dates is shown in the table below. This is the date of the next adjustment to the dividend rate, if the issue is currently paying a fixed rate for a limited time; which in practice is successive terms of 5 years. Issues that adjust quarterly are considered “Currently Floating”.

MAPF is, of course, Malachite Aggressive Preferred Fund, a “unit trust” managed by Hymas Investment Management Inc. Further information and links to performance, audited financials and subscription information are available the fund’s web page. The fund may be purchased directly from Hymas Investment Management. A “unit trust” is like a regular mutual fund, but are not sold with a prospectus. This is cheaper, but means subscription is restricted to “accredited investors” (as defined by the Ontario Securities Commission). Fund past performances are not a guarantee of future performance. You can lose money investing in MAPF or any other fund.

A similar portfolio composition analysis has been performed on the Claymore Preferred Share ETF (symbol CPD) (and other funds) as of July 31, 2017, and published in the August, 2017, PrefLetter. It is fair to say:

• MAPF credit quality is much better
• MAPF liquidity is lower
• MAPF Yield is higher
• Weightings
  • MAPF is less exposed to Straight Perpetuals
  • Neither portfolio is exposed to Operating Retractibles (there aren’t too many of those any more!)
  • MAPF is a little more exposed to SplitShares
  • MAPF is less exposed to FixFloat / Floater / Ratchet
  • MAPF is higher weighted in FixedResets, with a greater emphasis on lower-spread issues

Malachite Aggressive Preferred Fund’s Net Asset Value per Unit as of the close February 28, 2020, was $7.6268.

MAPF returns assume reinvestment of dividends, and are shown after expenses but before fees. Past performance is not a guarantee of future performance. You can lose money investing in Malachite Aggressive Preferred Fund or any other fund. For more information, see the fund’s main page. The fund is available directly from Hymas Investment Management.

The preferred share market continues to be underpriced relative to other capital markets, leaving a lot of room for outsized gains. The Seniority Spread (the interest-equivalent yield on reasonably liquid, investment-grade PerpetualDiscounts less the yield on long term corporate bonds) is extremely elevated (chart end-date 2020-2-12):

Click for Big

Note that the Seniority Spread was 385bp near month-end, equal to last month’s figure. As a good practical example of the spreads between markets, consider that CIU issued a long-term bond in early September yielding 2.963%, about 411bp cheaper than the interest-equivalent figure of 7.07% for CIU.PR.A, which was then yielding about 5.44% as a dividend. Shaw Communications issued 30-year notes at 4.25% interest on December 5, 2019, when their FixedResets, SJR.PR.A, were yielding 6.59% dividends.

As has been noted, the increase in the Seniority Spread over the past one or two years has been due not to an increase in yield (drop in prices) of Straight Preferreds over the year, but because the yield of the Straight Preferreds has remained relatively constant while the yield of long-term corporate bonds has dropped dramatically.

… and the relationship between five-year Canada yields and yields on investment-grade FixedResets is also well within what I consider ‘decoupled panic’ territory (chart end-date 2020-2-12):

Click for Big

In addition, I feel that the yield on five-year Canadas is unsustainably low (it should be the inflation rate plus an increment of … 1%? 1.5%? 2.0%?),and a return to sustainable levels is likely over the medium term.

It seems clear that many market players are, wittingly or not, using FixedResets to speculate on future moves in the Canada 5-Year yield. This is excellent news for those who take market action based on fundamentals and the long term characteristics of the market because nobody can consistently time the markets. The speculators will, over the long run and in aggregate, lose money, handing it over to more sober investors.

It should be noted that I have been unable to explain the very strong performance of Floor issues during the 2018-19 downdraft relative to their non-Floor counterparts. See the discussions on PrefBlog at LINK, LINK and LINK.

I believe the bear-market outperformance by the Floor issues is a behavioural phenomenon with very little basis in fundamentals. When interest rates in general move, FixedReset prices should not change much (to a first approximation, for issues priced near par), since in Fixed Income investing it is spreads that are important, not absolute yields. There should be some effect on Floor issues, which should move up slightly in price as yields go down since the ‘option’ to receive the floor rate will become more valuable. Adjustments due to this effect should be fairly small, however – and over the past year issues with a floor, that started the period being expensive, have simply gotten even more expensive, relative to their non-floored counterparts.

And the tricky thing about behavioural models of investing is that they can lose their explanatory power very quickly when an investment fashion shifts, whereas fundamentals will always be effective. Just to give an example from the preferred share market – until the end of 2014, FixedResets were priced relative to each other according to their initial dividend; when the reset of TRP.PR.A shocked a lot of investors, relative pricing became much more dependent upon the Issue Reset Spread, a much more logical and fundamental property. This paradigm shift was discussed extensively in PrefLetter.

FixedReset (Discount) performance on the month was -6.10% vs. PerpetualDiscounts of -1.39% in January; the two classes finally decoupled in mid-November, 2018, after months of moving in lockstep, but it still appears to me that yields available on FixedResets are keeping the yields of PerpetualDiscounts up, even though a consistent valuation based on an expectation of declining interest rates would greatly increase the attractiveness of PerpetualDiscounts (in other words, PerpetualDiscounts are now priced off FixedResets rather than off Long-term Corporates):

Click for Big

Floaters had a poor month, returning -8.54% for February and the figure for the past twelve months has deteriorated to -14.12%. Look at the long-term performance:

Click for Big

Some Assiduous Readers will be interested to observe that the ‘Quantitative Easing’ decline was not initially as bad as the ‘Credit Crunch’ decline, which took the sector down to the point where the 15-year cumulative total return was negative. I wrote about that at the time. but it became worse in August, 2019! On August 30, 2019 the HIMI Floater Index (total return) value was calculated as 1906.6; the index first surpassed this value on 2003-8-13. Thus, cumulative total return (that is, including dividends) was negative over a period of slightly-over sixteen years.

It seems clear that Floaters are used, wittingly or otherwise, as a vehicle for speculation on the policy rate and Canada Prime, while FixedResets are being used as a vehicle for speculation on the five-year Canada rate. In support of this idea, I present an Implied Volatility analysis of the TRP series of FixedResets as of February 28, which is comprised of six issues without a Minimum Rate Guarantee and two issues which do have this feature:

Click for Big

The two issues with floors, TRP.PR.J (+469, minimum 5.50%) and TRP.PR.K (+385, minimum 4.90%) are $2.48 and $4.12 rich, respectively. These figures are a little higher than last month’s figures; note the fact that their floors will not become effective unless five-year Canadas dip below 0.81% and 1.05%, respectively. As we approach these levels (as of month-end) and even pierce them (as of time of writing), we would expect something of an increase in fair value as noted above; but these levels seem elevated!

It will also be noted that the spread of a notional non-callable TRP FixedReset priced at par has increased from 434bp last month to 462bp this month, while GOC-5 has declined from 1.28% to 1.07%.

Lest this be considered a fluke, I also show results for the BAM series of FixedResets, which includes three issues with dividend floors: BAM.PF.H (+417, Minimum 5.00%); BAM.PF.I (+386, Minimum 4.80%); and BAM.PF.J (+310, Minimum 4.75%); these issues are all rich compared to their non-floor siblings, being rich 2.15, 3.32 and 4.19, respectively, much more expensive than last month’s figures of 1.07, 2.59 and 3.39. Note that we would expect all issues to be somewhat expensive according to this analysis, since the guarantee rate is rapidly being approached and (as of month-end) pierced by one of the issues.

Click for Big

It will also be noted that the spread of a notional non-callable BAM FixedReset priced at par has increased from 424bp last month to 462bp this month, while GOC-5 has declined from 1.28% to 1.07%. This is very similar to the effect seen for TRP.

Relative performance during the month was weakly correlated with Issue Reset Spreads for the “Pfd-2 Group” (10%) and the “Pfd-3 Group” (11%) issues:

Click for Big

… and results over the quarter were poorly correlated (below 10%):

Click for Big

As for the future, of course, it’s one thing to say that ‘spreads are unsustainable and so are government yields’ and it’s quite another to forecast just how and when a more economically sustainable environment will take effect. It could be years. The same caution applies for an end to the overpricing of issues with a minimum rate guarantee. There could be a reversal, particularly if either Trump’s international trade policies or the novel coronavirus cause a severe recession or even a depression. And, of course, I could be just plain wrong about the sustainability of the current environment.

On the other hand, I will pass on my observation that international interest in the Canadian preferred share market is increasing, as other Floating Rate indices globally are doing much better. Consider, for example the Solactive Australian Bank Senior Floating Rate Bond Index, which “provides exposure to the largest and most liquid floating rate debt securities issued by selected Australian banks. The index is comprised of investment grade floating rate debt securities denominated in AUD and calculated as a Total Return Index” (LINK although the index constituents currently all have a remaining term of less than five years), and the S&P U.S. Floating Rate Preferred Stock Index.

Yields on preferred shares of all stripes are extremely high compared to those available from other investments of similar quality. As I told John Heinzl in an eMail interview in late November, 2018, the best advice I can offer investors remains Shut up and clip your coupons!

I think that a broad, sustainable rally in FixedResets will require higher five-year Canada yields (or a widespread expectation of them), since paradigm shifts generally require a trigger (a Wile E. Coyote moment, as they say!) … and although I’m sure this will happen eventually, it would be foolish to speculate on just when it will happen.

These calculations were performed assuming constant contemporary GOC-5 and 3-Month Bill rates, as follows:

I will also note that the sustainable yield calculated above is not directly comparable with any yield calculation currently reported by any other preferred share fund as far as I am aware. The Sustainable Yield depends on:
i) Calculating Yield-to-Worst for each instrument and using this yield for reporting purposes;
ii) Using the contemporary value of Five-Year Canadas to estimate dividends after reset for FixedResets. The assumption regarding the five-year Canada rate has become more important as the proportion of low-spread FixedResets in the portfolio has increased.
iii) Making the assumption that deeply discounted NVCC non-compliant issues from banks (and insurers, until November 2019), both Straight and FixedResets will be redeemed at par on their DeemedMaturity date as discussed above.