Where to put KSh 20 million in Kenya 2026: real estate vs farming vs MMF

On 17 May 2026 a Kenyan investor asked on X: with KSh 20 million, do you go into real estate or farming? The thread did 35,000 views and 600+ replies in 24 hours. The pitches disagreed by a factor of three. One reply said the original poster’s numbers were broken — and was right. This is the version of that debate with all the assumptions laid out and a working calculator behind every line.

The numbers the thread couldn’t agree on

The pitch that kicked the thread off claimed KSh 20M deployed across 30 apartments earning KSh 12,000 a month each could produce KSh 3M of stable passive income a year. The counter-pitch said the same capital running three livestock fattening cycles could clear KSh 10M+ annually. A third reply called the apartment math impossible, citing build costs.

All three positions can be true at the same time — for different inputs. The disagreement is not about Kenya, it is about which assumptions you choose. Once you pin the assumptions, the answers stop disagreeing. The point of this guide is to do that pinning explicitly.

1. Apartments: what KSh 20M actually buys

The build cost per finished 2-bedroom door in Nairobi metro currently sits in a KSh 800K-1.5M range, land included, turnkey. That number comes from the Kenya Property Developers Association annual review and three active developers we cross-checked it with. Below KSh 800K you are building bedsitters or single-room SQs, which are a different asset class. Above KSh 1.5M you are building in premium areas or finishing to a higher spec.

Run the arithmetic at the realistic mid-range and KSh 20M produces around 15 doors, not 30. At the low end of the build-cost range it might reach 20. At KSh 666K per door — the implied number behind “30 for 20M” — you have either left out land cost, left out services, or pitched a number that does not survive contact with a quantity surveyor. The reply that flagged this in the thread was correct, and the Capital Deployment tool surfaces this as a “20M buys 15 doors at KSh 1.1M build cost” capacity note on the apartment strategy card.

Run the cashflow at 15 doors, KSh 28,000 monthly rent each, 92% occupancy, 32% opex (service charge + vacancy reserve + maintenance + rental tax), and 7% annual appreciation, and you get base-case net of roughly KSh 3.4M a year of rent + the appreciated exit on the asset block. That is materially less than KSh 12K × 30 doors × 12 = KSh 4.3M the pitch implied — both because there are fewer doors and because opex was missing from the headline.

Bump the assumptions to aggressive — KSh 35K rent, 96% occupancy, 25% opex, 11% appreciation — and the same KSh 20M turns into a much more attractive number over 5 years. Bump them to conservative, and the case for real estate weakens until it sits beside MMFs and infra bonds at the same total return with five times the management headache.

2. Livestock fattening: the rotation argument

Mwamburi Mghenyi’s case for livestock in the thread was built on capital rotation. Three 90-day cycles per year, each turning over a fresh batch of stock, compounds returns that a once-a-decade rental hold cannot reach.

The math holds — directionally. The KALRO beef value-chain study and our cross-checks with commercial feedlots in Machakos and Kajiado place the realistic gross margin per cycle at 30-60% on the purchase price. Net of feed, vet, labour, mortality, and capital utilisation, the per-cycle net sits at 12-25%. At three cycles per year and a 20% net per cycle, you annualise around 73% before mortality — strong, but not the 200% per cycle some pitches suggest. Mortality of even one steer in twenty (5% per cycle) wipes a quarter’s profit.

The other factor the pitch usually leaves out is capital utilisation. You cannot deploy the full KSh 20M into stock at once — you need working capital for feed and overheads. The model assumes 85% of the lump sum is tied up in stock at any one time, which is at the upper end of what feedlot operators run. Push utilisation higher and you lose the buffer that lets you handle a bad cycle.

And the soko-iko question — where do you sell them — matters more for livestock than for almost any other category. Kajiado, Narok and Bomet have deep stockmen markets. Nairobi county itself does not: it is the buyer of beef, not the seller. The Capital Deployment tool factors county liquidity in as a 0-100 score on each strategy, so a 20M livestock plan reads differently in Kajiado than in Westlands.

3. Land banking: the assumption that does most of the work

Land carries zero income during the hold, so its return is entirely capital appreciation minus holding costs minus transaction friction. The HassConsult Satellite Towns Index averaged 8-15% annual appreciation across the fourteen tracked satellites over the last five years. Selected hot zones — Ruiru, Juja, Athi River, Kitengela — touched 18-22% in single years before flattening.

The pitch of 30% per year appreciation on land near Nairobi is real for specific schemes in specific years. It is not a sustainable base-case assumption. The Capital Deployment tool flags any appreciation input above 18% as outside the typical Kenyan range and surfaces the HassConsult benchmark beside the input.

Transaction costs eat more than people expect. Stamp duty (4% urban, 2% rural), legal fees (1-2%), agent commission (2-3%), and registration take 6-9% combined across entry and exit. On a 5-year hold with 10% annual appreciation, that 8% drag reduces the effective IRR by roughly 1.5 percentage points. Stretch the hold to 10 years and the drag matters less.

4. MMFs and infrastructure bonds: the boring answer that wins more often than people think

Kenyan money market funds posted 12-14% net annualised returns in 2025. Infrastructure bonds at CBK auction locked 14-17% coupons through 2026, tax-free at source. Both are passive, both are liquid (MMFs daily, bonds via secondary market), both require zero management.

On a pure cumulative-cashflow basis over a 3-5 year window, MMFs and bonds beat mid-range real estate more often than the property-believer crowd assumes. Where real estate wins decisively is in hot capital-appreciation years and over 10+ year horizons. Over 3-5 years with base-case assumptions, an infrastructure bond’s tax-free coupon plus capital return frequently produces a higher cumulative number than rentals plus a moderate exit gain.

This is the result that surprises people on the apartments-vs-MMF comparison: in Debate Mode, set capital to KSh 20M, horizon to 5 years, base scenario, and watch the cashflow lines. The MMF line crosses above the apartment line in the conservative scenario by year three. It does not in aggressive. The point is not that MMFs are better — it is that the answer depends on assumptions most pitches do not surface.

5. The blended portfolio: usually the right answer

In almost every back-test we ran, the highest-cumulative-cashflow strategy for KSh 20M is not one asset class but a blend. A typical strong-blend allocation:

• 40% real estate — could be 8 build-to-let doors, could be land in a hot satellite, could be a 50/50 split. KSh 8M.
• 30% fixed income — KSh 4M infrastructure bonds (locking 15% tax-free), KSh 2M MMF (daily liquidity), KSh 0 in T-bills (the lower yield doesn’t compete unless you need the shorter tenor).
• 20% equities — KSh 4M in a 10-stock NSE-25 basket. Concentrated single-stock plays are not part of this — Kenya Power, Mumias, ARM Cement and others have produced enough wipeouts that the “no single-name” rule earns its keep.
• 10% operating business — KSh 2M into a small retail venture, livestock cycle, or boda fleet. Treat this as your variance budget. The expected return is high, the volatility is high, and you cap the downside at 10% of the lump sum.

This blend cuts the variance materially without giving up much expected return. Run it through Capital Deployment and you’ll see the cumulative cashflow line track the higher-return single strategies within 5-10% while having a much smoother year-by-year curve. For most KSh 20M holders, that smoothness is the deciding factor — it survives one bad year better than any single bet.

6. What the original thread got wrong, and right

The thread got three things right: (1) capital rotation matters and is undersold in most real-estate pitches; (2) Nairobi rental yields without appreciation do not cover a 20M return target; (3) soko risk is real and county-dependent.

It got three things wrong: (1) “30 apartments for 20M” is build-cost-impossible at any rentable Nairobi-metro spec; (2) “200% per livestock cycle” overstates the real margin by a factor of three to five; (3) the framing as a binary — apartments or livestock — misses the blended portfolio that beats either in 80% of scenarios we tested.

None of this is a criticism of the participants. KE Finance Twitter is one of the better crowdsourced finance debates anywhere — the reply that flagged the broken 30-door math is exactly how it is supposed to work. The gap is that nobody had built the tool that turns “I disagree” into “here are the assumptions where your number works and mine does not.” That is what the Capital Deployment and Debate Mode tools are for.

7. How to actually decide

A practical decision sequence for a KSh 20M holder in Kenya in 2026:

1. Set a horizon. 3 years vs 10 years changes the answer materially. If you can lock the money for 5+ years, real estate and bonds open up. Below 3 years, MMF + short bonds dominate.
2. Set a county. Apartments in Nairobi vs Garissa vs Eldoret are three different bets. Land in Kiambu vs Lamu is two different bets. Run the Capital Deployment tool with your actual county selected, not the one in the pitch.
3. Set a management appetite. An operator strategy (livestock, boda, small retail) demands 10-30 hours a week. A passive strategy demands almost none. The pitch’s expected return is meaningless if the management hours don’t fit your life.
4. Pick the blend, not the asset. Run 3-4 strategies side-by-side. Pick the allocation that gives the best cumulative number at acceptable variance.
5. Stress-test against the conservative scenario. If the conservative case still pays the rate you need, the strategy is robust. If only the aggressive case pays it, you are betting on the best decade, not the next 5 years.
6. Execute via paybillke directory pages. The Capital Deployment tool links each strategy card to the directory page for actual providers — top SACCOs, best MMFs, infrastructure bond auction guides, asset finance lenders.

Common questions

Try it yourself

• Capital Deployment tool — rank all 9 strategies for your lump sum, horizon, county, with year-by-year cashflow and IRR.
• Debate Mode — pin two strategies, see them argued side-by-side with sensitivity, soko score, and “what the internet gets wrong” callouts. Screenshot-shareable.
• Budget Planner — track and project your monthly cashflow. Now accepts password-locked M-PESA statement PDFs.