Commercial Markets: Absorption, Leases, Offices, Industrial, Retail, and Hotels

A full office building, a busy warehouse, and a sold-out hotel all use space, but their demand drivers and adjustment speeds differ. An office tenant may commit for years; a hotel sells a room for one night. Commercial-market analysis starts with a common accounting framework, then asks what produces demand in each sector and how quickly leases, rents, and construction can respond.

Learning goals

Calculate net absorption, distinguish leasing turnover from occupancy growth, interpret rent indices, and connect sector-specific demand to office, industrial, retail, and hotel space. Explain why a statistic can improve while another deteriorates.

1. Start with the market’s stock and occupied space

Track one consistently defined geography, property segment, and period. Mixing prime offices with all office space, or quarterly absorption with annual construction, can produce misleading comparisons.

Symbol	Measure	Typical unit
S_t	Completed stock at period end	Square feet or square metres
v_t	Vacancy rate	Fraction of eligible completed stock
C_t	Construction starts during the period	Floor area per period
Ab_t	Net absorption	Change in occupied floor area during the period
L_t	Average lease term	Years, if annual turnover is being estimated
N_t	Renewal rate	Fraction, preferably area-weighted for floor-area calculations

Occupied space is O_t = (1-v_t)S_t. Therefore:

Ab_t = (1-v_t)S_t - (1-v_(t-1))S_(t-1).

Net absorption is a change in occupied space, not the number of signed leases or the amount of new construction. Vacancy is also different from availability: an occupied unit offered for future lease or sublease may be available without yet being vacant. Data providers’ conventions must be checked.

A market that grows and becomes more vacant

Last year stock was 1,000,000 square feet with 10% vacancy, so occupancy was 900,000. This year stock is 1,100,000 with 15% vacancy, so occupancy is 935,000. Net absorption is positive 35,000, even though vacancy rose. New stock outpaced the gain in occupied space.

Measure	Earlier year	Later year
Stock	1,000,000	1,100,000
Occupied space	900,000	935,000
Vacant space	100,000	165,000
Vacancy rate	10%	15%

The stock identity is current stock = previous stock + completions - removals, with conversions and reclassifications treated consistently. Starts enter a construction pipeline; they are not immediate additions to usable stock. A formula using starts from an earlier period assumes a delivery lag and needs adjustment for delays and cancellations.

Teach it back: Why is rising vacancy compatible with positive net absorption?

Self-check: demand growth versus supply growth

Occupied space can increase while completed stock increases even faster. In the example, occupancy rises 35,000 but stock rises 100,000, adding 65,000 to vacant space.

2. Leasing turnover is not net absorption

If a firm moves from one existing 10,000-square-foot office to another of the same size, the destination records leasing activity while the origin loses a tenant. After the move, market-wide occupied space can be unchanged. Renewing in place can also generate a lease transaction without adding occupied space.

The supplied lecture turnover approximation is:

Gross absorption proxy = S_t(1-N_t)/L_t.

This estimates annual non-renewal turnover from stock, renewal probability, and average lease term under stylised steady-state assumptions. It is not a universal identity for all gross leasing. Including renewals, using occupied stock rather than total stock, counting new entrants, pre-leases, expansions, or weighting leases differently changes the calculation. A new lease can also be signed well before occupancy begins.

With stock of 1,000,000 square feet, renewal rate 60%, and lease term 5 years, the proxy is 80,000 square feet per year. An occupied-stock version at 90% occupancy would give 72,000, illustrating why the denominator convention matters.

The lecture’s implied average lease-up time is:

v_t / [(1-N_t)/L_t].

At 10% vacancy, it gives 0.10/(0.40/5) = 1.25 years. This is vacant stock divided by a particular estimated flow, not a direct observation of each listing’s marketing time. It requires compatible units, a reasonably stable flow, and an appropriate mapping from turnover to vacant-space take-up. If renewals are 100%, the proxy flow is zero and this ratio cannot supply a finite lease-up estimate; it does not prove that actual leasing ceases.

3. Measure comparable rent, not just the average deal

A market’s average signed rent can rise because newer, better-located buildings account for more transactions, even when like-for-like rents do not change. A hedonic rent index controls observed features such as floor area, age, submarket, lease length, and contract structure while estimating time effects.

Also compare rents on a consistent economic basis. Headline rent, initial rent, average rent over the term, and effective rent after concessions can differ. Rent-free months, landlord improvements, escalation clauses, operating expenses, and gross versus net leases matter. Hedonic controls reduce some composition effects but cannot remove unobserved quality or contract differences automatically.

Teach it back: If this quarter’s leases are mostly in new buildings, why might the average rent overstate general market rent growth?

4. Office demand is employment times space intensity

An intuitive starting point is occupied office space = office-using employment x space per relevant worker, with adjustments for geography, vacancy, and the definition of a worker. The supplied FIRE category means finance, insurance, and real estate; business services include functions such as legal work, consulting, advertising, and data processing. Actual office usage varies within each category.

Suppose 10,000 office workers require 150 square feet each: demand is 1,500,000 square feet. Employment rises 10% but space per worker falls 10%. Demand becomes 11,000 x 135 = 1,485,000, a 1% decline. Employment growth alone does not establish space-demand growth.

Rental elasticity describes how desired space responds to rent, holding other inputs fixed. Technology, hybrid work, layout, and business organisation can shift the relationship itself. Assigned workers, daily attendance, desks, and occupied space are different measures; a hybrid arrangement does not imply floor area falls one-for-one with days worked remotely.

5. Lease length, block size, and the lease-or-own choice

A long lease commits the parties to a future rent path. Expected market rents matter, but so do escalation, incentives, credit risk, flexibility, break options, and the value of income certainty. A long-term lease rate is not simply today’s spot rent multiplied by a fixed term premium.

The supplied roughly 2%+ annual-rent difference per additional lease year, its relationship with vacancy, and larger-block discounts are historical lecture findings or hypotheses, not universal pricing schedules. If a log-rent regression coefficient on term were 0.02, one extra year would imply exp(0.02)-1, approximately 2.02%, conditional on its controls. That is not an extra 2% annual escalation automatically imposed by the contract.

Larger suites can spread leasing costs over more area or attract a negotiated discount. Scarce contiguous blocks can instead command a premium. Always compare like-for-like space and lease obligations.

For leasing versus ownership, the shorthand i_c P versus R compares an annual capital charge on purchase price P with annual rent R. It is only a first screen. A proper comparison discounts lease payments and ownership cash flows over the same horizon, including maintenance, taxes, acquisition and sale costs, residual value, financing, and flexibility. A property’s cap rate and a corporate discount rate represent different cash flows and risks and should not be substituted blindly.

Custom facilities also create hold-up risk: after a tenant makes a hard-to-move investment, bargaining over future rent or renewal can become difficult. Ownership can reduce one dependency while creating capital commitment and disposal risk. Lease accounting and tax treatment depend on the actual framework; do not assume leases automatically stay off the balance sheet or that every rent payment has identical tax treatment.

6. Industrial demand: track the production and logistics system

Manufacturing, distribution, and R&D properties serve different processes. For warehouses, inventory levels, handling, storage density, throughput, location, and delivery requirements matter. For manufacturing, output, capital intensity, production design, and supply chains matter. Worker count alone can be a poor proxy when automation changes productivity.

Inventory velocity is often measured by sales divided by inventories, with a consistent valuation basis and period. Just-in-time practices may reduce average inventory for a given sales flow, while faster throughput, distribution requirements, and resilience strategies can create other space needs. Faster inventory movement does not mechanically imply more or less warehouse floor area.

Two useful identities

Warehouse space/worker = (space/inventory value) x (inventory value/worker).

Manufacturing space/worker = (space/output) x (output/worker).

These level identities imply exact additive changes in logarithms, not exact addition of ordinary percentage changes.

Supplied component changes, interpreted as ordinary percentages	Exact combined change	Why the supplied sum differs
Warehouse space per inventory value falls 60%; inventory value per worker rises 70%	0.40 x 1.70 - 1 = -32%	-60% + 70% = +10% omits the interaction
Manufacturing output per worker rises 70%; space per output falls 30%	1.70 x 0.70 - 1 = +19%	+70% - 30% = +40% omits the interaction

If the lecture figures are intended as log changes, their sums are 0.10 and 0.40 log points. Those imply ordinary growth of about 10.52% and 49.18%, respectively, rather than exactly 10% and 40%. Label the measurement convention before interpreting the figures; the ordinary-percentage claims cannot all be exact simultaneously.

Inventory value also changes with prices and product mix. A physical-space analysis should distinguish nominal value growth from real quantities and storage requirements.

Import flows can support port-linked storage and distribution, but the supplied claim that imports create far more warehouse demand than exports is a context-dependent empirical proposition. Goods, routes, handling, dwell time, inland distribution, and re-export patterns matter. A port can be a pass-through node without holding a proportional share of regional inventory.

Teach it back: Why can manufacturing space per worker rise even when each unit of output needs less space?

7. Retail: spending and physical space are different demands

Retail spending depends on incomes, households, tourism, prices, product categories, and shopping channels. The supplied income elasticity of about 0.8 is a historical estimate, not a fixed law. Locally, an elasticity of 0.8 means a 1% real-income increase is associated with approximately 0.8% higher real retail spending, holding other inputs fixed. Under a constant-elasticity model, a 10% increase implies 1.10^0.8-1, approximately 7.92%.

Cross-city correlations involving labour-force participation or climate do not establish universal causal effects. Income, demographics, visitors, local prices, and the types of retail measured can confound the relationship.

Spending does not translate one-for-one into store area. Sales per square foot, format, rent affordability, customer visits, and lease commitments matter. Online shopping can shift demand toward fulfilment and logistics while physical shops adapt through collection, returns, showrooms, and services. Some formats lose space demand while others gain; apply the retail location framework to the specific shopping journey.

8. Hotels sell room-nights, not just rooms

A hotel reprices and sells access on a much shorter cycle than a multi-year office lease. Keep the period consistent: a 100-room hotel over 30 days has 3,000 available room-nights, assuming every room is available throughout.

Metric	Formula
Supply	Available room-nights in the period
Demand	Sold room-nights in the same period
Occupancy	Sold room-nights / available room-nights
ADR, average daily rate	Room revenue / sold room-nights
RevPAR, revenue per available room	Room revenue / available room-nights = ADR x occupancy

These familiar STR-style metrics use room revenue, not necessarily total hotel revenue including food, events, and other services. RevPAR is a revenue measure, not profit. Treatment of unavailable rooms and other reporting details should remain consistent across the comparison.

A hotel example

Our 100-room hotel sells 2,100 room-nights over 30 days and earns 252,000 of room revenue. Occupancy is 70%, ADR is 120, and RevPAR is 84. Check: 0.70 x 120 = 84, and 252,000/3,000 = 84.

If discounting raises occupancy to 80% but ADR falls to 100, RevPAR becomes 80, lower than before. More occupied rooms do not necessarily mean higher revenue per available room or profit. Serving additional rooms also has operating costs.

9. Recoveries do not follow one fixed sequence

Following a demand shock, a hotel may rebuild occupancy through discounts before restoring rates. That is one plausible recovery pattern. Events, customer segments, supply changes, inflation, pricing strategy, and competition can produce others. Room rates may recover before occupancy in some settings.

Mean reversion is a property of a specified time series relative to a benchmark; it does not guarantee a return to a particular pre-shock occupancy or room rate. Structural change can move the benchmark. Separate a historical lecture pattern from a forecast about a new shock.

Across commercial sectors, ask how quickly contracts reset, how fast new space arrives, and whether the underlying demand driver has changed permanently. Office leases can delay cash-rent adjustment, while daily hotel pricing can respond rapidly; neither feature removes operating risk or guarantees efficient forecasting.

Feynman study lab

Explain it simply

Start with how much completed space exists and how much is occupied. Net absorption measures the change in occupancy; leasing volume also includes churn. Each property sector then needs its own demand story: office workers and work patterns, industrial production and logistics, retail spending and shopping channels, or hotel room-nights. Compare consistent rents and revenues before drawing conclusions about value or profit.

Teach it back without looking

Explain how the office example has positive absorption and rising vacancy. Then explain the hotel with higher occupancy but lower RevPAR. Finally answer:

Why can a market report lots of leasing activity without a comparable increase in occupied space?

Check your explanation after you try

Some leases renew existing occupancy, and some relocations fill one building while emptying another. Pre-leases can also precede occupancy. Gross transaction activity therefore differs from the net change in occupied market stock.

Transfer questions

1. Stock rises from 500,000 to 550,000 square feet while vacancy stays at 10%. What is net absorption?
2. Using the lecture proxy, S = 600,000, N = 50%, L = 3 years, and v = 10%. Estimate annual turnover and implied lease-up time.
3. Office employment rises 20% while space per worker falls 10%. What is the exact space-demand change?
4. Space per inventory unit falls 20% and inventory per worker rises 50%. What happens to warehouse space per worker?
5. A hotel has 2,000 available room-nights, sells 1,500, and earns 180,000 room revenue. Find occupancy, ADR, and RevPAR.
6. Does a longer lease always command a higher annual rent after all concessions and options are considered?

Check the transfer questions

1. 45,000: occupied space rises from 450,000 to 495,000.
2. 100,000 square feet per year, with implied lease-up time 0.6 years, subject to the proxy’s assumptions.
3. +8%, since 1.20 x 0.90 = 1.08.
4. +20%, since 0.80 x 1.50 = 1.20.
5. Occupancy 75%, ADR 120, and RevPAR 90.
6. No. Expected rents, incentives, risk, flexibility, and market conditions can change the term relationship.

Find the gap and try again

Identify where you hesitated: stock versus flow, turnover versus absorption, asking rent versus effective rent, ordinary versus log changes, or occupancy versus RevPAR. Rework that example and explain it again without formulas. Change one input and predict the effect before calculating. Tomorrow, reconstruct the absorption identity and explain one demand driver and one measurement limitation for each sector.