The direct answer
A bankable farm business plan is not a better-formatted plan. It is a plan in which every number can be traced to something outside your own optimism.
Most Ugandan farm loan applications are not rejected because the idea is bad. They are rejected because the plan asserts a yield, a price, a cost and a sale that nothing supports, and an appraiser has no way to test any of it. The idea may be excellent. The document is untestable, so it fails.
The correction is a discipline, not a template. For every number in the plan, you must be able to answer three questions: where did this figure come from, what happens if it is wrong by a third, and what have I already done that shows it is achievable.
You can see the standard directly in what a Ugandan development lender publishes. Uganda Development Bank's credit checklist for mainstream facilities asks, among other things, for a customer facility application letter on company letterhead, a resolution to borrow with the articles and memorandum of association and certificate of incorporation, "Customers Feasibility report/business plan (business plan guide/outline is available on the website)", "A copy of each of the last 2 years audited/draft accounts (from an ICPAU/BOU listed audit firm)", "Cashflow projections and projected financial statements (Income Statement, Balance Sheet)", "Bank statement for the last twelve months", loan statements for existing loans, "Security for the proposed loan, including current valuations of the assets and any encumbrances", a credit reference bureau report for the company and all its shareholders and directors, "Current & historical off taker agreements/purchase agreements", proof of social security compliance where applicable, know-your-customer documents including operating and trading licences, "Profiles of Directors & key technical personnel", and a tax clearance certificate.
Look at what that list is actually asking for. Six of those items are evidence about the past. Only two are about the future. A plan that is entirely about the future is a plan with almost nothing a lender can verify.
This article contains no projected returns, no expected profit, no yield estimate and no interest rate, and it does not tell you whether to borrow. It tells you what a plan must contain and to what evidence standard.
Practical summary
- Fix the legal entity and the documents first; a plan cannot be appraised without a borrower.
- Write the assumptions register before you write the plan.
- Source every assumption, and mark honestly the ones you cannot source.
- Build the cash flow month by month, not year by year.
- Show the downside case, because the appraiser will construct one whether or not you do.
- Assemble the evidence pack that the checklist actually asks for.
- Have someone who does not want the loan read it and try to break it.
Step 1: the borrower must exist before the plan does
An appraiser cannot lend to an idea. UDB states that applicants must be "incorporated entities, Co-operatives, or registered farmer groups" and that "Every Ugandan registered business enterprise engaged in an existing or start-up business qualifies for UDB products and services if the business falls in the UDB defined priority sectors", agriculture among them.
That has an ordering consequence most applicants get wrong. Registration, tax registration, a bank account in the entity's name and twelve months of statements through it are prerequisites, not paperwork to sort out after approval. The checklist asks for twelve months of bank statements. Those months either happened or they did not, and no amount of planning creates them retrospectively.
If your entity is three weeks old, the honest position is that you are not yet ready for a mainstream facility, and the useful work now is the twelve months of clean transacting that will make you ready. Falcon's article on registering and formalising a farm business sets out that sequence.
Step 2: write the assumptions register first
This is the single practice that separates plans that survive appraisal from plans that do not, and almost nobody does it.
Before writing a word of narrative, list every number the plan will depend on, in one table.
| # | Assumption | Value used | Where it came from | Evidence held | Confidence | If wrong by a third, what breaks |
|---|---|---|---|---|---|---|
| 1 | Area to be planted | Title or lease reference, survey | ||||
| 2 | Yield per unit area | |||||
| 3 | Farm-gate price | |||||
| 4 | Input cost per unit area | Dated supplier quotations | ||||
| 5 | Labour requirement and cost | Own records | ||||
| 6 | Post-harvest loss rate | |||||
| 7 | Volume actually sold, not produced | Off-take agreement or buyer letter | ||||
| 8 | Payment terms and days to payment | Contract or past bank statements | ||||
| 9 | Repairs, maintenance and replacement | Own records or quotations | ||||
| 10 | Working capital needed before first income | The cash flow itself |
Fill the fourth and fifth columns honestly. An assumption sourced from your own three seasons of records is strong. One sourced from a dated supplier quotation is strong. One sourced from a variety trial result is medium, because trial conditions are managed conditions. One sourced from "what farmers here say" is weak but stateable. One with nothing in the column is a guess, and you should mark it as a guess.
Marking a guess as a guess does not weaken the plan. It is the single most credible thing in it. An appraiser who finds one honestly flagged assumption trusts the other nine more. An appraiser who finds one unflagged fantasy stops trusting all ten.
The two assumptions that most often destroy farm plans are the yield figure and the volume actually sold. Yields taken from trial results or from a supplier's brochure are researcher-managed or seller-stated figures. Peer-reviewed work on African smallholder conditions has shown directly why they should not be carried across: after adjusting for farmer selection and behavioural responses, estimated yield gains from an agronomic trial for one crop fell to being not significantly different from zero, while remaining high for another. Volume sold is not volume produced, because of loss, rejection at grade, and buyers who do not take everything.
Step 3: build the cash flow month by month
Annual figures hide the thing a lender is actually assessing.
Farm income is lumpy. Costs are not. Between land preparation and first payment there is a period with outgoings and no receipts, and its length is set by the crop and the buyer's payment terms, not by your intentions. The working capital requirement is the deepest point of that trough, and a plan that does not show it is asking for the wrong amount of money.
Build the cash flow with one row per month for at least the full cycle, and preferably for the loan tenor. UDB's checklist asks for "Cashflow projections and projected financial statements (Income Statement, Balance Sheet)", and asks special programme applicants for revenue projections over a minimum five-year loan term.
Three rules make it defensible.
Put receipts when the money arrives, not when the crop is sold. If the buyer pays thirty days after delivery, the receipt is thirty days after delivery. If the buyer has historically paid late, use the historical behaviour, not the contractual term.
Put costs when they are paid. Inputs bought on supplier credit sit at the repayment date, not at planting.
Include the loan repayment schedule in the cash flow. A plan showing the business is viable but not showing whether it can meet the repayment in the month it falls due has not answered the question being asked.
Step 4: show the downside
The appraiser will build a downside case. You can either build it yourself and demonstrate that you have thought about it, or let them build it and discover what they assumed.
Show three columns for the key figures: the case you expect, a case where the main assumption is materially worse, and the point at which the business cannot meet its obligations.
| Variable | Case used in plan | Materially worse case | Point at which repayment fails |
|---|---|---|---|
| Yield | |||
| Price received | |||
| Volume actually sold | |||
| Input cost | |||
| Days to payment |
Then state, in one paragraph, what you would actually do in the worse case. Reduce area. Sell an asset. Stagger sales. Delay a phase. Draw on a named reserve. This paragraph is worth more than any part of the projection, because it is the only part that describes the applicant rather than the arithmetic.
Never present a plan with only a good case. It signals either that you have not tested it or that you are hiding the test.
Step 5: the evidence pack
Assemble the documents the checklist actually asks for, and assemble them before you submit rather than in response to a query. Working from the published UDB mainstream checklist:
| Category | What is asked for | What it proves |
|---|---|---|
| Entity | Certificate of incorporation, memorandum and articles, resolution to borrow | There is a borrower, and it has decided to borrow |
| Track record | Two years of audited or draft accounts from a listed audit firm; twelve months of bank statements | The business exists and its money moves through an account |
| Plan | Feasibility report or business plan, following the lender's own published outline | The proposition, and the assumptions behind it |
| Projections | Cash flow projections and projected income statement and balance sheet | The numbers, in the form the appraiser reads them |
| Existing obligations | Loan statements and agreements for all existing facilities | Total exposure, not just this application |
| Credit history | Credit reference bureau report for the company and all shareholders and directors | Repayment behaviour, including personal |
| Market | Current and historical off-taker or purchase agreements | That there is a buyer, and that there has been one before |
| Security | Security offered, with current valuations and any encumbrances | Recovery position |
| Compliance | Tax clearance certificate, social security compliance where applicable, PAYE records in some cases, trading and operating licences | The business is formal in substance |
| People | Profiles of directors and key technical personnel | Who will actually run this |
| Regulatory | Approvals or clearances from other regulatory bodies and state agencies | Nothing external blocks the project |
Two observations that matter more than the list itself.
"Current and historical off taker agreements" is the item most farm applications cannot produce, and it is the one closest to the heart of the appraisal. A letter of intent from a buyer, obtained before you apply, changes the character of an application more than any improvement to the narrative.
Use the lender's own outline. UDB notes that a business plan guide or outline is available on its website. Writing to a lender's own structure removes an entire category of avoidable friction, and it is free.
Step 6: what the narrative should and should not say
The narrative sections are short and they exist to make the numbers legible.
Include: what the business does and where; the land and its tenure position, with the verification actually done; the production plan and its basis; the market and the specific buyers, with what they have said in writing; the people and what they have done before; the risks and the mitigations; the amount requested, the purpose, the tenor and the repayment source.
Exclude: anything you would not be able to defend in a meeting.
Three claims to remove from every draft.
Guaranteed returns. Do not state or imply that a return, profit, income, yield or price is guaranteed or assured. It is not, and an appraiser reads it as inexperience rather than confidence. Falcon publishes no such claim anywhere, including in this article.
Market size as a substitute for a buyer. That a commodity has large national demand tells the appraiser nothing about whether anyone will buy yours. One named buyer beats any market statistic.
Unattributed figures. Every number in the narrative should exist in the assumptions register with a source.
Decision framework: is this plan ready to submit
| Test | Ready | Not ready |
|---|---|---|
| Entity | Registered, with tax registration and twelve months of statements in its name | Entity new, or trading through a personal account |
| Assumptions | Every figure in a register with a stated source and confidence level | Figures appear only in the projection |
| Yield basis | Own records, or a clearly labelled adapted figure with the adjustment explained | A trial or brochure figure carried across unchanged |
| Sales basis | A named buyer, ideally with something in writing | An assertion that demand exists |
| Cash flow | Monthly, including the repayment schedule and the working capital trough | Annual totals |
| Downside | A stated worse case, a failure point, and what you would do | One case only |
| Evidence pack | Assembled before submission | Assembled when the lender asks |
| Review | Read by someone who tried to break it | Read only by people who want it to succeed |
If more than two rows sit in the right-hand column, the plan is not ready, and submitting it early spends a relationship you will want later.
Field record: assumption evidence file
Keep one physical or digital folder per plan, indexed to the assumptions register.
| Assumption # | Evidence item | Type, for example quotation, contract, own record, published source | Date of evidence | Where the original is held |
|---|
Date of evidence is the column that matters. A supplier quotation from two seasons ago supports nothing. If an appraiser asks how old your input costs are and the answer is "recent", the plan has just weakened. If the answer is a date, it has not.
Search evidence for this article
In Google Keyword Planner with the location set to Uganda, the exact term agricultural business plan returned an average monthly search band of 10 to 100, and farm business plan also returned 10 to 100. commercial farming in uganda returned 10 to 100. Bands are reproduced exactly as returned and none has been converted into a single number.
A Uganda-geolocated Google search for agricultural business plan uganda returned a document-sharing site copy of a farm plan in first position, a property company's agribusiness overview in second, MAAIF's Agriculture Sector Strategic Plan in third, a commercial blog offering bankable plans in fourth and an NGO proposal sample in fifth. The result set carried videos, short videos, images and people also search for.
That result set is the argument for this article. Three of the five leading results supply a document to copy. One is a national sector strategy, which is a different kind of document entirely. None explains the evidence standard an appraiser applies to the assumptions inside the template. A copied template with unsourced numbers is exactly the document that gets rejected.
Uganda-specific importance
Ugandan agricultural lending is constrained less by the absence of capital than by the difficulty of appraising farm propositions. Seasonal income, informal record-keeping, land held under tenure arrangements that complicate security, and buyers who deal in cash all make verification hard. An appraiser facing an unverifiable plan does the rational thing and declines.
That is why the leverage sits with the applicant. Almost everything on the published checklist is producible by an ordinary Ugandan farm at low cost, given twelve months of intent: transact through the entity's account, keep dated quotations, keep production records, get one buyer to put something in writing, and file it. The applications that succeed are frequently not the largest or the most sophisticated; they are the ones where the appraiser could check things.
Limitations and common mistakes
Limitations. This article states no interest rate, fee, tenor, equity contribution requirement, loan size, projected return, expected profit, yield or price, because none was verified and because Falcon does not publish such figures. It is not financial advice and does not recommend borrowing, any lender or any product. The checklist described is one Ugandan development lender's published mainstream requirements at the access date; commercial banks, SACCOs, microfinance institutions and supplier credit have different requirements, which are the subject of a separate Falcon article. Bank of Uganda operates an agricultural credit facility; its terms could not be verified in this research and are therefore not stated anywhere. This article does not repeat gross-margin calculation method or the investment due-diligence checklist, both of which are published Falcon articles.
Common mistakes.
- Starting with a template. The template is the last step, not the first.
- No assumptions register. If the numbers exist only inside the projection, they cannot be defended.
- Carrying trial yields into a farm plan. Trial conditions are managed conditions, and the difference has been measured.
- Projecting volume produced rather than volume sold. Loss, rejection at grade and partial off-take all sit between the two.
- Annual cash flow. It hides the working capital trough, which is the number you are actually borrowing against.
- Omitting the repayment schedule from the cash flow. The appraiser's central question is whether you can pay in the month it falls due.
- Presenting only a good case. It reads as untested.
- Claiming a guaranteed return. It is the fastest way to be read as inexperienced.
- Trading through a personal account. It destroys the twelve months of statements the checklist requires.
- Applying before the evidence pack exists. A first impression is spent once.
- Undated evidence. A quotation with no date supports nothing.
When to bring in a professional
- An accountant, for the financial statements and projections, and well before the year end that will be audited.
- An auditor from a recognised professional body, because the checklist specifies accounts from a listed audit firm.
- An advocate registered in Uganda, before signing any facility agreement, security document or off-take agreement.
- A qualified agronomist, to sense-check the production assumptions before they enter the register.
- A valuer, for any asset offered as security, because current valuations are asked for explicitly.
- The lender's own relationship officer, early, to obtain their published outline and confirm current requirements.
