Milking it: what is the cost of another billion litres?

Last year, Irish dairy farmers produced a record 8bn litres of milk – 7% more than in 2018 and almost 50% or 2.5bn litres more since the lifting of milk quotas in 2015. This strong performance was a result of both rising dairy herd numbers and improvements in productivity.

The dairy herd has expanded 40% to approximately 1.5m head since 2015. Improvements in productivity came from higher returns in milk solids primarily as a consequence of investment in breeding.

This has seen a surge in exports, which are up €1.5bn since 2014 and reached a record €4.4bn last year. It has been seen as a true industry success story. And farmers are benefiting also, with average dairy farm incomes now in excess of €61,000 according to Teagasc’s National Farm Survey last year. Ireland’s 18,000 dairy farmers received milk cheques totalling €2.6bn last year. To put this in context, direct payments to all Irish farmers amounted to €1.8bn last year. In order to process this extra milk, some €2.5bn has been invested while about €1bn has been invested at farm level.

Further expansion

Processors are now asking suppliers their expansion plans in the coming years. Despite the already impressive 50% growth, there seems to still be an ambition among a lot of dairy farmers to grow further – albeit at a lower rate.

To date, very few have not grown, with most growth coming from almost every farm growing a little. For example the average cow herd has gone from less than 70 cows to more than 90 cows in the space of five years. Some models are now suggesting that because many milk suppliers have expanded to their full capacity in terms of land and labour that fewer milk suppliers will substantially expand into the future. No doubt environmental concerns (covered extensively in other parts of this week’s Irish Farmers Journal look likely to limit expansion.

However, a casual look at cow density per county, and the lack of a profitable beef sector, could see more land switch to dairy over the next five years. Currently Ireland has on average 11 acres for every dairy cow in the country (see map). Even taking a modest increase in the density of cows where every dairy cow has one acre less on average (ie 10 acres per cow), the number of cows would increase by 200,000 to 1.7m, which would see Ireland produce an extra 1bn litres.

This increase is likely given that counties such as Cork have a cow on every 5 acres and counties with good-quality land have much lower cow density. Looking at the Netherlands, it has an average of 6 acres for every dairy cow – twice as dense as Ireland.

Additional cows could also easily come from mixed herds becoming specialised dairy units. In Glanbia’s catchment, 60% of the suppliers have mixed herds. If further pressure (economic or environmental) comes on these, some could switch.

Over the last five years while they may make headlines, new entrants have been small – around 100 in the case of Glanbia as an example. Surveys by some of the processors who have a “must take all co-op members’ milk” rule suggest dairy farmers still have more to come. Two of the country’s largest processors, Glanbia and Dairygold, are each expecting growth of a further 20% by 2025. That amounts to around 600m litres and 300m litres respectively.

The industry is expecting the growth to slow from the recent 5-6% average levels per annum. But even if we take a modest growth level of between 2% and 3% for each of the next five years, that will see 200-250m litres per annum. It is easy to see how another billion litres will arrive at processors’ gates.

Of course, there will be challenges to that growth from year to year – weather, economics, environmental regulations, etc – but the industry needs to plan and have a discussion around what the future of the dairy sector might look like and future challenges it will face. More importantly, it places a conundrum at the processors’ door – because they will need to invest (again).

A distinct feature of the Irish dairy sector is the highly seasonal milk production – a consequence of the preferred, predominantly grass-based system. At processor level it causes over investment in order to process milk through the peak supply weeks (usually around 15 weeks).

The challenge faced by most processors around the country, is that this investment may only be needed for less than six weeks per year

The decision now is whether continuing expansion in primary processing (which is low return) is still required or desired. Without doubt ongoing investment in primary processing slows down and delays other investments in higher-margin business activity.

If there is significant growth, the industry will need to decide how the processing capacity will be structured in terms of product type and funding model. Peak week processing capacity requirement is usually 3.2% of annual milk volume. So increasing by 1bn to 9bn litres will see peak week processing capacity increase by 32m litres to 288m litres per week over the peak.

A simple rule of thumb is that it costs about €12m for each additional million litres of milk processing capacity in the peak week. So in order to process an extra 1bn litres, an investment of about €350-400m will be required. To put this in context Glanbia invested €343m to process an extra 1bn litres in the last five years.

The challenge faced by most processors around the country, is that this investment may only be needed for less than six weeks per year. There are investments ongoing that will provide adequate capacity for the next three to five years depending on growth. For example, Aurivo is midway through a €50m investment programme, Tipperary is investing €30m in a dryer, Arrabawn is also investing €30m while Glanbia is investing in cheese plants in Portlaoise and Belview. Dairygold, meanwhile, is completing a cheese investment with Tine.

But the question for the industry in the next couple of years is who pays for the next round of investments to process the next billion litres. Should it be every farmer (as it has been so far) or should it be levied on only those expanding?

If the cost is borne by only the extra 1bn litres over the peak (usually around 15 weeks) it works out at 8c/l. That is a hefty cost for any litre of milk to bear. If it is based on allocating the cost over all the milk received per annum, that cost falls to around 0.5c/l. And here starts the debate. Should the farmer who is not expanding be levied or pay for stainless steel he doesn’t need?

It also gives rise to a larger question – what is the opportunity cost of investing that €350-400m? Is expansion the best return on investment or if 0.5c/l was invested on every litre could more than 0.5c be returned on every litre?

The seasonal mode of operations entails a series of implications throughout the dairy products supply chain, which reduces the cost advantages of grass-based milk production. More seasonality also increases volatility through more reliance on low-value powder commodities.

Would it make more sense to have a single peak processing plant that has scale and the ability to be turned on and off to match the supply curve be a more efficient way of processing the next billion litres?

The other option is to look at a more evenly distributed pattern of raw milk deliveries, which would allow for a higher degree of capacity utilisation and a targeting of the higher-margin, less price-sensitive markets. The debate needs to start now.