Why Concentration in Land Is Not Diversification
Understanding Risk
1. The Universal Investor Problem
Every Ugandan investor knows the land mantra: “Buy plots in different areas and you’re diversified.” Buy one plot in Kira for quick flips. Purchase another in Entebbe to benefit from tourism upside. Get a third in Ntinda to meet middle-class rental demand. Three plots feel like smart risk management. Then cement prices spike 40%, tenants vanish during economic slowdowns, and 24% bank loans crush cash flow across all three. What felt like diversification reveals itself as concentration in one correlated asset class.
This isn’t unique to Uganda. Investors worldwide chase “different” stocks within tech or crypto, only to watch entire sectors evaporate. Concentration disguised as spreading risk destroys more wealth than outright market crashes. The universal problem: humans mistake owning similar assets in different wrappers for true diversification.
2. The Core Concept: Correlation Creates Risk
Diversification works when assets zig, others zag. Land plots, even across Kampala suburbs, move together. They face identical macro forces. These include construction material costs, interest rates, tenant solvency during slowdowns, and government titling delays. Three Kira plots don’t diversify land risk; they amplify it through management overhead without return benefit.
True diversification pairs assets with low correlation.
To easily describe correlation. Imagine you and I are connected. If I step to the right, there’s a very high chance you’ll step to the right too. That’s what it means to be correlated.
- Land provides steady 8-12% annual appreciation but stalls during credit crunches.
- USE bank stocks like Stanbic Bank deliver dividends through real estate slumps.
- Agricultural SACCO deposits offer stability when urban plots sit vacant.
The magic happens when returns offset each other.
Simple math preview:
Portfolio risk depends on individual volatilities AND their covariance. Three perfectly correlated land plots (correlation ≈ 0.9) barely reduce risk VS. one plot of Land + equities (correlation ≈ 0.3) and the risk drops 40-60%.
3. The Analytical Framework: Covariance Rules Everything
You can ignore the math below. All it means is that having assets whose prices move differently does a lot to protect your investments than having assets whose prices move similarly.
Modern Portfolio Theory (Markowitz, 1952) proves portfolio risk equals:
Three land plots scenario: w₁=w₂=w₃=33%, Cov≈σ₁σ₂, ρ≈1.0 (perfect correlation). Portfolio σ ≈ individual plot σ. No diversification benefit.
Land + equities scenario: 70% land (σ=25%), 30% Stanbic (σ=18%), ρ=0.3. Portfolio σ drops to 21%—16% risk reduction despite lower equity weighting.
Ugandan land reality: Regional plots correlate 85-95%. 2020 COVID showed this brutally. Almost all land transactions froze simultaneously while bank stocks recovered 35% by 2022.
Geographic “diversification” within one asset class fails systematic stress tests.
4. Quantitative Evidence: Simulations Don’t Lie
Consider 100M UGX across three strategies over 10 years (12% land mean return, 25% volatility; equities 11%, 20% vol, correlation(ρ)=0.3):
Equal-weight three land plots: Median outcome is 310M UGX. There is a 5% chance that the returns are less than or equal to 125M.
70/30 land/equities: Median 340M UGX. There is a 5% chance that the returns are less than or equal to 210M. The probability of returns being less than 125M is even smaller, with only a 22% severe loss probability.
| Strategy | Median 10yr (M UGX) | 5th %-ile (M UGX) | Risk Reduction |
| 3 Land Plots | 310 | 125 | Baseline |
| 70/30 Land/Equity | 340 | 210 | 68% better tail (less risk) |
Monte Carlo confirms: Correlation less than 0.5 slashes the downside by 1.5 standard deviations. Land’s high idiosyncratic risk (tenant defaults, titling) demands equity ballast.
5. Limitations: When Diversification Breaks
Correlation estimates fail during crises. 2022 global rates crushed bonds AND equities simultaneously. Uganda’s land adds unique traps: 1.5% stamp duty on sales, 6-year holding minimums, political evictions. Illiquidity prevents tactical shifts when USE rallies.
Estimation error plagues small samples. Five years of data can’t predict decade-long correlations. Transaction costs (2-3% round trip) erode theoretical benefits for retail investors trading annually.
Yet partial diversification still beats concentration. Even correlation (ρ)=0.7 cuts risk 25%. The alternative. Naked land bets (lack of diversification) courts permanent capital destruction.
6. The Timeless Lesson
Diversification isn’t owning three similar plots; it’s owning cash flows that behave differently. Thoughtful investors build portfolios where land’s illiquidity complements equity liquidity, SACCO stability offsets stock volatility. Correlation, not geography, creates safety.
Action: Map your “diversified” holdings by correlation matrix. If the correlation of the assets is greater than 70%, they move together. You should re-balance toward true offsets. Uganda’s growth favors patient combiners, not frantic land collectors.
References:
- Markowitz, H. (1952). “Portfolio Selection.” Journal of Finance.
- Uganda Securities Exchange historical returns data.
- Bank of Uganda real estate lending statistics.
