Statement on Brookfield Asset Management and its family of subsidiaries

Too complex. Bruce Flatt is undoubtedly a genius (in John Malone style), but compare Brookfield with the relative simplicity of Berkshire (despite Berkshire controlling a much larger asset base) and there is such a huge night and day difference. My general issue with large conglomerates with lots of partially owned but controlled subsidiaries is that you run into agency issues with the boards having common members, but having to guess which arm of the company is going to be advantaged (or whether it mostly flows to management).

So a long time ago, I excluded Brookfield and its subsidiaries from being investment candidates – too difficult. I’ve always taken a superficial liking to Brookfield Property Partners (TSX: BPY.UN) (their takeover of General Growth was well timed), but even that entity has a snake’s pit of issues relating to structure. There’s a whole bunch of other REITs trading on the TSX (and indeed the NYSE) that give property exposure with a lot less complexity.

The DREAM preferred share comes to an end

Dream Unlimited (TSX: DRM) had a class of preferred share (TSX: DRM.PR.A), which by virtue of their split from Dundee Corp (TSX: DC.A) had an unusual characteristic – it had a par value of $7.16/share, and was redeemable by the company or the shareholder at any time. In the meantime, it paid out a quarterly coupon of 12.53 cents per share, or 7% on par. The company retained an option to redeem the preferred shares for their own stock at 95% of a prior trading range of DRM stock or cash (or a $2 floor).

This preferred share has been trading for quite some time, and by virtue of DRM’s relatively stable balance sheet, was never in threat of suspending its dividend.

As such, yield-hungry investors could use this preferred share as a cash parking vessel. Even better yet, the dividends that were paid were eligible dividends which would qualify for the dividend tax credit in non-registered accounts. Over time, however, DRM.PR.A became over-utilized and started to trade at a significant premium to par:

The peak was on August 31, 2019 when some poor soul paid $7.56/share for this, or a 40 cent per share premium over par value. This investment would have taken over three quarters of dividends in order to pay itself off. Indeed, at this price the current yield would have been 6.63% for a perpetual investment.

All good things, however, come to an end. Today, DREAM Unlimited announced a substantial issuer bid on their common shares and also the following paragraph:

The Company also intends to redeem all of its outstanding First Preference shares, Series 1. As at November 11, 2019, there were 4,005,729 Preference shares, series 1, issued and outstanding. They may be redeemed at the option of Dream, at any time, at a price of $7.16 per share, plus all accrued or unpaid dividends up to but excluding the redemption date.

As such, one of the best cash parking vessels on the TSX will be off the ledgers. I would expect the shares to crash 4% tomorrow in trading. Fortunately I sold the last of my DRM.PR.A in 2018.

To my knowledge, there is only one other financial instrument that trades in a similar manner, which is Birchcliff Energy’s preferred shares (TSX: BIR.PR.C) which are redeemable by the holder as of June 30, 2020. They also give out a 7% coupon. They are also trading above par. Although the premium is very modest at present, when adjusting for the dividend dates it effectively makes these shares at a tiny discount. There is more balance sheet risk with Birchcliff given its spacing in the natural gas industry, hence why it is not trading wildly above par value at present, in addition to a potential share conversion price at a floor of $2/share (Birchcliff common closed at $2.21 today). I’ve held some of these since February 2016.

Why Marijuana producers will all go to zero

Here’s an interesting press release from a marijuana producer company Alefia (TSX: ALEF, ALEF.DB) – the relevant snippet I’ve quoted below:

2019 OUTDOOR HARVEST HIGHLIGHTS

10,300 kg of dried flower harvested
1,000 kg per acre yield in Zone 1, which was planted in June 2019
$0.08 cash cost per gram to harvest (unaudited)
$0.10 all-in cash cost per gram to harvest, including facility capital costs (five-year amortization) (unaudited)
Cannabinoid content (THC and CBD per gram) of harvested flower was strong, at levels near to the cannabinoid content in identical strains harvested indoor
Quality assurance testing to date is successful, including for microbial content, pesticides and contaminants

“Our inaugural 2019 outdoor harvest was successful due to the commitment and capabilities of our team. I’d like to thank our on-site growers who navigated the challenging environment of starting the cultivation season late into the year and ultimately delivered an excellent harvest that we are measuring in tons,” said SVP of Production Lucas Escott.

The key line here is the $0.10 per gram “all-in” cost per gram to harvest, which bakes into some “half-baked” amortization scheme (pun most definitely intended).

Simple math follows.

1 kilogram (kg) is 1,000 grams.

So this batch of 10,300 kg of dried flower marijuana cost $1.03 million using their numbers.

How much is 10,300 kg?

Apparently 5.2 million Canadians have consumed cannabis over the past three months.

Consumption statistics are not that easy to find, but apparently in 2017 (when Marijuana was still technically illegal in Canada) the annual consumption was about 20 grams per user (with the distribution of consumption highly skewed towards the high frequency user in a typical Pareto distribution – I’d argue that these people are more likely to have their own sources, but let’s ignore this for now). This works out to 104,000 kg per year of consumption (if legal).

The latest statistics (link 1 – October 2018 to June 2019, link 2 – July and August 2019) show the annualized consumption at around 155,000kg (medical and non-medical usage – the pretense of medical usage has gone completely away after legalization).

So let’s say it is 155,000kg and rising. What I find particularly amusing is that if distributed evenly among 37 million Canadians, that’s 4.2 grams for every man, woman and child – apparently this is good for about 10 joints per individual.

Of course, not every Canadian is going to smoke marijuana. Statistics suggest that about 1 in 6 Canadians use it. The addressable population for Cannabis is about 6 million people in Canada – or about 26 grams per person. Assuming this 1 in 6 number is constant (I don’t see how non-smokers can convert into smokers too easily), that works out to about 26 grams per smoker – how much higher is this going to go?

Later in Alefia’s press release, we have the following:

Based on the 2019 results, the Company estimates that it can produce 1,200 kg per acre for a total of 102,000 kg of dried flower in 2020 at its expanded 3.7 million sq. ft. (86 acre) outdoor site, at full capacity. The modest increase in the expected yield per acre for 2020 is due a number of factors which should improve the overall outdoor grow operation, including commencing cultivation several weeks earlier relative to 2019.

So we have one company that is making a claim they can produce 2/3rds Canada’s annual consumption of (legal) marijuana, and implying they can do it for a relatively low cost (10 cents per gram, all-in).

If this is true, then there are a few implications, especially considering that growing marijuana doesn’t appear to involve much in the way of patent-able or proprietary technology (it is a weed, after all!).

One is that there is going to be a massive over-supply of marijuana, and there will be a huge “race to the bottom” effect as price leaders attempt to leap-frog each other to dump their supply into the market place. This is already happening.

Two is that because the marginal cost of production is effectively nothing, that the value chain in producing is going to capture precisely zero profit beyond a cost of capital – and indeed, that will only happen when other inefficient producers get squeezed out of the market due to oversupply – is the company that is able to produce at 6 cents per gram all-in going to have a competitive advantage when all others can do it at 8 cents? Sure it will, but how much value will they be able to extract from that advantage? Not a lot.

Three is that governments are going to make a huge amount of profit on volumes (one of the winners of the value chain, and also having a huge financial incentive to encouraging as much volume as possible to be transacted in the legal market). (CRA Cannabis Excise Duty Information) At the federal layer, the government stands to make a minimum of $1/gram sold, or 10% of the product cost.

Four is that the low price of producing cannabis is going to create its own markets for ultra-low priced product, but this has to happen before it hits the excise tax layer – hence, the production of oils and other cannabis knock-off products that try to find some way to use what is otherwise worthless biological inventory.

Where is the profit going to be in the product? Obviously it is going to be in branding and marketing like most commodity products – tobacco has its Marlboro, and if/when Cannabis has their equivalent, that brand will probably end up making money. These brands take decades to build, sort of like Coca Cola and Pepsi. Until then, good luck – everybody in the marijuana production industry is going to lose.

Subscribing to comments – fixed for now

I am happy to report that the “Subscribe to Comments” feature on the site now appears to be functional. Thank you Marc for reporting this logistical issue. I have taken the liberty to subscribe people that attempted to subscribe for the past couple posts to those posts (so they do not have to manually confirm).

(Update, November 12, 2019: It’s broken again. Grrrrrrrrr……..)

On a side note, it is intriguing how email has become less and less of a reliable utility as the internet has aged, especially with the advent of centralized free mail providers (e.g. Google Mail, Hotmail, etc.) which effectively can shut off most of the internet when they blackball mails from less centralized (and usually compromised) servers. The issue then becomes how these free mail providers make a mint by harvesting the information. If you are a high profile person of interest, using such services is absolutely insane, but it still happens. You hear stories about this all the time (example). Properly securing your email is not as trivial as it may seem – on a shared webhost, they can obviously access your information at will (but at least it is not likely to be retained by Google or Microsoft and the like), so the only real defense is running your own private server and ensuring that sensitive communications are given proper levels of security.

Of course, when there is a need, there is a market. The question is – trust.

Also, on a slightly related note, just imagine if you were some foreign hedge fund that made their specialty living (illegally or otherwise) hacking into publicly traded companies’ financial infrastructure and you got wind of their earnings releases in advance of their release to the market. It doesn’t even have to be full financial statements, it could be a single line item, such as consolidated revenues. Such information would be a cash generating machine, or at the very minimum give you a huge heads-up on whether to long or short. The potential gain for these firms makes it easy to think that it does happen.

Corus Entertainment – quick look

Corus Entertainment (TSX: CJR.B) is a well known company. They have various media assets in Canada on television and radio. At one point I was considering a purchase, but held back because they are highly leveraged and I didn’t have a solid grasp on the risk/reward profile – I suspect they do have competitive advantages but the broadcasting industry is shifting so much (“cord-cutting”, Netflix, etc.) it is difficult to tell whether it is sustainable.

Reading their year-end financial statements, there are a few interesting wrinkles which caught my attention.

0. Fiscal year ends August.

1. They historically have generated a lot of cash. FCF in FY2018 was $344 million, and $307 million in 2019. In 2019, they chipped away $250 million on their debt (which stood at a FY year end balance $1,732 million). Most of this debt was issued to facilitate the acquisition of Shaw Media. $258 million is due November 2021, $869 million on May 2023 and $639 million on May 2024. Clearly they won’t stand much of a chance of paying back the 2023 and 2024 tranches, but one would presume if they rake in $300 million a year in free cash, this won’t be a problem to refinance. The effective rate on the debt is 4.3%.

If for whatever reason this future operating cash flow were to drop at a more accelerated pace, however, the equity is going to suffer badly.

But needless to say if one believes that $300 million is the norm, $1.42/share in cash means the Class B shares trade at a 3.5x multiple…

2. The amount of dividends issued is structured in a peculiar manner:

Corus slashed their dividend 80% last year so they can concentrate their capital on deleveraging. However, the amount of dividends going to non-controlling interests is quite high.

The concept of non-controlling interests is not easy to explain in accounting terms, so I will try here.

Let’s pretend you own a holding company, but this company’s only asset is the ownership of 70% of voting and common shares of an operating entity. If the operating entity makes $100 of income, you consolidate the operating company’s financial statements into your own, but $30 of that income is attributable to non-controlling interests. Shareholders of the holding company effectively only “see” 70% of the operating company’s income.

This is most prominent in a case like Interactive Brokers (Nasdaq: IBKR) where a shareholder of IBKR (76.7 million shares outstanding) owns 18.5% of IBG LLC, which is the entity that actually owns most of the assets. In the 3 months ended June 2019, for example, IBG LLC reported $210 million in net income, but $178 million goes to non-controlling interests (the 81.5% that owns IBG LLC, mainly Thomas Petterfy) while IBKR holders effectively see $32 million. Indeed, IBKR holders will only be able to “cash in” if IBG LLC is generous enough to distribute earnings to it (which they do through a nominal dividend, and control is not an issue because the entity has the same controlling shareholders).

Indeed, when we look at Corus’ income statement, we see that various entities contribute a 13.6% slice on net income:

So one obvious question is the following: what is the agreement governing the non-controlling interest and Corus? It appears that the non-controlling interest has favourable agreements with respect to cash distribution than common shareholders. What entitles these non-controlling interests a $30 million slice of income each year when common shareholders get 56% of the dividends paid by Corus?

I haven’t been able to figure this out.

However, on page 51 of their financial statements, we have the list of subsidiaries, and the non-controlling interest must come from these entities:

I never knew the Food Network and HGTV was so profitable!

What sort of agreements are stripping so much value away from common shareholders? I tried looking into the MD&A and AIF, but couldn’t find any relevant answers.

3. Here is the biggest issue I have with Corus – alignment of interests.

Shaw used to own 38% of the non-voting Class B shares of the company (the only shares which are publicly traded). They dumped this stake for CAD$6.80/share on May 31, 2019. However, the Shaw family trust still owns 85% of the Class A voting shares (3,412,392 shares outstanding) which means they have an economic stake in the company of 1.6% but still total control. Now that Shaw is virtually out of Corus, the incentive structure is completely mis-aligned with common shareholders. Right now the 24 cent/year dividend is the only thing they have going for it. One wonders what sort of value-stripping agreements might take place in the future (similar to my suspicions on the relatively high cash drain coming from non-controlling interests – where is this value going?).

It is one thing if the controlling shareholder has a significant economic interest in the firm they are leading (e.g. one would suspect that Genworth Financial is not going to take actions that will hurt the economics of the underlying Genworth MI entity). It is completely another thing if somebody has control but no economic interest – this sort of alignment asks for common shareholders to get the short end of the stick.

If anybody has answers on this one, or if I’m completely out to lunch, I’d like to know.