Description

Elevator pitch: From time to time, I come across a stock that makes me think: "this business shouldn't exist as a standalone company". Torex is one of these companies. Before insulting me with things like "gold bug", "barbarous" and/or "Neanderthal" - three words which I probably deserve - focus on this pariah security for a second. Like it or not, there's a market for such a useless piece of metal as (for some people) gold is. This company mines and sells that metal. It has no debt, free cash flow yield is in the double digits (at $1,500/oz gold price and with a 21 years life of mine) and it's worth much more money for a geographically diversified gold major than what Mr Market is saying today.

Torex is a mid-sized gold producer operating in Mexico. The shares are listed in Canada (TXG) and the US (TORXF). It reports in USD. A brief history of Torex can be found in the Appendix of this writeup.

The investment case for Torex is pretty straightforward:

- High-quality assets.

- Healthy balance sheet.

- Cheap stock price.

The reasons why the stock is cheap are clearly identifiable:

- Single-mine company.

- Location of the mine perceived as risky by stock market investors.

- Stock market places zero-confidence on Torex proprietary mining technology.

- Low insider-ownership + diversified shareholder base + high FCF = recipe for capital allocation mistakes.

These ingredients make a strong case for Torex being acquired by a gold major in the near future. This doesn’t mean that we can’t make money here by owning the asset. Torex is a clear laggard of the recent gold price rally, so there’s catch-up potential and margin of safety is much bigger than in richly-priced gold stocks. But, all things being equal, our return (and my sleep) worsens every day that goes by and it’s not acquired.

This writeup is structured in four segments:

1. Asset base.

2. Balance sheet.

3. Valuation.

4. Reasons why the stock is cheap.

1. Asset base

Jurisdiction

If the real estate motto is “location, location, location”, the gold mining motto should be “jurisdiction, jurisdiction, jurisdiction”. NAFTA + Australia, Peru and Chile are good places to own a gold mine. As of today, Torex assets (two mining complexes, El Limón and Media Luna) are in Mexico. Potential acquisitions that Torex might do in the future will be restricted to North America according to the CEO.

Size

You just need the thing to be large enough to matter. It’s not just that having a large gold deposit is important to secure at least a handful of years of gold production, but operating leverage is meaningful enough to need at least a couple of million ounces of resources to justify the upfront capex investment and to be able to feed the processing mill. Either that or you buy existing infrastructure (gold mines that used to be in production in the past but had to close due to mining costs exceeding the gold price back then).

Torex fits the size bill with 4.9MOz in the Measured and Indicated Resources category and an extra 3.3MOz in the Inferred category. The average size of new gold discoveries over the last decade was 5.3MOz. Having enough size is especially relevant when investing in single-asset mining companies as it puts a “valuation floor” on the stock price (provided that the balance sheet is healthy - more on that later in the writeup).

Mark Bristow considers an asset to be “Tier 1” if it can produce >0.5MOz annually over at least 10 years. Torex is close to this feat as it will comfortably do 300-450KOz of gold per year over the next 21 years with the current resource base (which will be easily increased further going forward as I explain below).

Low operating costs (adjusted by operational complexity)

Similar to Munger’s willingness to pay a fair price for a good business rather than a cheap price for a so-so one, better to accept a few hundred dollars an ounce of higher operating costs for a low risk operation (open pit, well-mapped ore body and a simple, non-acidic metallurgy) than seek for low AISC deposits at all costs.

Underground mines are just more risky to operate than open pits as more things can go wrong below the Earth crust than above it. Similarly, it’s probably a better idea to accept higher AISC for an old mine that’s being reopened and where detailed maps are already in place than working on a greenfield project.

In most cases, the largest determinant of operating costs is the grade of the deposit, measured in grams of gold per tonne of rock mined. Open pit projects with at least 1g/t of gold are usually good assets to own. Provided that the metallurgy of the deposit is not very complex, these mines tend to be well placed in the gold industry cost curve (think first or second quartile).

For underground mining, grade needs to be higher to justify the extra infrastructure cost. As opposed to open pit deposits, it’s much harder to give a particular number for what’s a good and a bad grade, as the depth of each deposit is unique. Having said that, underground mining below 2-3 g/t is rarely a good idea. As a rule of thumb, the deeper the asset, the higher the grade needed to make the project economically viable.

Torex AISC are low (650-750$/oz on a Life of Mine basis) relative to the global average ($900-1,000/oz), sitting comfortably in the first quartile of gold producers. More relevantly, the assets are a mix of open pit (El Limón Guajes), and underground (Sub-Sill, El Limón Deep and Media Luna).

Interestingly, although Media Luna is an underground project, it’s not very deep relative to other underground deposits. The ore body is about 600-700m below the surface. To put this into context, the deepest gold mines in the World are 3-4km below the Earth’s crust (https://www.mining-technology.com/features/feature-top-ten-deepest-mines-world-south-africa/).

An ex-Torex geologist whom I spoke to pointed out that Media Luna should be thought of more as a deep open pit rather than as an underground mine. The grade of Torex geological assets is significantly above industry standards (1.5g/t on average vs 4.05g/t for Torex).

Both Torex assets (El Limón Guajes and Media Luna) are fairly easy to mine as they are skarn deposits. A skarn deposit has a sandwich-like structure where you have a clear differentiation between layers where there’s gold and layers where there isn’t. This reduces the “surprise factor” that’s typical from more complex deposits where gold is all over the place, making skarn deposits more efficient to mine.

Torex is currently pouring gold from El Limón (Guajes, Sub-Sill and El Limón Deep). Media Luna will ramp-up production in 2024. The Mexican government already permitted exploration work in the area. Given that the asset is underground and that the processing plant that will be used is the same as that being used for El Limón complex, the risk of not getting environmental approval to operate the project is very low. Moreover, Torex assets are located in the renowned Guerrero Gold Belt, where Agnico Eagle and Equinox Gold are also present.

Exploration potential

There’s an interesting company called Evolution Mining, listed in Australia (ASX:EVN). Its business model is akin to the feeding pattern of a street dog - they buy the assets that the gold majors don’t want anymore. The reason why both EVN and the seller is happy with the transaction is that EVN overheads are lower and its operational focus is higher (hard to justify for a major to dedicate a meaningful amount of human resources to an expiring mine). That makes the mine more profitable to operate for EVN than for the gold major.

I rate EVN management highly (and so do other gold industry people I’ve spoken to during my research). They look for margin of safety in the assets they acquire via exploration potential rather than a low multiple on existing resources (or cash flow). According to their experience, gold mining is just too unpredictable to rely on financial models so you better buy an asset for a fair price and seek for meaningful expansion potential of the resource base that’s beyond what the seller of the asset has documented for the project.

Torex scores highly in this aspect. I had an interesting conversation with a former Torex exploration geologist who spent 7 years working on the assets. He explained that the exploration upside potential for Torex comes from:

a) Media Luna will triple its resource base. The exploration team at Media Luna saw its budget cut by the management team to focus capital expenditures on the ramp-up of El Limón. This cut should be seen in the broader context of exploration budget cuts that we’ve seen in the industry over the last few years, especially after gold fell by 15% in price from the 2016 peak of $1,300/oz. The current resource base of Media Luna is just below 4MOz (including inferred ounces).

b) Only 25% of the land package has been explored. This is particularly interesting as the existing portfolio of assets are skarn deposits. These “sandwich-like” deposits don’t occur in isolation in nature but close to a porphyry deposit. Porphyry deposits are shaped as a big cone. Skarn deposits are the result of this big cone pushing against the Earth crust and leaving traces of its metallic content. There’s absolutely no guarantee that Torex will find the porphyry in the land package that it owns, but the geologist I spoke to showed almost certainty about it. To put this in perspective, the picture below shows the existing asset base (ELG + Media Luna) in the context of the whole land package. The areas highlighted in pink in slide 9 (https://www.torexgold.com/assets/docs/presentations/Corporate%20Presentation%20-%20March%202020.pdf) are the result of the magnetic survey conducted in the area. It could be gold, it could be copper, it could be a worthless piece of metal. But given that I’m not banking any of these into my valuation of the company, it’s a nice situation to be in.

2. Balance sheet

Torex gross debt amounted to $155.2m as of the end of Q1 2020. Decommissioning liabilities are worth $19.2m and cash on hand was $135.7m. Let’s say that gold goes to $1,300/oz (-25% vs current price) and stays there forever. In that scenario, Torex will generate over $675m in cash from operations over the next 3 years, more than enough to cover the $610m in capex that Torex will need to to fund the ramp-up of Media Luna. If gold stays at $1,700/oz, then Torex will generate almost 40% of its current market cap in FCF over the next three years.

3. Valuation

Given the long life of mine (21 years excluding any of the two sources of resource expansion mentioned above) and the fact that the company has been pouring and selling gold since 2016, Torex can be valued as a going concern business (with 0% terminal growth). The two key financial drivers of Torex valuation are the gold price and the USDMXN rate (as 50% of the cost base is denominated in MXN). At the current combination of $1,700/oz and 22.2 USDMXN, Torex discounted cash flows at 7% WACC are worth C$40 per share, which is more than double the current price.

A more simple way of looking at the valuation of this company is to look at FCF yield on a “normalised” basis for both gold and the USDMXN rate. With $1,500/oz gold and 20 USDMXN, Torex yields over 7% on average over the next three years. Very relevantly, this FCF estimate includes Media Luna upfront spend. To give a sense of the impact of this factor, I expect capex to peak at $277m by 2023 and then fall to $100m onwards. $100m is conservative when compared to the company guidance of $85m per year of sustaining capex.

To add further to the margin of safety of Torex valuation, all these numbers assume that Media Luna is developed using traditional methods. But this is not the plan. Media Luna will be developed using Torex in-house technology for underground mining (Muckahi). The difference in project economics is substantial - management expects to reduce upfront capex and life of mine operating costs by 30%. To put this into the context of profitability, Media Luna Preliminary Economic Assessment guides for an increase of IRR at the project level from 27% to 46%.

4. Reasons why the stock is cheap

Single-mine company

The stock market doesn’t like geographical concentration of mining assets. Eggs and baskets tell that it’s more probable that something goes wrong on the operational side when you only have one mine than when you have a bunch of them.

Mitigants: there’s no mitigant to this risk for obvious reasons. Torex just needs to be acquired (preferably) or start an acquisition spree (more on this below). Hence we shouldn’t expect that Torex ever trades at what a conservative DCF says it should but always stay below that (think 10-20% discount to NAV). At the current price of C$17/sh, Torex is trading at 40% discount to my DCF-based NAV $1,500/oz. At $1,700/oz, the discount is almost 60%.

Location of the mine perceived as risky by stock market investors

Guerrero is one of the poorest states in Mexico. And it’s also among the most violent ones. Famous for its beaches (Acapulco), there’s not much to do as a local aside from working for a drug cartel, for a hotel company or for a mine. The fact that the local population is poorly educated compounds the problem. Torex in particular suffered a mine blockade in 2017, which doesn’t help either.

After speaking with several people about this issue, I’ve come to the conclusion that is very possible that we see more disruption going forward. There are two potential sources of conflict: an incident with drug dealers that surround the area or a new incident with local communities similar to what happened in 2017.

Mitigants: The CEO is well aware of this issue and is risk averse. This means that the mine has multiple entry points and, more relevantly, that Torex always operates with a large stockpile of resources that’s waiting to be processed. Of course it’s not an efficient inventory management policy. Agnico Eagle and Equinox Gold own gold properties in Guerrero too. This fact, alongside with my channel checks, provides me with enough comfort to believe that we could see a bid for Torex in the near term. Simply put, Torex shouldn’t exist as a stand alone company but should form part of the portfolio of assets of a major, which can diversify the operational risks associated with Guerrero and access a much lower cost of capital as a result. My view is that this should happen sooner than later. It was risky to bid for Torex before 2016 as the company was in the middle of El Limón Guajes ramp-up. Plus the company was indebted back then. A similar situation will occur in 2024, when the company ramps up Media Luna.

Stock market places zero-confidence on Torex proprietary mining technology

“Mining is really a logistics business, a material handling business. We do drilling, blasting, and ground support up front, but everything else is moving stuff.” - Fred Stanford, former CEO and current Exec. Chairman of Torex

Fred Stanford has been described to me as “a mining geek”. He spent 30 years trying to improve underground mining techniques. In 2015, he invented a new mining system which he called Muckahi (a word he used to motivate miners back when he started in the field). The system is being tested for the first time at El Limón Deep this year. There were some positive initial results in mid 2019 but the system is still an untested concept.

This article describes Muckahi in detail: https://magazine.cim.org/en/projects/big-dream-narrow-focus-en/. In a nutshell, Muckahi is about substituting diesel-fueled trucks with electric conveyor belts anchored to the ceiling of the underground tunnels excavated to mine the asset. The technology is new to the gold industry but it grabs key components from other mining industries, particularly from the coal mining process. Its key positives are:

a) Lower upfront capex and quicker ramp-up. Up to 30% reduction according to management. The use of conveyor belts allows for steeper tunnels (10 degrees in a normal mine, 30 degrees using Muckahi). This is due to the fact that trucks would suffer too much mechanically going up and down a 30 degrees tunnel day in and day out. Plus diesel consumption would be sky-high. The steeper the tunnel, the shorter in length. As the required infrastructure is lower, it takes less time to start an underground mine using Muckahi. According to management, ramp-up time can be reduced by up to 80% using this method.

b) Lower operating costs. Up to 30% reduction vs conventional method according to management. Diesel is more expensive than hooking up a conveyor belt from the power grid. Plus trucks are mechanically more fragile, which means lower down times with conveyors. Additionally, the absence of trucks makes the atmosphere much less contaminated, so the power spent in ventilation is lower. Finally, the inclination of the conveyor belts allow for the generation of excess electricity according to Stanford.

c) Much cleaner. Stanford claims that Muckahi can reduce greenhouse gas emissions by 95% relative to conventional methods. The sources of reduction are clear (no trucks, less ventilation).

I spoke with a Mexican operations engineer now working for Pan American Silver in Mexico and with familiarity of Torex assets and Muckahi in particular. In his view, the use of that technology at Pan American Silver mines in Mexico would really move the needle. And he wasn’t particularly concerned with the potential extra risk of the new method. Although the machinery is more complex than trucks, it takes significantly less tunnels. So it’s fair to at least give some credit to management claims that actually it's a lower risk to develop Media Luna under Muckahi than via traditional methods.

It takes a glance at Torex stock price to see that this method is completely out of the valuation. One caveat is that Muckahi cannot be applied to every underground gold mine on the planet - the geology of the ore body needs to be such that the rocks can be broken down into fine grains in order to then put it on the conveyors. Another way to look at this is that the market doesn’t trust the use of a new mining method in an asset (Media Luna) that needs to take the lead of El Limón Guajes from 2024 onwards. The company is trying to either acquire a small mine in the US that’s well suited for Muckahi or license the use of the technology to an existing miner which can use it. I wouldn’t be too enthusiastic about this latter possibility because the IP rights of the technology are owned by Stanford for any other use that doesn’t compete with Torex business. Hence hard to underwrite that if a licensing agreement comes up, Torex will make money out of it.

Mitigants: Neither me nor the stock market is ascribing any value to this technology, so any surprise should be an upside one rather than a catastrophe.

Low insider-ownership + diversified shareholder base + high FCF = recipe for capital allocation mistakes

Torex Board owns 0.5% of the company, which I find very disappointing (and so does the stock market). Marlin Sams Fund (a NY-based hedge fund that owns 4.1% of Torex shares) wrote a letter to the Board in 2019 which can be found here: https://www.businesswire.com/news/home/20191007005727/en/Marlin-Sams-Fund-Releases-Letter-Board-Directors. Although the acquisition of TMAC didn’t go through (the Chinese miner Shandong Gold ended up buying a stake instead), the fact that insider ownership is so low at a point in time when the company will have so much cash to allocate is clearly a negative.

The shareholder base is so diversified (largest holder is VanEck ETF with 10.3% of the company, then Blackrock - 8.8% - and Sprott - 6.1%) that there’s no proper control of governance. On the other hand, it shouldn’t be too problematic for an industrial player to acquire Torex without having to fight too fiercely on the price offered.

Mitigants: Torex governance isn't really that bad relative to Canadian standards, which to be fair are really bad compared elsewhere. Hoping for Torex to be acquired is the main strategy here (with all the caveats attached to mixing the words “hope” and “strategy” in the same sentence). For what is worth (probably little given the small insider ownership), I’m pretty impressed with the background of the new CEO. She joined Torex in October 2018 from Vale, where she was Director of Business Strategy for Canada. Fred Stanford knew her from Vale. I found her mindset to be risk-averse and detail oriented, two features that are critical to operate a mine in Guerrero state. Her trajectory is interesting too (lawyer by training, ended up in operations because she loved the practical aspect of mining).

Conclusion

At $1,500/oz and 22 USDMXN, the stock comfortably yields circa 15% stabilised FCF (after normalising for Media Luna upfront capex and subsequent ramp-up). These numbers exclude further exploration potential in the property and don’t consider the 30% cost and capex reduction that Muckahi would bring. And, of course, don’t consider the current gold price of $1,700/oz.

With Agnico Eagle and Equinox present in the same area and channel checks confirming that Guerrero isn’t really that bad relative to other Mexican states, plus a widely diversified shareholder base and low insider ownership… I would be knocking at the CEOs door on a daily basis to bid for this company if I worked at the corporate development department of a gold major.

Appendix I: Brief history of the company

1998: A JV formed between Miranda Mining and Teck Resources won the government-sponsored auction to operate Morelos Norte (current Torex assets).

2000: El Limón Guajes and Media Luna deposits were discovered by the JV partners. Exploration and metallurgical tests of these assets continued until 2009.

2009-2010

- Glelchen Resources acquired 100% of the property from Teck Resources and Goldcorp (which had bought Morelos Norte stake from Miranda Mining previously) and changed its name to Torex Gold.

- Fred Stanford came on board and took the role of CEO of the company.

2016: Production started on March 30, ahead of schedule and spending less capex than initially budgeted.

2017-2018

- A group of 20 people blocked the mine for 3 months (November 2017 - January 2018), preventing normal production. The reason for the blockade was a dispute between the union that legally represents Torex workers (CTM) and the union that pretended to represent them (Los Mineros).

- Unfortunately, three Torex employees were killed and seven were kidnapped by representatives of Los Mineros.

- By the end of January 2018, the Governor of Guerrero State lifted the blockade of the mine and operations resumed at the mine. By April of 2018, Los Mineros union withdrew its demands. No other social issues have been reported in the property since then.

2019: the company announced that CEO Fred Stanford will be moving to the Executive Chairman position by June 2020. The current COO, Jody Kuzenko, will take the CEO role that same date.

Appendix II: Supply perspective for the gold industry

“Most investors spend the bulk of their time trying to forecast future demand for the companies they follow (...). Capital cycle analysis, however, focuses on supply rather than demand (...). Because most investors (and corporate managers) spend more of their time thinking about demand conditions in an industry than changing supply, stock prices often fail to anticipate negative supply shocks” - Edward Chancellor, Capital Returns

Contrary to popular belief, supply does matter for the price of gold given that 70% of annual demand is satisfied with mining supply (30% recycling). Comparing the last 10 years of gold exploration with the previous 20:

- Yearly exploration budgets have risen by a factor of 4.0x.

- It’s 15.7x more expensive to find an ounce of gold than it used to be in the past.

- Only 20% of the gold produced in the 2009-2018 period has been replaced (vs 97% in the 90s-00s).

- New discoveries are 30% smaller. A smaller mine is less economical to develop than a bigger one.

- The average grade of new discoveries (measured as grams of gold per tonne of dirt) has remained constant, which compounds the “ever smaller new mines” problem 

- It takes 2x the time (20 years) to get the permits and develop a mine once it has been found.

At $1,700/oz, the current price is at least half what it should be to incentivise gold mine exploration and development. Consider the following hypothetical journey that a gold miner would need to pursue to start a mine (rough estimates using the above table and industry sources):

- Go through a 20-years journey of exploration and development of a 5.3Moz mine that can produce about 300Koz per year over 17.6 years since operations start at the site.

- Spend about:

* $231.5/oz in exploration.

* $100/oz in development expenses (mostly permitting) and pre-production capital expenditures.

* $908/oz in operating costs annually including maintenance capex (S&P Global Marketing Intelligence).

- Hope for the gold price to be high enough to make a decent ROI, let’s say 15% driven by:

* A high degree of uncertainty involved in the exploration and permitting process 

* The material operational risks involved once the mine goes into production

Putting aside for a sec the well-known modelling principle of “garbage-in, garbage-out”, a simple DCF considering the above inputs gives an incentive price of $3,500 to $8,500 an ounce, depending on the discount rate that we want to plug into the model (10-15%).

The point I'm trying to make by doing this silly exercise is that while the bulk of investors focus on Jay Powell to predict the price of gold, the industrial perspective is pretty much ignored and this leads to a constant fear of gold falling spectaculary (think $1,000/oz or less). I think that mindset is wrong. For the first time in decades, we have a situation were supply scarcity in a rising exploration budgets and widespread high-grading paradigm puts a solid floor on the price of gold.

But given that the macro arguments in favour and against gold have a higher sex appeal, consensus misses the industrial picture. As long as we don't decide to destroy our oceans (https://www.theatlantic.com/magazine/archive/2020/01/20000-feet-under-the-sea/603040/) - low probability event according to my checks with miners given the significant environmental impact, gold miners should be fine when it comes to assessing the downside risk of the price of what they sell.

Catalyst

Free cash flow

Acquisition