Friday, 16 December 2016

Market Linked Principal at Risk Notes or Cineplex Convertible Debentures?

I overheard a couple of colleagues at work discussing market linked products on Friday. The conversation was typical of retail logic in my opinion, which somehow escalated market linked GIC's into the realm of shiny happy goodness. I think they are anything but. Retail logic goes something like this: heads, I make the market return, tails I don't lose much. The problem with this line of thinking is that both outcomes depend on a multitude of factors all converging at the same time in order for an investor to achieve the optimal outcome, which in the best case is usually between a 3% or 4% CAGR (for GIC's), and maybe, a 6% to 7% CAGR for linked notes (principal at risk).

The factors are:

1) The type of product (notes or market linked GIC's, principal at risk or principal protected)
2) The payout formula (set by the payor, and adjusted by the payor for each new series, contingent on the payor's expertise )
3) The underlying basket (index, Cdn banks, etc.)
4) Where we are in the overall cycle

Using principal at risk notes for example, from my understanding, an investor can lose principal if the underlying basket declines and stays down through maturity. These are slightly different than straight market linked GIC's, but the concepts are similar, limited upside participation with the added feature of downside protection, but only up to a point.

Here's a concrete example offered by BMO, BMO Autocallable Principal at Risk Notes, Series 27 (I can't believe they've had 26 previous series of these things!), link here

Here are a few excerpts from the prospectus supplement illustrating the payoff matrix (did I mention that these notes are auto-callable, meaning that BMO has the right to call the notes early on each valuation date if the index is > the reference return?)

BMO then explains four payoff scenarios with illustrations. I'm going to reproduce the scenarios illustratively and then provide payoff matrices for all four scenarios based on the prospectus supplement:

And here are the payoff matrices in excel demonstrating CAGR under all four scenarios, plus an added feature demonstrating average return and standard deviation of returns comparing the note scenarios to the index scenarios:

A few observations on this analysis:
  • In scenarios 1 and 2, BMO calls the notes early once the index level exceeds the auto-call level. There is no participation beyond the auto-call date (obviously), but the index investor gets to participate in additional upside and continues to collect a dividend stream beyond the auto-call dates (he/she is not prisoner to BMO's structured products desk). The notes investor has reinvestment risk.
  • In all four scenarios, the index investor collects dividends. The BMO note participant does not, and gives up a 3% yield. This yield softens the blow in scenarios 3 and 4 for the index investor, and enhances the overall return to the investor in scenario 2
  • The CAGR are the returns over the time period invested. Scenario 4 looks tame because it's a CAGR of -9.71% , but the reality is that the index is -40% by the fifth year and the note participant only gets $6K back! Imagine retail Joe going into his branch in the fifth year and being advised that of his $10K (which he thought he was putting into something safe) is now $6K. The response to this comment will be that -40% are highly unlikely, therefore it's worth taking the risk. My response to this response to this comment is, if this is indeed the case, stop being cute and buy the flipping index, not the notes.
  • The average index return is higher and the standard deviation of returns lower for the index vs. the notes under all 4 scenarios.
Update - December 18

I was thinking about the initial observations I made. Statistically, I should be comparing the annual returns on the index to the notes in each year under comparison, not just the CAGR returns at the end of each scenario. In scenarios 1 and 2, the annual returns on the notes are zero when the index is < reference level, but the index shows a negative return. Once the index > reference level, both the notes and the index show positive returns.  Here are the average return and standard deviations for reference under all four scenarios calculated based on annual experience returns during the holding period:

This shows an entirely different picture! The notes by year have a much lower average return and standard deviation (explained by virtue of 0 return in most of the years under study).

An investor must draw their own conclusions from the above.

Onto Cineplex Convertible Debentures

My conversation with my colleagues got me thinking of possible alternatives to the above.

Cineplex has an issue of 4.5% convertible debentures outstanding, maturing December 31, 2018. These are priced to yield around 1.5%, so absolutely nothing to write home about. At today's ask of $105.50, an investor buying these convertibles is locking in 1.5% over two years (not much better than a GIC, but not as bad as a market linked GIC which may pay nothing, or a principal at risk note which could loss of principal);  Here's the math:

I borrowed the calculator from Professor Damodaran, link here

I assumed that Cineplex would be rated close to other BBB issues, yielding 2% in the market for two year maturities. The value of the straight bond with two years to maturity is around $104.85 & the value of the option is $1.46. The conversion ratio is $1,000 / $56 = 17.86 shares for every $1,000 face value.

The optionality here is interesting. The stock has been stuck in a trading range of between $48 & $52 for almost two years, see chart below:

Arguably, the stock is expensive trading at a PE of 24x ttm, but I've seen worse, especially in consideration of the company's virtual monopoly over exhibition across Canada.

Q4 historically has been the strongest quarter, and on an adjusted EBITDA basis, YTD adjusted ebitda is tracking at $167.2M per the company's most recent investor fact sheet.

TTM EBITDA  is around $288M, so at today's closing EV, this would put EV / EBITDA at around 12.6x (on a TTM basis). The company has managed to grow adjusted EBITDA at close to a 10% CAGR since 2006, and has most recently invested in gaming / recreation outside of traditional exhibition, so at today's multiples, I can only assume that investor's expect growth from areas outside of traditional exhibition.

This brings me to complete and utter speculation on my part: What would an acquirer pay for a virtual exhibition monopoly in Canada as the Canadian dollar continues to weaken vs. the USD? Whistler Blackcomb was purchased for close to 16x EV/EBITDA by Vail Resorts, so my guess is, anything's possible. At 15.6x EV/EBITDA, this would put Cineplex at $4.5B, or around $62.50 per share.

I'm speculating on this outcome here only because I find it strange that the stock has basically been pinned to $50 for the last year, and the strike price on the convertibles is less than $6 away. With a cheaper Canadian dollar, a US acquirer could use strong currency to buy a cash flow stream (albeit in Canadian dollars). I also noticed what appeared to be a lot of volume in the convertible debentures themselves today ($3M face value traded, vs. average volume of basically nothing).

At a take out price of $62.50, the convertibles should trade for 62.50 x 17.85 = $111.50 vs. $105.50 currently. The risk is limited to a 1.5% YTM (assuming no default risk), vs. a minimum 7% return if Cineplex gets taken out within two years ($60 profit on the convertibles + interest of $90 over the two years).

Wishful thinking I suppose. Full disclosure, I own the debentures (I do not own principal at risk notes or market linked GIC's)

Update - December 18

Another correction in my initial analysis: The minimum return is not 7% if Cineplex gets taken at $62.50 within two years!  If Cineplex got taken out at $62.50 on Dec 31st, investors would receive $62.50 x 17.85 + accrued interest up to the closing date of the takeout. If the closing date is 6 months, from closing, this works out to 1/2 the annual coupon from date of announcement, so total P/L would be:

$62.50  x 17.85 + (4.5% x 1/2) x $1,000 = $1,116 + 22.50 = $1,138 vs. $1,055 as of Friday's close = $83

($1,138 / $1,055) ^ (2) - 1 = 16%

<I'm going to ignore PV'ing the $1,138 for 6 months)

Another alternative is no takeout but price continues higher between now and December 31, 2018, assume $62.50 again on December 31, 2018:

PV of $62.50 x 17.85 <received on Dec 31st 2018>
PV of $45 received at the end of Dec 31st 2017 <assuming annual payment for simplicity>
PV of $45 received at the end of Dec 31st 2018 <assuming annual payment for simplicity>

I'll use 2% as my discount rate:

($62.50 x 17.85) / (1.02)^(2) + $45 / (1.02) + $45 / (1.02) ^(2)

= $1,073 + $44 + $43.25 = $1,160

vs. $1,055 as of Friday's close:

P/L = $1,160 - $1055 = $105

Over the two years:

($1,160 / $1055) ^(1/2) - 1 = 5%

No comments:

Post a Comment