Refurb programme scenarios: forecasting downtime, capex, and payback

Refurb programmes are where property portfolios win or lose returns. Done well, they lift rent, reduce voids, improve tenant quality, and strengthen long-term value. Done badly (or modelled badly), they create the most common "surprise" in real estate finance:

• NOI dips more than expected
• cash troughs arrive earlier than planned
• capex overruns compound at exactly the wrong time
• payback quietly stretches from "18 months" to "three years"

The fix is not a more complex spreadsheet. It is a repeatable scenario model that makes three things explicit and measurable:

1. Downtime (what income you lose and for how long)
2. Capex (what you spend, when you spend it, and what is committed)
3. Payback (what incremental NOI you earn back-and how confident you are)

This blog lays out a practical approach you can use across a single asset or a portfolio of SPVs-without turning forecasting into a bespoke project every month.

---

Why refurb scenarios fail in the real world

Most refurb modelling breaks for predictable reasons:

• The programme is treated as a capex budget, not a temporary operating shock (downtime + lease-up ramp).
• People forecast a single "void period" and forget the messy middle: phased works, tenant churn, rent-free periods, delays, and letting incentives.
• Capex timing is simplified into one month, so the model misses the cash trough.
• Payback is calculated on "headline rent uplift" instead of incremental NOI (after downtime, incentives, and operating cost shifts).
• Across many SPVs, cash is assumed to be "available" when it is actually restricted or trapped.

A refurb programme is a mini-investment cycle. You need an investment-grade view, not just a cost tracker.

---

The three core variables you must model

1) Downtime

Downtime is not one number. It is a chain:

• Offline time (unit unavailable during works)
• Lease-up time (available but unlet)
• Economic lag (let, but not earning full rent due to incentives / rent-free)

If you only model "physical occupancy," you will almost always overstate near-term performance.

Minimum viable downtime inputs

• units / sqm affected (and whether the programme is phased)

• Start date, end date (or duration) per unit/cohort
• Lease-up assumption (time-to-let)
• Incentives (rent-free weeks / lease incentives / step rent if relevant)

Output that matters: monthly lost revenue and monthly occupancy impact (average, not just month-end).

---

2) Capex

Capex needs two lenses:

• Total cost (budget realism)
• Cash profile (liquidity realism)

Most portfolios underestimate cash pressure because they treat capex like a total, not like a draw schedule.

Minimum viable capex inputs

• Budget by category (hard costs, professional fees, permits, contingency)
• Timing curve (e.g., 20%/50%/30% across months, or weekly if needed)
• Committed vs uncommitted capex (what is already contractually locked in)

Output that matters: monthly cash outflow and committed capex pipeline (what is coming even if you pause discretionary work).

---

3) Payback

Payback is only meaningful if you define it properly.

If you calculate payback using rent uplift only, you will overstate returns. In refurb reality, payback is driven by incremental NOI, not just incremental rent.

Minimum viable payback inputs

• Expected rent uplift (or rent per sqm / unit)
• Stabilised occupancy assumption post-refurb
• Incentives at re-let (rent-free, leasing fees)
• Incremental operating costs (sometimes refurb reduces maintenance; sometimes it increases service/amenity costs)

Output that matters: incremental NOI over time, payback month, and return-on-cost (where appropriate).

---

The minimum viable refurb scenario model

To make this scalable across assets and SPVs, keep the structure consistent. A simple model has five building blocks:

Block A: Asset + cohort register

Instead of forecasting every unit line-by-line, start with cohorts:

• "10 units in Building A - kitchens/bathrooms"
• "Floor 3 - common areas + lobby"
• "Retail unit - fit-out upgrade"

Each cohort has:

• size (units/sqm)
• start date / duration
• expected rent change
• capex budget
• letting assumptions

This is the key to scaling: you can model 5-20 cohorts per asset instead of 200+ units.

---

Block B: Downtime engine

For each cohort, translate programme dates into a monthly profile:

• % offline
• expected lease-up ramp
• incentive period (economic lag)

Best practice: model average monthly occupancy impact (day-weighted), not just "month-end occupancy," because revenue is earned over the month.

---

Block C: Capex cash curve

For each cohort, apply a draw profile and split:

• Spent (actuals)
• Committed (signed contracts)
• Forecast (planned but not committed)

This supports both planning and control: you can see what is still optional.

---

Block D: Stabilised income and NOI uplift

Estimate incremental NOI, not just rent.

A practical way to frame it:

• Incremental revenue = (new rent - old rent) - stabilised occupancy
• Less incentives = rent-free / leasing fees spread over the relevant period
• Less incremental OpEx (or add savings, if refurb reduces repairs/void costs)

This creates an "NOI ramp" rather than a single jump.

---

Block E: Outputs that decision-makers actually need

Minimum viable outputs per asset (and rolled up to portfolio):

1. Monthly NOI impact (dip + recovery)
2. Monthly capex spend and cash trough timing
3. Payback month (when cumulative incremental NOI covers cumulative capex)
4. Stabilised uplift (incremental NOI once programme is settled)
5. Downside sensitivity (what breaks if assumptions shift)

---

Payback: what to calculate (and what not to oversimplify)

A simple payback calculation (useful early)

• Payback (months) = Total capex - Stabilised annual incremental NOI - 12

This is quick, intuitive, and board-friendly.

What to add so it does not lie

Because refurb programmes have a dip before the uplift, payback should usually be based on cumulative cash/NOI, month by month:

• cumulative capex spend (cash out)
• cumulative incremental NOI (cash in / earnings uplift)
• payback occurs when cumulative incremental NOI - cumulative capex

Why it matters: two programmes can have the same stabilised uplift but very different payback because the downtime curve is different.

---

Common mistakes that distort refurb scenarios

Mistake 1: Treating downtime as a single "void period"

Reality: downtime is offline + lease-up + economic lag.

Best practice: show downtime in three layers so everyone can see what is driving the dip.

Mistake 2: Assuming capex is "evenly spread"

Reality: capex is lumpy, and programme cash curves often create a trough.

Best practice: model a monthly (or weekly for near-term) cash draw profile and track committed spend separately.

Mistake 3: Ignoring incentives and leasing friction

Reality: you often "buy" the uplift with rent-free and fees.

Best practice: explicitly model incentives so "economic occupancy" and NOI ramp reflect the real earnings profile.

Mistake 4: Mixing capex and opex inconsistently

Reality: one entity codes refurb invoices as repairs; another capitalises properly, and your portfolio story becomes noise.

Best practice: lock a consistent definition (capex vs opex, one-offs) and enforce it via mapped reporting.

Mistake 5: Modelling returns without modelling constraints

Reality: in multi-SPV groups, cash can be restricted or trapped, and funding decisions are entity-specific.

Best practice: tie refurb scenarios into SPV-level cash planning and facility constraints, not just portfolio totals.

---

Best-practice reporting: how to present refurb scenarios monthly

If you want refurb scenario reporting that builds trust with investors and leadership, publish a consistent "programme page" each month with:

1) Programme status

• cohorts in progress / completed / delayed
• % complete vs plan
• key risks (planning delays, contractor availability, letting pipeline)

2) Capex control

• budget, spent, committed, forecast
• variance to budget (with reasons)
• next 4-8 weeks expected cash outflows

3) Downtime and income impact

• units/sqm offline
• occupancy impact (average)
• revenue/NOI impact vs baseline
• lease-up progress and incentives exposure

4) Payback and return snapshot

• payback month (base case)
• stabilised incremental NOI
• downside case results (see below)

---

The "3 toggles" that make scenarios genuinely useful

Keep scenario planning simple but powerful. Three toggles usually cover most of the real risk:

1. Downtime sensitivity: +2 weeks offline and/or +1 month lease-up
2. Capex sensitivity: +10% cost and/or 1-month cash timing shift
3. Uplift sensitivity: -5% rent uplift and/or higher incentives

If leadership can see what happens under these toggles, you get faster, cleaner decisions:

• do we phase differently?
• do we pause discretionary cohorts?
• do we need equity injection vs intercompany funding?
• are we drifting toward covenant pressure?

This is where "what-if" scenario planning becomes a practical operating tool rather than a quarterly spreadsheet exercise.

---

Scaling this across many SPVs

Refurb programmes rarely live in one entity. They span multiple SPVs, multiple bank accounts, and multiple close timelines.

That is why scalable refurb scenario forecasting needs:

• a one-stop view across SPVs
• standardised chart of accounts + mappings so capex and one-offs roll up consistently
• portfolio-level and SPV-level reporting (so you can see where cash pressure actually hits)
• scenario planning that links downtime -> NOI -> cash flow -> returns

When that foundation exists, refurb scenarios stop being bespoke "model builds" and become a repeatable portfolio workflow.