Over the full cycle
Does the farm pay?
Compare projected revenue with operating costs and investment over ten years, allowing for the timing of those cash flows.
Research
The economics of the whole coffee tree
Every coffee tree grows two harvests. The world pays for one. The Whole Tree Economic Model tests whether selling the leaf and fruit could pay for the shift to regenerative farming.
The model in brief
Could another harvest pay?
The bean
Start with what the farm earns from coffee today.
The rest of the tree
Add the revenue from leaves and fruit harvested from trees the farm already pays to grow.
The real cost
Subtract the costs of picking and processing, and any loss of bean income. Then test the result against changes in price, yield and cost.
- 10test acres at Copper Hill Farms
- 6management scenarios
- 10years modeled
- 1,000simulations per scenario
The results depend on the model’s assumptions. Read the assumptions and limits or find the thesis citation.
The problem
The shift to regenerative agroforestry can require new planting, different management, and cash to carry the farm while yields and costs change. A grower needs income through that transition as well as a viable business afterward.
The thesis asks whether leaves and fruit from trees the farm already tends could help finance that change.
Finding 1
Six ways to manage the same ten test acres
Mean ten-year net present value (NPV) per hectare and the share of 1,000 simulations with positive NPV. All six scenarios use the same 10 test acres; results are expressed per hectare.
Swipe sideways to see the full chart.
Under the study’s assumptions, conventional management produced negative NPV in all 1,000 simulations. Regenerative management with coffee leaf had the highest mean NPV. The separate stress test also produced positive ten-year NPV with bean revenue set to zero.
View the figures as a table
| Scenario | Mean NPV per hectare | Positive runs |
|---|---|---|
| Conventional | −$119,128 | 0 / 1,000 |
| Regenerative | $52,353 | 1,000 / 1,000 |
| Regenerative + fruit | −$32,644 | 0 / 1,000 |
| Regenerative + leaf | $195,307 | 1,000 / 1,000 |
| Stress test: no bean revenue | $79,032 | 1,000 / 1,000 |
| Whole tree | $141,944 | 1,000 / 1,000 |
Finding 2
Putting the model at farm scale
Ten-year NPV for a 4.05-hectare reference farm, approximately 10 acres. The figures represent projected cash flows after costs, discounted to present value.
Swipe sideways to see the full chart.
The thesis reports a difference of about $1.06 million in ten-year NPV between conventional and whole-tree management over the same reference area. Prices, yields, costs, and access to buyers determine how useful that comparison is for another farm.
View the figures as a table
| Scenario | NPV for reference farm |
|---|---|
| Conventional | −$482,463 |
| Whole tree | $574,874 |
The cash-flow test
Getting through the transition
A farm can have a positive ten-year NPV and still run short of cash during the transition. The thesis tests long-term viability and cash flow during the transition separately.
During the transition
Can the grower keep going?
Examine annual income while yields and costs change, alongside the household’s needs and access to finance.
For a grower, both questions matter. A promising long-term return still needs a workable plan for the years before it arrives.
Source: thesis methods, pp. 50–51; transition analysis, pp. 95–98.
Finding 3
Where the money comes from
Modeled first-year gross revenue for the 10 test acres under the whole-tree scenario. Operating costs are accounted for separately.
Swipe sideways to see the full chart.
In this model, coffee leaf and cascara supply about 68 percent of gross revenue. The leaf contributes nearly twice the bean revenue. Buyer demand and the prices assumed in the model are essential to that result.
View the figures as a table
| Source | Gross revenue | Share |
|---|---|---|
| Coffee leaf | $47,648 | 61% |
| Coffee beans | $25,399 | 32% |
| Cascara | $5,247 | 7% |
| Total | $78,294 | 100% |
Source: thesis scenario S6 and Appendix 9, Table 28, pp. 203–204 and 215.
Finding 4
The idle-season advantage
Harvesting and processing leaf between cherry harvests could put existing mills and dryers to work during quieter months. The thesis calls this temporal complementarity. Putting it into practice depends on suitable harvest timing, equipment, crews, and buyers.
Swipe sideways to see the full chart.
The thesis illustrates peak cherry harvest from November through February and a potential leaf-processing window from April through September. March and October separate those windows. Cascara arrives with the cherry harvest and can compete for the same people and machines. Actual schedules vary by farm and season.
What drives the result
- The leaf leads in this model. Regenerative management with beans and leaf produces the highest mean NPV among the six scenarios.
- Adding fruit brings another set of costs. The regenerative-plus-cascara scenario includes bean revenue and still produces negative NPV under the study’s assumptions. Harvest-period processing constraints help explain the result.
- Costs and infrastructure matter. The model compares conventional inputs with regenerative management and assumes access to shared processing. Those conditions shape the result.
- Resilience is an assumption to test. The hurricane scenarios assume lower damage under regenerative management. Field conditions and the severity of an actual storm will determine the outcome.
Questions for growers
- Assess the transition. Work out the costs, likely yield changes, and cash needed for regenerative management on your farm.
- Test the leaf. Establish a buyer, quality requirements, processing capacity, and the effect of harvesting on the tree. The model’s entry case uses one annual leaf harvest from routine pruning.
- Evaluate the fruit. Consider cascara where its price, processing costs, and harvest-season demands make it worthwhile.
The base case assumes no bean-yield penalty from that single pruning harvest. The thesis’s higher-frequency scenarios include added labor, quality changes, and possible bean-yield penalties. The farm’s response needs to guide the harvest plan.
What it would take
The thesis proposes support in three stages:
- Build capacity. Develop suitable shared processing and the skills to produce consistent-quality leaf.
- Carry growers through the transition. Combine a workable cash-flow plan with finance, extension support, and conservation funding where available.
- Develop the market. Establish buyers, quality standards, and repeat orders for coffee-leaf products.
The sequence matters. Growers need processing capacity and a route to market before taking on the costs of a new harvest.
What the thesis adds
- Two tests of viability. A farm has to pay off over time and survive the years it takes to get there. The framework separates the economic barrier from the financial one.
- Temporal complementarity. Off-season leaf processing could earn income from capacity that sits idle between cherry harvests.
- Infrastructure as finance. Existing mills and cooperatives could help growers earn revenue from leaves and fruit to support regeneration.
Methods & limits
These are modeled estimates based on a study plot of 10 test acres at Copper Hill Farms, approximately 4.05 hectares. They describe the case study and its assumptions; results on another farm depend on its costs, yields, harvest practices, and buyers.
Read the main assumptions and limitations
- Time and value. The analysis uses a ten-year horizon, a 3.57 percent real discount rate, and costs stated in 2024 US dollars. NPV brings future cash flows into present-value terms.
- Units. The six-scenario comparison reports NPV per hectare. The farm-scale illustration covers 4.05 hectares. The revenue chart reports gross first-year revenue for the reference plot of 10 test acres.
- Markets. The base case assumes finished coffee-leaf product prices of $8 to $10 per pound and access to processing and buyers. Those are model inputs that require validation in practice.
- Harvest. The base case uses one annual leaf harvest from routine pruning. Harvest frequency, tree response, and effects on bean yield need field validation.
- Simulation. Each scenario was run 1,000 times within the model’s selected parameter ranges. A 100 percent positive result describes those simulated runs; outcomes depend on whether the assumptions hold.
- Transferability. Puerto Rico’s labor costs, specialty pricing, and farm conditions shape these results. Each new location needs its own economics.
Thesis · 2026
Monetizing Coffee Leaves and Fruit to Fund Regenerative Agriculture: A Case Study in Puerto Rico
Harvard Extension School · Harvard DASH
Field & laboratory
At Copper Hill Research Station, the work turns to practical questions about botanical processing, coffee-leaf composition, and how leaf harvest affects the crop.
From research to market
The field and laboratory work at Copper Hill informs my work on the Specialty Coffee Leaf category.