Modeling a path forward for expanding dairy farms
Print
EmailDairy farms must consider many factors when considering expansion.
Dairy herds with less than 500 cows face economic challenges due to their relative lack of scale. While many such farms remain profitable, difficult decisions emerge when key facilities like a milking parlor become worn out and re-investment is needed. In these situations, smaller dairy farms often face a dilemma: shift to niche marketing or expand the farm. New MSU research explores the most profitable strategies for expansion and has generated a tool for commercial farms to use for their own planning.
Over the last 20 years, there has been steady consolidation within the U.S. dairy industry, resulting in about 40,000 fewer dairy herds in 2023 than 20 years prior (Figure 1). This decline in total herd numbers has been primarily driven by a steep decline in dairy herds that milk less than 200 cows. Many herds within that size range have had to make the difficult decision to either expand to remain competitive, find an industry niche or exit the dairy business altogether. The primary drivers forcing these decisions are the economies of scale disadvantages faced by small dairy producers as the industry has evolved. As herd size increases, net returns per pound increase while total costs per pound decrease, with fixed costs being spread across a greater number of cows. While the small dairies (< 200 cows) have rapidly declined, the industry has an increasing number of herds milking more than 1,000 cows. Despite the decline in U.S. dairy herds, milk production has increased from about 170 billion pounds of milk in 2003 to over 220 billion pounds in 2023.
There are some dairies in Michigan that have been very successful by finding niche markets, including a recent MSU Dairy Farm of the Year. Location, skill sets, interests of the owners and marketing expertise are all critical factors for farm-to-consumer marketing efforts. MSU has several programs to support such efforts, and for some farms, this is a fruitful path to take.
However, it’s important to also acknowledge that many farms have not succeeded in niche marketing efforts, and the reality is that 99% of dairy products are sold through conventional commercial supply chains. As such, many smaller farms will likely reach a point where they will have to decide to expand or stop milking cows. Why? The most common reason is that milking parlors have a finite lifetime—typically 20-30 years at most—and they are very expensive to replace. We estimated that a typical 250-cow dairy today would have to spend approximately $130,000 to build a new, standard double-10-stall milking parlor, and very few 250-cow dairy farms have sufficient profit margins to pay for that reinvestment, particularly with interest on a loan for this project.
In this situation, these farms often recognize that a modest additional investment will provide a parlor with capacity to milk twice as many cows, and this often becomes the path chosen to dilute the loan payments across more animals. However, even if the decision is made to expand the farm, there are several strategies that can be considered.
Understanding these economic pressures on dairy farms and the trends within the industry, our goal was to determine the most profitable, efficient and least risky approach to expansion for small dairy farms through economic modeling. There were four expansion strategies that were evaluated within this study:
- DOUBLE: Double land, cows, barn capacity, milk storage, and manure storage capacity.
- BUYFORG: Bypassing the land investment, scale up the herd size and the infrastructure at a greater rate while purchasing forages to feed the same traditional Midwest diet as in the first scenario.
- BUYCOMM: Bypassing the land investment, scale up the cows and the infrastructure at a greater rate while feeding a byproduct-heavy diet, feeding only the forages that can be produced on current acreage.
- ROBOT: Scale up land, cows, and infrastructure as much as the capital allows while investing in robotic milking systems.
To compare these options in an “apple-to-apples” manner, each of the scenarios utilized the same total investment cost of about $5 million dollars, allowing the farm to remain at 50% equity after borrowing for the expansion. To evaluate these scenarios, a 250-cow base herd (prior to expansion) was defined based on average Dairy Record Management System (DRMS) statistics for herds between 100-500 cows in the Midwest. Table 1 defines the key characteristics for each scenario compared to the base herd.
As Yogi Berra once said, “It’s tough to make predictions, especially about the future!” To help account for the uncertainty in projected costs and values for this farm, we used a process called Monte Carlo simulation. In this type of simulation, a distribution of possible prices is “sampled” by the computer for each simulation. Therefore, although the expected interest rate we entered would be the most common value used in a simulation, values above and below that expected mean would also be selected sometimes, following a bell-shaped curve. To generate those normal distributions of expected values, means and standard deviations for at least the last three years were found using USDA records, peer-reviewed research, and Extension articles for over 75 prices and costs which impact farm economics annually (milk price, heifer value, interest rate, etc.). We examined the outcomes by running each scenario through a Monte Carlo simulation 10,000 times, pulling different values from each of those 75 variables each time. The output modeled the economics of the farm for 10 years after the initial investment.
Key Result #1: Avoiding additional land purchase and scaling the herd more dramatically was the most profitable financial investment and resulted in the least debt after 10 years.
Figure 2 represents the average annual net profit for each of the 10 years following the investment. Scenario BUYFORG, bypassing the land investment and scaling up at a greater rate while maintaining a forage diet, resulted in the greatest average net profit each year. This is likely due to the ability to milk more cows than in scenarios DOUBLE and ROBOT and the reliance on a forage diet, which results in less feed cost variability. Scenario BUYCOMM, which relied on byproducts within the diet, resulted in the most variable annual net profit, while ROBOT led to the least variability in annual net profit. Further, we evaluated the total return on investment after 10 years (Table 2) and found that BUYFORG was the only scenario that resulted in a positive return on investment (ROI) in that time frame. The two scenarios with the greatest number of cows (BUYFORG and BUYCOMM) resulted in the greatest ROI but also had the greatest variation in their ROI.
When it came to solvency, we looked at the debt-to-asset ratio after the 10-year simulation (Table 2). We found that BUYFORG had the least amount of debt remaining at the end of the 10 years. This indicates that BUYFORG has the greatest ability to meet long-term financial obligations compared to the other scenarios.
Profitability does not equal cash flow. Yet, the BUYFORG scenario led to the greatest probability of positive cash flows each year compared to the other scenarios. Approximately four out of every 10 years, BUYFORG had a positive cash flow (Figure 3).
Key Result #2: The robotic milking scenario provides the most predictable outcomes.
Scenario ROBOT led to the most predictable outcomes, indicating that there is less variation around the mean. Having predictable outcomes becomes important for producers who rely on the certainty of an outcome due to not having the ability to take on a lot of risk. Whether the outcome is desirable or not, having a more predictable outcome aids producers in making smart and sustainable decisions for their operations.
Key Result #3: Sensitivity analysis revealed few scenarios that would change the ranking of options.
Sensitivity analysis was carried out to determine at which point a change in a variable alters the overall conclusions. Feed price, milk price, robot milk production, land appreciation rates, interest rates, cow value and labor costs were all evaluated with net profit per farm annually. All sensitivity analysis was based on only 1 variable change and no other changes to the model or inputs.
Milk price sensitivity graphs (Figure 4) show that the more cows a farm milks, the greater the annual net profit when milk price increases. Figure 4 graphs the annual net profit by farm expansion method, denoted by color. The black dotted line represents how each of the scenarios rank based on the way milk was priced in the simulation ($2.70/cwt milk fat and $2.49/cwt. milk protein). The yellow dotted line represents the point at which the milk price change would result in the re-ranking of the scenarios. Milk prices (both fat and protein prices) would have to decrease by 9% compared to the prices used in our model, for scenario ROBOT to become the most profitable scenario annually.
Predicting the milk production change per cow when switching from conventional parlor milking to robotic milking systems can have a significant impact on profitability projections. Our base model included a 7.5% expected increase in milk production per cow for cows milked on robots vs. a parlor. However, we also evaluated how scenario ROBOT would rank if there were no milk production advantage or if the benefit was greater. Sensitivity analysis revealed that if there was no milk production increase for robots compared to conventional milking, the robotic milking system would be the least profitable annually. Milk production would have to increase by 20% or more relative to conventional milking for scenario ROBOT to become the most profitable option.
Feed cost graphs showed that the more cows a farm feeds, the greater the reduction in annual net profit as feed costs increase. Sensitivity analysis revealed that total feed costs would have to increase by 15% for scenario ROBOT to become the most profitable annually.
Other than feed cost, milk price and robotic milk production change, no other variable tested within the sensitivity analysis yielded a reranking of the scenario outcomes.
Online decision-making tool
The decision and methods to expansion are not a one-size-fits-all scenario. Each dairy is unique, and one expansion method may not be most profitable or least risky for all operations. Therefore, we have developed an online extension tool that can be used by dairy producers to enter their own inputs and run the simulation. This will provide more insight for specific operations, accounting for unique factors like local land price and access to capital. MSU Extension educators are available to assist in utilizing and interpreting simulation results. To access the tool, visit our website.
This work is supported by the Michigan Alliance for Animal Agriculture (M-AAA), from AgBioResearch and MSU Extension at Michigan State University, in partnership with the Michigan Department of Agriculture and Rural Development.
