Description

E-L Financial is an opportunity to be long a ~60% discounted liquid diversified portfolio of SPY, NA, Global, and Value equities, with ~5% CAGR NAV accretion from high capital returns on top. Because ELF's circular ownership is obscured in this company, both the true cheapness, true accretion, and potential imminence of a catalyst is underestimated.  

The holy grail of investing according to modern financial theory is beating passive benchmarks by either 1) a lot on average or 2) in a persistent low risk or risk-free manner. Here’s a way to get both. The catch? Assume you beat passive benchmarks in a boring way (no news flow, not picking any individual businesses, low liquidity), and with a ticker that gets nobody excited at peer stockpicker cocktail parties: did it really happen? 

Thesis recipe:

1. (Very) High NAV discount

2. (Extremely) high conviction in said discount thanks to liquid assets portfolio

3. (Very) High capital returns in the period since CV-19 (i.e. when the intrinsic value discount blew out)

4. ELF has a myriad of circular holdings, but because this is quite obscured in the last decades of filings, the prospective return drivers are underestimated:

   1. True NAV discount is not 52% but 63% because of circular holdings
   2. True accretion of special capital returns is even higher 
   3. True control of main shareholder is much closer to "end game" levels because of circular holdings 

While a large derating has occurred mostly since 2019 and during CV-19 lockdowns (the stock traded around ~1.0X BV, and higher since the 2000s), ELF NAV has CAGR’d 12.5% since 1969 (that’s more than half a century!), and 11.2% in the last 10 years. I expect faster compounding ahead.

On a high level, E-L Financial is a holdco with 70% of NAV in a diversified stock portfolio and 30% in Canadian life insurer “Empire Life”. 

My SOTP, and the pie chart (using gross asset values) show this a bit more granular:

The stock portfolio (~70% of GAV) consists of 

• Corporate investments, of which

  • 900MCAD in Vanguard S&P500 (VOO). ELF shows up as a top owner of VOO with 1.65M shares. This alone is more than 300 CAD p.s. in value 

  • 2.8BCAD in direct & indirect N-American & global equities (think Neuberger Berman, Burgundy AM), and owning own shares via private co’s (see circular holdings later on)

• United Corporations Limited (54.9% owned & consolidated)

  • a stake in one of the world’s oldest closed-end funds worth 660MCAD at market price in this SOTP (or 1.1BCAD at underlying NAV)

• Economic Investment Trust (24.7% owned & equity method accounting)

  • Another very old closed-end fund, worth 180MCAD at market price (or 250M at NAV) 

• Direct stake of 37.2% in Algoma Central worth 215MCAD at market price (equity method accounting). This shipping company in a “profitable niche” the Great lakes was recently detailed in an Alluvial Capital blog post 

Side note: in reality ELF owns more than 50% of EVT and (interestingly) Algoma central via its several stakes in its private companies (these are accounted for in the value of Corporate investments).

Empire Life (~30% of GAV): Based on general fund and segregated fund assets, Empire life is a top 10 life insurer in Canada, with a 6% market share in its main markets. ROE seems average, and we won’t dig deeper into insurance.

The company reported a year-end NAV of 1786 CAD per share. 

My current SOTP has ELF at 1764 CAD per share. The difference because of 

1. YTD performance of all investments (+360MCAD impact) 

2. UNC stake at market price instead of consolidated, and the NAV discount is currently ~40% (-440MCAD impact)

Remark: if the company were to report NAV today (so not incorporating point 2), I believe IFRS NAV would be 1888 CAD p.s.

The NAV discount is hence 52% on a 'run-rate' IFRS basis (using look through NAV of consolidated UNC), or 49% by my numbers (UNC at market price).

Management & capital returns

The Jackman family controls E-L Financial. On Bloomberg it is easy to underestimate its control: the complexity adds difficult to know the exact percentage, but the last proxy statement with this disclosure (2020) stated:

Companies, in which The Hon. H.N.R. Jackman has an indirect interest, control, in the aggregate, 2,945,765 or 77.15% of the outstanding Common Shares of E-L Financial. Although The Hon. H.N.R. Jackman is associated with some of these companies, he does not control any of them and, accordingly, does not have beneficial ownership of the Common Shares of E-L Financial held by them. 

Management compensation vs NAV (or mgmt ownership) is a drop in the ocean here (e.g. 1M salary for Mr. Jackman). 

What is a management sitting on 7B of liquid assets supposed to do? Focus on sound capital allocation decisions! Buybacks & dividends when the NAV discount blows up. This is exactly what management has done. ELF has conducted a flurry of NCIB, large SIBs, divi’s and special divi’s since CV-19 has left the stock severely undervalued.

The share count has dwindled since 2020 through buybacks. Today this number is 2946/3557= 83% and free float has shrunk by 30%. For the purpose of a buy-out offer, I understand the shares Mr. Jackman controls through the many circular holdings are not allowed to vote in a minority vote. However, this gives a sense how close we could be to a buyout.

The cherry on top: ELF has dialled up the capital returns of its controlled CEF entities (EVT, UNC) as well!

Below an aggregate view of NCIBs, SIBs, dividends and special dividends:

Focusing on ELF only, the average capital return since CV-19 has been 258MCAD p.a., or on average 9% p.a. capital returns.

Circular ownership

E-L Financial owns its own shares via intermediate holdco’s.

ELF and its partly owned entities UNC, EVT all own ELF directly and via private companies, the most important being:

1. Ecando

2. Dominion & Anglo “D&A” 

3. Canadian & Foreign Securities “CFS”

I will mention a few important details:

1. While all these vehicles own ELF, they also own stakes in each other
2. EVT (the publicly traded CEF) owns ELF directly (this stake will show up as 'Burgundy Asset Management' in ELF's ownership filings, EVT's previous asset manager), but also via Ecando. This can be found in EVT's annual report
3. Importantly, these intermediate links in the ownership chain also own more Algoma Steel and EVT, rendering ELF in control of more than 50% of both vehicles

Some of these three own a stake in each other. Only old accounts are available for D&A. However, it is not necessary to weed out the impossible circular structure to get to a quite accurate estimate on self-ownership.

As the company has never sold its own shares but only accumulated, it’s safe to say it owns at least 708k shares, as the old Canadian IFRS ‘shares outstanding’ calculation in 2011 suggests (hat tip to SafetyinNumbers  at Corner of Berkshire & Fairfax). We use 

Another way to estimate the quantum of self-ownership is to triangulate from the annual report and other public holdings of ELF.

We will estimate a higher boundary.

1. ELF self-ownership via corporate holdings

It’s interesting to scrutinise note 4 “Investments - Corporate”, including “the investments held at the corporate level of E-L Financial and the investment holdings of its subsidiary, United Corporations Limited”. Remark: this financial note excludes holdings in equity accounted holding EVT.

The level 3 assets subsection has this intriguing footnote:

The footnote is telling us two important things:

• ELF owns a lot of illiquid ‘private companies’ holding (liquid) common stocks as their main purpose, and these
• are being valued at the underlying share prices

However, we think the Canadian Level III assets item is interesting, at 836MCAD. The correct way to think about this number, is to find a higher boundary of ELF’s self-ownership through private companies.

Yet, ELF owns more of Algoma Central via private company ‘Amogla Holdings’ in its corporate portfolio as well:

Correcting for this Canadian Level 3 asset, we get 700MCAD of Canadian Level 3 assets that are potentially ELF shares. ELF traded at 900 CAD at YE22, yielding a suggested higher boundary of 778M shares self-ownership.

But as we will see next, ELF also owns itself via its stake in EVT (not accounted for above).

2. ELF self-ownership via stake in EVT

EVT owns approx. 461 k ELF shs

On an economic look-through basis, ELF owns 24.7% of EVT, hence 114M ELF shs.

The total higher boundary of ELF self-ownership is hence 778M + 114M = 892M shs.

Conclusion

In my opinion (and while I have done this type of exercise before in previous write-ups), the structure of self-ownership here seems impossible to pin down exactly. Doing this (if at all possible) would perhaps budge the true NAV discount by a percentage point or two but will triple this thesis' length. I'd be interested to move this to the comments section.

At the mid-point between our lower (708k) and higher (892k) boundary, we get 800k self-owned shares.

True NAV after accounting for circular ownership

It is important to realise all ELF self-owned shares are accounted for at market price.  

To adjust for true value

1. We deduct 800k * 900 CAD per share @31/12  = 720 MCAD from the NAV

2. Economic shares outstanding becomes 3557 - 800 = 2757 shs out

The new SOTP value becomes

1. Adjusting the company value per share:       2435 CAD per share

2. Adjusting my own SOTP (UNC marked down at market price):   2276 CAD per share

With the share price at 900, the NAV discount becomes 63% using the IFRS logic (looking through UNC NAV), or 61% using my SOTP logic.

Accretion

Accretion = NAV discount x (capital return as % of valuation)

Annual NAV accretion (on top of the Vanguard-like return) is hence 9% p.a. capital return x 63% = 5.7% if the (special) buybacks / (special) dividend continue. That’s almost double what indexers expect to earn on average.

Does VIC really need/want a catalyst (such as a privatisation) with that type of accretion?

Catalyst

Privatisation: as discussed, while this catalyst's imminence is invisible by looking at Bloomberg, the true float has shrunk to only 17% of shs out (after accounting for all circular ownership stakes). E-L Financial owns a portfolio of ultra liquid assets, so a privatisation would be easy to conduct, and at a 60% NAV discount a starting point, highly attractive for the main shareholder. 

Maybe the privatisation is done in slow motion: special dividends were roughly half of the torrent of capital returns in the last 3 years, suggesting the main shareholder is just shareholder friendly and ELF is not going anywhere. 

Lastly (and much like Fairfax Financial’s share swap), as ELF’s sizable circular holding is marked to market, this stock’s main price driver (and buyout KPI), i.e. NAV, will go up by 1.1% as the price rises 10%.