Thursday, 14 January 2016

Canadian Preferred Shares: Bottomless Pit or Misunderstood Opportunity?

I guess the best place to start is a visual comparison of two periods in time.  Here's the ishares index:






















I've highlighted two areas of interest which, at first glance, appear comparable, at least to the "technical" trader.  The school of thought is something along the following lines:

"The index has broken support from 2008, therefore it's a sell"

While I have to acknowledge the existence of participant muppetry, it doesn't mean I have to agree with it.

Garth Turner published an interesting piece explaining the space in the attached link.  I'm reproducing excerpts from the piece below (circa July 2015):

"FIRST, WHY THE DECLINE? 

Preferred shares are supposed to be a steady asset class, rarely declining dramatically and providing investors with a consistent return. Unfortunately, this has not been the case in 2015 with the S&P/TSX Preferred Share Index falling by a gut-wrenching 14%. So what has changed? Why has this traditionally defensive asset class behaved with so much volatility?

First, let’s begin with some history. Prior to the credit crisis of 2008–09 Canadian preferred shares were mostly of the perpetual variety. Perpetual preferreds have a fixed, usually very attractive, dividend, but they can also be redeemed by the issuer—the issuer can, at their option, “call” them back from investors at par in exchange for cash. The attractive dividend usually prevents major price declines and the call feature usually prevents major upside price moves. Hence, perpetuals historically had predictable price movements rarely straying far from their redemption, or par, values. However, perpetuals also have one key weakness: no fixed maturity date. This weakness became apparent during the credit crisis when investors fled from perpetuals because there was nothing compelling issuers to ultimately pay investors a set price for their shares. 

<DA: THIS IS OUR FIRST CLUE THAT ALTHOUGH 2015/2016 MAY ECHO 2008 IN TERMS OF INVESTOR CONCERNS REGARDING ISSUER CREDIT RISK, THE TWO PERIODS AREN'T EXACTLY THE SAME>

As confidence in markets eroded, particularly in the financial sector, which is a major issuer of preferred shares, investors, who had no maturity feature to protect against the downside, watched their perpetuals plunge in value. 

In the aftermath of the credit crisis, investors were stuck in a world of rock-bottom interest rates and depressed bond yields. They’d also had enough of perpetual preferreds. They wanted a product that at least offered a quasi-maturity feature and some protection against rising interest rates, as the accepted wisdom following the credit crisis was that interest rates had only one way to go: up. Thus rate-reset preferreds were born. Rate-resets could still be called by the issuer, but they offered investors the opportunity to “reset” their dividend, typically every five years, at a yield equal to the Government of Canada (GoC) 5-year bond plus a pre-determined premium spread. So, for example, if in five years’ time GoC 5-year bond yields moved to 3% and the premium spread on a rate-reset preferred was 200 basis points (2 percentage points), investors had the option to reset the dividend to the equivalent of a 5.0% yield. Investors loved these new products because they ostensibly offered protection against rising interest rates (if bond yields moved higher, investors stood to receive a higher dividend down the road) and, in theory, provided the maturity feature investors desired because the reset spread would incentivize issuers to simply redeem, or call, their preferred shares at par rather than support the new reset yield, which might be above current market rates.

Eventually, the majority of the Canadian preferred share market moved to a rate-reset structure and perpetuals fell out of favour. Currently, the market is roughly two-thirds rate-resets. 

<DA - THIS IS OUR SECOND CLUE THAT THE CURRENT MARKET IS DIFFERENT IN TERMS OF COMPOSITION.  IN 2015/2106, THE INDEX IS 2/3rds RATE RESETS.  THE MARKET IS MAKING NEW LOWS NOW FOR DIFFERENT REASONS VS. 2008>

Unfortunately, of course, interest rates did not move higher, they stayed low and the reset spreads, which were initially modest (100–150 basis points) made redemptions improbable—“resets” were likely to be at yields lower than prevailing market rates, so there was no advantage for issuers to call. In other words, the wave of redemptions (or “maturities”) that investors were hoping for seemed unlikely to occur. The Bank of Canada unexpectedly cutting interest rates in January 2015 and then again in July 2015 has only compounded the problem. Interest rates and bond yields have plunged and preferred share values have fallen almost in lock-step (see chart below). Compounding the problem even more is that preferred shares tend to be widely held by retail investors who, unfortunately, often sell out of fear without fully understanding market conditions. The final difficulty is that the preferred share market is illiquid, which can magnify downward price action, particularly when retail investors are running for the exits."


 So, my overall takeaways are as follows:


  • Index is plagued by the worst features attached to both 2008 and 2015/2016, namely issuer credit concerns and lower resets in face of tumbling GC 5 years (more on this shortly).
  • Space was (and likely still is) perpetuated by retail thirst for yield expecting issues to trade close to par without understanding reset features in context of GC 5 years
  • Initial retail expectation that high enough reset spreads would ensure issuer call issues if GC 5 years rose, ensuring that issues trade close to par (in retrospect, it's interesting that no signal was taken that non-investment grade issuers offering high reset spreads were issuing high reset spreads for a reason).  As GC 5 years have dropped, the exact opposite has happened.  Non investment grade issuers have tapped into a cheap source of funding and have chosen to extend their reset issues, so investors who had initially hoped for issues to get redeemed are stuck with extended non-investment grade issues paying .5% + x bps.  It's no wonder the junk in the index has dropped  so precipitously 
  • New issues have come to market recently with minimum reset floors, which have perpetuated selling on non-floor issues
  • It's interesting that investors are thirsty for imbedded puts with no upside - these new issues are basically collared, investors would rather buy the sure thing with no prospect for capital gain vs. going through the wreckage and sorting out what may be salveagable

Here's the GC 5 year.  Notice the slope of the drop, since 2001. 






Now, I can't write a blog about Canadian preferreds without mentioning James Hymas.  I follow his blog regularly, link here, and I think of him as a rocket scientist in terms of evaluating preferreds. He writes a regular monthly newsletter and runs the Malachite Aggressive Preferred Fund.

Mr. Hymas had the following to say regarding bank and insurance company issued preferreds back in 2011, link here, courtesy of Globe and Mail:


He suggests looking at bank and insurance company perpetual preferred shares. Generally, perpetuals have no fixed redemption date and offer no rate-reset potential. Really, they're a lot like open-ended bonds with no maturity date.
Mr. Hymas argues that most perpetuals issued by banks are a different animal because the 2022 regulatory deadline is almost like a drop-dead date for redemption. In fact, he regards them as being nearly as good as retractable preferred shares, which have a preset date for redemption.
Retractables are considered a desirable kind of preferred share because they offer an escape hatch that perpetuals lack.
There's some uncertainty right now about whether insurance companies will be bound by the same rules as banks on preferred shares. Mr. Hymas thinks they will be, and he therefore suggests insurance company perpetuals as another potential landing spot for people selling bank-issued rate resets.


Mr. Hymas goes into much more detail regarding why he believes the OSFI / NVCC rules will eventually end up getting applied to certain insurance companies, but I don't want to do Mr. Hymas a disservice and let the cat out of the bag - his thesis is mentioned regularly on his blog.  He makes a number of recommendations in his newsletters regarding potential insurance preferrered candidates for consideration, and I'm going to analyze one of his recommendations in my own way as follows.

Here's Intact Financial Series A Preferreds:




















This issue has the following features:

  • Initial dividend of 4.2%, or $1.05 per share
  • Resetting to GC 5 year + 172 bps on Dec 1, 2017 (so just under 2 years to reset at a potential $.55 dividend, almost a 50% reduction)
According to Mr. Hymas, there is a probability that the "OSFI Advisory Treatments  of non-qualifying capital instruments" gets extended to this and other specific insurance issues.  In this case, these issues may have to be redeemed by the issuers at par.  For analytical purposes, Mr. Hymas has imposed a theoretical deemed maturity for this issue of 1/31/2025.  I note that the OSFI rules already extend to banks with non NVCC compliant issues, which explains the differential in price between bank non NVCC issues and insurance non NVCC issues with similar reset spreads.

For my purposes, I wanted to determine if there's a margin of safety in the event that Mr. Hymas is proven right in respect to insurance preferreds.  Here are my results as pertaining to IFC.PR.A:

Take one:


  • Reset dividend at $.555 in 2018 = 2.22% x $25, no change through next reset date
  • Model redemption in Dec 2024 (close enough to Jan 2025)
  • Use a high enough discount rate to account for risk, I used Long gov't + reset spread for 5.72%
  • Overall PV = $19.89 vs. current market of $14.40















Take two:

  • Reset dividend at $.555 in 2018 = 2.22% x $25, no change through next reset date
  • Model redemption in Dec 2024 (close enough to Jan 2025)
  • Use a high enough discount rate to account for risk, I used 5% + 3.5% current seniority spread for 8.5% (=interest equivalent yield  - long corporates as offered by Mr. Hymas )
  • Overall PV = $16.27 vs. current market of $14.40

















Take three - what is the market currently pricing in terms of worst GC 5 year using highest discount rate?  Approximately -1.25% over the next 9 years.
















Concluding thoughts

Maybe my analysis is simplistic, maybe I'm putting too much weight on the possibility of Mr. Hymas being right about NVCC rules getting extended, but I believe there is a margin of safety in buying selective issues with the prospect for capital gain.  I have to weigh the possibility of a negative GC 5 year getting to -1.25% indefinitely vs. buying IFC.PR.A today to yield 4% at $14.40.  I've already made my decision (and I'm taking my lumps), but I'm willing to bet against the consensus here.






5 comments:

  1. I have been buying preferreds every day (including some IFC.PR.A) it seems like! Maybe some people want to get out ahead of what might be another rate cut next week by the BOC but it sure seems like a fat pitch. The hardest part is being patient for most people (including me!).

    ReplyDelete
  2. I've tried to spread it around by buying all of the fixed reset insurance preferreds. Started buying in early December and added some more lately. I'm probably early, but, whatever. Guess we'll see. Heads they are non NVCC compliant, tails I collect a 4% yield here. Definitely taking lumps, and it's not comfortable.

    ReplyDelete
  3. Its an incredible decline that we have seen. People who own the S&P/TSX or S&P 500 think they have it bad but the TXPR was down ~14.5% last year and is down almost 13% already this year! Its possible interest rates will stay this low for years but I'm willing to bet on either higher rates, NVCC or tighter credit spreads. Just seems like a lot of ways to win and I like that.

    ReplyDelete
    Replies
    1. Reminds me of Howard Marks' newsletter on risk vs. liquidity. Risk does not equal beta or volatility. Risk is a pure function of liquidity. When liquidity disappears, you get a glimpse into the risk abyss. This certainly rings true for these preferreds no matter how high a discount rate we use to PV them. We may think there's a theoretical disconnect (and we're probably correct), but nothing we think can stop anyone who hasn't thought any of this through from hitting the bid.

      Delete
  4. I suppose the key to succeeding here is to not lose your mind while everyone else is, and the only way I know how is to limit position sizes so I don't blow up. I'm up to about 10% of assets, and I think that's enough for me. re: non NVCC issues, seems that fixed resets seem to be pricing in > -1% GC 5 yrs at current prices. IFC.PR.A doesn't reset til late 2017, so there's still almost 2 yrs at 4.2% before it resets, and it doesn't matter b/c investors want to own collars on new junk issues. Factor in current spike in credit spreads (+330 bps?) and no one wants to own this stuff (except us idiots). See you on the other side!

    ReplyDelete