Many in the ag industry, including me, have persistently anticipated a significant production cutback from the dairy industry as milk and product prices continually declined from fall 2014 records. However, as the following chart indicates, that has not been the case. Indeed, as the blue line indicates, the 9.327 million-head total seen during the March-May period was quite high by historical standards, falling only modestly below the 9.334 million highs posted in 2008. The situation might be different if herd size were the only factor affecting the market, but the red line on the chart representing monthly milk production shows one big reason why prices remained very weak into late spring. That is, milk production has clearly trended higher in the new millennium, despite some rather dramatic fluctuations in the size of the milking herd. Indeed, the May total set a fresh all-time monthly record at 18.645 billion pounds.
Grain and soybean price declines from summer 2012 highs to early 2016 lows probably encouraged producers somewhat, since those implied commensurate cuts in feed costs. Energy costs also fell sharply over the past two years. Still, past market shifts have indicated that fluctuations in milk prices are the usually the most important factor affecting producer decisions to boost or to cull dairy cow herds. Conversely, these market factors have rather obviously built farmer incentives to boost their productive efficiency, thereby reducing their cost per gallon of milk produced.
This second chart depicts the upward trend in milk production per cow over the past 17 years. Indeed, the surge from 1,579 pounds/head in May 1999 to the May 2016 figure at 1,999 pounds marks a 26.6% improvement. We can certainly argue that improved cow genetics, better husbandry and economies of scale have combined to boost industry efficiency, but I have come to suspect that the advent of robotic milking is having a vast and greatly underrated impact upon the industry.
Recall our image of dairy farmers; these men and women rise in the middle of the night in order to do the early-morning milking, having to handle the cows, process the milk and clean the equipment before sunrise, only to have to repeat the process in the afternoon. They couldn't take a day off, much less a vacation, without making arrangements with someone, who had better know their way around a dairy, to handle all those chores. We can bet they're still pretty well tied down, but robotic milking is clearly changing the way they operate.
Cows now come to the milking parlor on their own schedule. They're milked and fed with little human intervention. They may be milked numerous times per day (the average is around 2.5), thus improving their productive potential.
Given these shifts, it's easy to see how productivity of the individual farms and the industry as a whole are increasing. Moreover, this phenomenon is probably causing two other significant changes in the adapting enterprise's cost structure. First, the installation of robotic milkers isn't cheap, which almost surely translates into increased fixed costs for the farm operation. Second, this implies a significant reduction in the operation's labor requirements. That may allow for more rest for the owner-operator and possibly the elimination of an extra hand from a large dairy. The increased fixed costs suggests producers will be much more reluctant to reduce production and/or close an operation if times get bad. Indeed, as long as they're covering fixed costs, they seem likely to continue operating and 'paying the note' on the machines. Also, the advantages conferred by the robots give those with small herds to at least expand to the 50-60 head level commensurate with the capacity of the smallest robots. As with most operations, economies of scale should also apply. Thus, the advent of robotic milkers will increase the pressure upon those with smaller and/or older operations. The trend rather clearly points toward larger herds.
Again, this move to robotic milking may help explain the U.S. dairy industry's stubborn refusal to trim herds significantly over the past two years. The chart above puts the latest herd totals in a shorter-term context, showing the recent plateau at 9.327 million head was slightly larger than the 2015 peak and well above the 10-year average. In fact, with Class III milk prices climbing from their May low at $12.76/cwt to $13.22/cwt in June and CME futures indicating that the 'cheese milk' price will reach $$16.50/cwt in August, the market now seems to be offering producers significant incentives to expand production. Thus, while some cow culling generally occurs during the second half of the year, one now has to wonder if that will be the case during the coming months.
These developments and ideas are causing me to rethink my expectations for U.S. milk production this year and next. I had previously thought the USDA was too optimistic in forecasting moderate increases in 2016 and 2017 milk output, at 212.4 billion and 215.6 billion pounds, respectively, from 208.6 billion in 2015. I now suspect the industry will easily top those projections, especially if forthcoming prices live up to the expectations implicit in CME futures.