Diversification and Location Risk in Property Portfolios
Location and concentration decide whether a loss arrives at one building or at several at once. When properties share a market, the drop in income and value lands on all of them together, and loans sized to value shrink in step. Adding properties lowers that risk only when their local demand conditions differ, so a count of buildings or a map measures very little. Concentration needs a separate share for value and for income, and debt balances need one too.
How a local shock becomes a portfolio funding gap
A local shock becomes a portfolio problem when several buildings depend on the same demand conditions and their loans come due while values are low.
Local oversupply and reduced demand act on both cash flow and value, as does competition from other available properties. The attractiveness of a particular building to prospective customers matters as well. Lower income cuts the cash available to carry debt. Lower value cuts the proceeds a lender advances against the same building, because a loan sized as a percentage of value falls when value falls.
Two tests answer different questions. An operating stress asks how far net operating income (NOI) drops. A refinancing test compares replacement proceeds with the balance due at maturity. A portfolio can show positive NOI after the shock and still owe more than the new loans will repay, because the lender sizes the loan on the stressed value.
Concentration links the two tests across buildings. When several holdings sit in one market, their repayment gaps arrive together, and the cash requirement is the sum of those gaps. A sibling property in the same market cannot be counted as a source of funds at its old value. Its own sale or refinancing proceeds have to be rerun under the same stressed conditions.
That is the mechanism behind a difficulty familiar to owners with one local market: when valuations fall across the whole portfolio, refinancing becomes hard because every asset is asking for the same thing at the same time.
The housing downturn showed the pattern at scale. Investors with extreme exposure to one property market or one property type were hurt most, and owners holding only Las Vegas properties took the heaviest damage when Nevada valuations crashed.
The diagnostic runs in a fixed order. Stressed income comes first, then value from the stressed cap rate. Debt supported by that value follows, and the balance due is set against it. The unfunded remainder is what is left. Skipping a step hides where the shortfall comes from.
“The housing meltdown and financial crisis exposed a number of investors who had an extreme exposure to a particular property market or type of real estate.”
Measuring concentration with a different denominator for each question
Concentration is a ratio, and the denominator changes the answer.
One large listed industrial owner reported that 30.0% of its consolidated operating properties, carrying $22.7 billion of gross book value, sat in California. Those same properties were 23.2% of operating square footage and 31.7% of NOI. One state produced three different weights, and the income weight was the largest of them.
The same owner measured exposure market by market against a 10% threshold of total assets and of revenues. Only one Southern California market crossed it. A state-level cut and a market-level cut therefore flag different things, and the unit of geography is a choice that belongs on the schedule.
A usable exposure schedule tags each property with its metropolitan area and submarket, then with its tenant industry. It records value and NOI in separate columns, along with debt balance and maturity. Each market’s share is computed per column. An NOI share above the value share means the market earns more than its capital weight suggests, so an income shock hits harder than the value share implies.
Equity adds a further layer. Equity equals value minus debt, so every dollar of value lost falls on equity. Borrowing magnifies the percentage change in equity value: the same percentage drop in value is divided by a smaller equity base. Two owners with identical property allocations carry different equity exposure when the debt behind those properties differs.
A market holding a large share of the debt balance needs its own refinancing test, with attention to when that debt matures. Value and NOI shares say nothing about whether the balance due is large relative to what the market will support.
Why the choice of city can outweigh the choice of property type
Return volatility differed more among cities than among property types in one sample, so a property-type allocation may settle less of the risk than it appears to.
In one sample period, Phoenix showed the highest standard deviation in returns at 3.3% and Minneapolis the lowest at 2.0%. Those are sample statistics. They show that city selection can matter inside a single property type and give no fixed volatility for either city.
Property-type spread leaves geography intact. An owner holding offices and warehouses in one metropolitan area still shares that area’s supply and demand conditions. Geographic spread leaves tenant-industry exposure intact in the same way: buildings in different cities that lease to the same industry respond to the same industry stress.
Building characteristics narrow the comparison further. A building’s location and design, together with its condition and tenant types, are the most important sources of risk and return for an individual investment, and not all of that risk can be diversified away. Two buildings in one metropolitan area can carry different exposure through design differences or through condition and lease roster.
Cycle analysis becomes more useful for forecasting and decisions as it moves from national levels toward metropolitan areas and individual property locations. A national property-type view cannot answer whether a subject building’s income holds up against the demand and competing properties in its own market.
Deal stage changes the question. An acquisition test compares price with assumed income. An operating test examines income under local demand and competition. A refinancing test carries the resulting value into debt sizing. The local assumptions need to stay consistent across all three.
“Differences in return volatility, measured by the standard deviation in returns over our sample period, are also materially larger among cities than among property types.”
Worked example: one market loses income across two $100 million portfolios
This hypothetical comparison puts two portfolios under the same local shock over one annual period. NOI throughout means in-place annual NOI after operating expenses, before capital expenditures and leasing costs, and with no debt service or sponsor adjustments. Each portfolio starts with $100 million of property value, $60 million of debt and $8 million of NOI. Fees are identical and left out of the arithmetic, and debt balances stay unchanged. Equity is $100 million minus $60 million = $40 million. The cap rate, which is one year of NOI divided by price, is $8 million divided by $100 million = 8%.
Portfolio A holds one concentrated market with $50 million of value, $30 million of debt and $4 million of NOI. Two other markets each hold $25 million of value, $15 million of debt and $2 million of NOI. Value adds to $50 million + $25 million + $25 million = $100 million, debt to $30 million + $15 million + $15 million = $60 million, and NOI to $4 million + $2 million + $2 million = $8 million. The concentrated market is $50 million divided by $100 million = 50% of value, $4 million divided by $8 million = 50% of NOI and $30 million divided by $60 million = 50% of debt.
The shock cuts NOI in one market by 20% and lifts that market’s cap rate from 8% to 10%. Lenders replace maturing debt at 60% of value in this illustration. In Portfolio A the NOI loss is $4 million multiplied by 20% = $0.8 million, leaving $4 million minus $0.8 million = $3.2 million. Value becomes $3.2 million divided by 10% = $32 million. That is a drop of $50 million minus $32 million = $18 million, or $18 million divided by $50 million = 36%, even though income fell only 20%.
Portfolio value is now $32 million + $25 million + $25 million = $82 million, and NOI is $3.2 million + $2 million + $2 million = $7.2 million, down $0.8 million divided by $8 million = 10%. Equity is $82 million minus $60 million = $22 million, a loss of $40 million minus $22 million = $18 million, or $18 million divided by $40 million = 45%. Property value fell $18 million divided by $100 million = 18% while equity fell 45%, because $100 million divided by $40 million = 2.5 times as much property as equity stands behind each dollar. Borrowing magnifies the percentage change in equity value.
At maturity the affected market supports $32 million multiplied by 60% = $19.2 million against $30 million owed, a gap of $30 million minus $19.2 million = $10.8 million. Each unchanged market supports $25 million multiplied by 60% = $15 million, exactly its balance. Total proceeds are $19.2 million + $15 million + $15 million = $49.2 million, and $60 million minus $49.2 million = $10.8 million confirms the portfolio gap.
Portfolio B holds the same $100 million, $60 million and $8 million across 4 equal markets. Each has $100 million divided by 4 = $25 million of value, $60 million divided by 4 = $15 million of debt and $8 million divided by 4 = $2 million of NOI. Each market is $25 million divided by $100 million = 25% of value.
The same shock hits one market. Its NOI loss is $2 million multiplied by 20% = $0.4 million, leaving $2 million minus $0.4 million = $1.6 million, and its value becomes $1.6 million divided by 10% = $16 million. Portfolio NOI is $1.6 million + $2 million + $2 million + $2 million = $7.6 million, down $0.4 million divided by $8 million = 5%. Value is $16 million + $25 million + $25 million + $25 million = $91 million, equity is $91 million minus $60 million = $31 million, and the equity loss is $40 million minus $31 million = $9 million, or $9 million divided by $40 million = 22.5%.
Replacement debt in the affected market is $16 million multiplied by 60% = $9.6 million against $15 million owed, a gap of $15 million minus $9.6 million = $5.4 million. Total proceeds are $9.6 million + $15 million + $15 million + $15 million = $54.6 million, and $60 million minus $54.6 million = $5.4 million confirms the portfolio gap.
Portfolio A loses $18 million of equity and faces a $10.8 million gap. Portfolio B loses $9 million and faces a $5.4 million gap. Both ratios equal 2: $18 million divided by $9 million = 2 and $10.8 million divided by $5.4 million = 2.
The protection depends on the shock staying local. If the same shock reached all 4 of Portfolio B’s markets, value would be $16 million multiplied by 4 = $64 million, replacement debt would be $64 million multiplied by 60% = $38.4 million, and the gap would be $60 million minus $38.4 million = $21.6 million. That is larger than the $10.8 million gap Portfolio A showed under its single-market shock. These results also assume the other markets stay unaffected and that debt is sized on value alone.
What geographic spread requires and what it cannot remove
Adding markets removes shared local exposure and adds underwriting and operating work at the same time.
Matching the diversification of the NPI benchmark by random selection may require as many as 100 properties. Expertise in selecting and managing the portfolio can reduce that number. The count is a feature of that exercise and gives no minimum for any actual owner.
Over a 20-year retrospective comparison, the ten stronger-performing metropolitan areas averaged 10.2% annual total return and the ten weaker-performing areas averaged 6.8%. The ranking is known only after the fact, so the gap describes what happened and sets no allocation rule.
Geographically diversified investing takes experience in unfamiliar markets and with the laws and regulations there. Someone has to judge competing supply in each new market and someone has to run the buildings after closing. A map with more pins and no answer to those two questions has measured spread without measuring capability.
Diversification has limits. Real estate can improve the risk and return trade-off of a broader portfolio, yet its correlation with other investments varies over time, so a historical correlation is a poor fixed input.
A shared-stress case belongs in the review once the allocation looks balanced. It shows the outcome when several markets weaken together. The distance between that result and a single-market stress measures how much of the apparent protection rests on independent local outcomes.
Pitfalls that make a spread portfolio look sturdier than it is
Counting addresses treats each property as an independent exposure. Buildings that depend on the same local demand conditions, or that lease to the same tenant industry, move together, and an ownership schedule sorted by address cannot show it. Grouping by the stress that drives income reveals the concentration.
A property-type allocation can overstate diversification when geography is missing from the analysis. Different uses in one metropolitan area share its location, and in one sample volatility differed more among cities than among property types. A location-by-type grid makes the overlap visible and leaves the size of each response as an explicit assumption.
A single denominator misleads. In the industrial owner figures above, one state held 30.0% of properties, 23.2% of square footage and 31.7% of NOI. A floor-area schedule understates the income weight by more than eight points, and a value-only schedule says nothing about the balance coming due.
Stressing NOI while holding the cap rate fixed answers only that scenario. Value equals NOI divided by the cap rate, so an income decline and a cap rate increase compound. In the worked example, income fell 20% and value fell 36%, which then set the replacement loan.
Counting a second property as a funding reserve at its pre-stress value breaks when it shares the affected market. Its stressed value, along with its own debt and sale costs, decides what remains, and in a concentrated portfolio the reserve shrinks in the same event that creates the gap.
Treating a new metropolitan area as a diversification certificate skips the work. Entering an unfamiliar market takes local experience with competing supply and with the laws and regulations there, and without it the underwriting of the new asset is weaker than the underwriting of the old one.
Checklist questions for a location concentration review
What share of portfolio value and of annual NOI comes from each market? A market whose NOI share exceeds its value share carries more income risk than its capital weight suggests.
Which properties depend on the same tenant industries or local demand conditions? A shared exposure can sit across different metropolitan allocations.
Which competing properties affect each building’s attractiveness to prospective customers? The relevant comparison is the particular building and its local alternatives.
How do design and condition differ among buildings in the same market? Those characteristics change individual exposure even when the metropolitan classification matches.
What value results when NOI falls and the cap rate rises together? Both assumptions have to be explicit before equity and refinancing effects can be measured.
How much replacement debt follows from the advance rate in the model, and when does each balance mature? The difference between proceeds and balance due is the modeled repayment gap.
What would an asset assigned a funding role actually produce under the same stress? Its stressed value, along with its own debt and costs, decides how much remains.
Who evaluates unfamiliar markets and who is responsible for operations there? Geographic expansion adds an execution requirement alongside the change in allocation.
What happens to the funding gap if several markets weaken together? The answer shows how much of the diversification result depends on separate local outcomes.