Salmon farming: a story of rent and capital

I believe long-term the salmon farming industry is toast, but it exemplifies a great deal of what competition and capitalism do to rents.

TLDR

• Despite being an elongated production cycle commodity, the salmon farming industry is no longer cyclical and generates growing profits. This is caused by environmental limitations to the current production system based on pens located in fjords. The companies that own licenses today enjoy enlarging rents from these licenses.

• However, these rents attract capital to alternative production systems: land-based and off-shore salmon farming.

• I predict that with sufficient capital and scale, these alternative systems will drive the experience curve until they reach a competitive cost with conventional farming, after which rents will be destroyed.

• The more time it takes to enlarge supply, the more rents are generated, and the more capital is attracted to destroy these rents.

• I don’t trade and expect the cycle to take at least 10 years. However, there are trading opportunities: arbitrage between producing countries and license-owners, rents lasting more than expected, shorting the conventionals, and finally shorting the alternatives. I do not explore these in detail but provide some ideas.

Time generates cycles

The salmon revolution started in the 1960s in Norway with the first sea-based farms. Since the 1980s, the industry expanded in Norway, and from the 1990s in Chile. Thanks to scale and productivity enhancements, the cost of producing salmon fell 80% since the 1980s, most of that during the first decade and a half. Between 1980 and 2008, the industry grew at a CAGR of 9%

Salmon is a product that is prone to supply-driven cycles because it has a very extended production period coupled with a very short sales period. Specifically, a salmon takes about three years from egg to harvest. It starts as an egg in a hatchery, lives its first 12 months as a fry or smolt in freshwater land-based farms, and then is moved alive via wellboat to open sea pens in fjords. For the harvest, it has to be moved again alive to a processing facility, where it is gutted, sometimes cut into fillets or steaks, and sent fresh via air freight boxed with ice.

The combination of these two factors (long production, short sales) creates an informational challenge for producers, who need to plan years ahead without knowing what other players will do.

The boom and bust dynamics of the industry for the first 30 years of its existence have led to a pretty concentrated landscape. Today, ten companies control 50% of the market.

Scarcity generates rent

Up to this point, salmon behaved like a regular commodity. However, nature had other plans. Since 2010, the industry has grown more slowly, profit margins have increased, and cycles have ameliorated, now more prone to supply shocks caused by disease rather than gluts caused by oversupply.

The reason for the cycle amelioration is the exhaustion of space to cultivate salmon. Conventional salmon farming requires particular natural conditions. First, the water temperature has to range between 8 and 14 degrees Celsius for the animals to thrive. Second, the pens must be on sea water but protected from ocean currents, so they must be in fjords.

Only a few locations in the world reunite these characteristics: the whole of Norway, the south of Chile, Scotland, Iceland, the Faroe Islands, Ireland, the north of Canada, the south of Tasmania, and the south of New Zealand. From these, only a few have optimal conditions, and some (like Iceland or the Faroe Islands) are very small. Norway concentrates 50% of the world’s market share, and Chile has another 25%.

As production expands and more cages are located close to each other, the risk of epidemics increases. Sea lice, different forms of bacteria, and algae blooms are the enemies of cultivated salmon. This has led to the establishment of license and biomass quota systems. It is unclear how much the industry combated these regulations because existing players are the primary beneficiaries of quotas.

A scarce resource will generate rent; that is Economics 101. In the salmon economy, licenses are the scarce resource, and the global companies owning them have benefitted substantially. That is, I believe, the reason companies like Mowi or SalMar have generated median ROEs close to 20% for the past decade.

Rent attracts onlookers

So far, the current situation is fantastic for the largest salmon producers. However, economics also indicates that competition will try to absorb the rents. If competition does not work, the State will do the job.

Norway, known for its egalitarian policies, has decided on two methods for distributing salmon license rents. First, Norway auctions the licenses and the permits to expand production. At their peak, these licenses cost as much as NOK 170 million (or NOK 215 thousand per ton permit). Today, they make up about 1/3 of the non-current assets of companies like Mowi. Second, this year, Norway decided to apply a resource rent tax of 25% (initially proposed for 45%) to the economic profit generated when the salmon are in the sea.

Second, new methods are being financed and tested worldwide to remove the environmental limitations of salmon production.

The most extended method is land-based cultivation in gigantic tanks. The main challenge for land-based cultivation is the enormous capital cost of building those tanks, given that salmon require space to swim and are large animals. Depending on the estimates, land-based cultivation may require three times more capital than sea-based cultivation. My estimates point to a current difference of $1.5 to $3 per kilogram from these capital costs (depreciation and interest). Many public companies in Norway and the US specifically focus on land-based salmon. Then, there are other, more experimental methods with enclosed cages in lakes and offshore pens. Again, the main challenge is capital.

If salmon supply keeps lagging demand, the differential rents will increase, independently of who gets them (corporations or the State). These rents are almost guaranteed to attract capital. Further, the alternative methods do not need to beat conventional farmers in cost, only beat their rent augmented price.

This differs from selling greenhouse tomatoes or hydroponic lettuce, because salmon is scarce and expensive. I believe that with time and capital, scale and experience will do their job and absolutely decimate the current cost of raising land-based or offshore salmon.

Back to cycles with a vengeance

Unfortunately for the innovators and their investors, the development of alternative methods for cultivating salmon will destroy the profits that those methods are trying to capture.

To begin with, untied from natural (or legal) chains, supply can expand and meet demand, destroying the rents generated by quotas and licenses.

This, in my opinion, will be aggravated by the capital intensity of the alternatives. Capital-intensive methods of production compound the effect of supply-driven cycles because they incentivize to increase volumes at the expense of margins, and to produce below the cost of capital.

Once a producer has a $300 million (today, potentially a billion or more in the future) land-based salmon facility, it wants to spread that depreciation across as many fish as possible to reduce its fixed cost component. Further, because capital cash flows are sunken costs, the producer is happy to keep production going at EBITDA-positive values, further harming overall profitability.

The cycle of the land-based salmon farmers (or their offshore cousins) will be worse than the fjord sea-based salmon farmers and will have two ‘times.’ One is tied to the biology of the salmon (three years of production planning), and the other to capital cycles.

Alternative strategies

This is, in my opinion, the future course of events. I am unsure about timing or players, but I am confident about the trends and incentives involved.

Personally, this blog and my personal investments are oriented by a buy-to-own mentality. This allows me to eliminate one substantial uncertainty factor: other people's opinions. If I own something for the long term, and I don’t care about stock prices, then my only source of risk is operational.

However, if a reader wants to play the expectations game, there are several subtrends that I have not explored in detail but could prove valuable:

• With the recent taxes, plus biomass quotas not growing since 2020, Norway seems poised to some level of stagnation in production. One country with spare licenses (40% of which are fallow) is Chile. Chile is not running at peak production because its costs are higher than Norway’s and because it’s more prone to epidemics. As the supply-demand gap enlarges, Chile producers (or their multinational matrices) could benefit if Norway's prices become less competitive because of taxes.

• Not all of Norway’s producers are equal. Some are more leveraged, less competitive, etc. The new taxes, applied at harvest-level profits, could take them out of business. Prominent players like Mowi could enlarge their market share and purchase profitable licenses (albeit at lower capital returns, given that these licenses already incorporate the rents created by scarcity).

• The alternatives can take a long time to reach competitiveness and scale. From my survey, land-based salmon farm projects from public and private companies are expected to produce about 400 thousand tons by the end of the decade, less than 20% of the current market size. This means that some producers will be able to enjoy rents for a long time. The problem with this speculation is how faster or slower than expected the alternatives develop and when the market expectations change regarding the extinction of rents for sea-based farmers.

• The alternatives will probably follow more hype and bust cycles as the technology rides the experience curve down. For cycle traders, it will be ripe with opportunity.