The Property Condition Assessment was paid for (probably between $5,000 and $15,000). It contains a professional engineer's read on every major building system, an estimated remaining useful life on each, and a forecasted capital reserve schedule that already laid out the work your underwriting assumed. The PCA does its biggest work in your first 90 days, calibrating what you bought against what's actually there. After that, it stays useful as a reference, but your own plan and quarterly walks become the active documents.

What the PCA actually gives you

What the PCA doesn't give you

It's a snapshot at one point in time. Three things to remember:

• Time has passed. The walk happened during diligence, often 60 to 120 days before you took title. Some items got worse. Some got fixed by the seller as a condition of close. You don't know which until you walk it yourself.
• The engineer didn't see everything. They saw a sample of units, the roofs they could safely access, the mechanical they could reach. There are systems they reported on by inference, not inspection.
• The cost estimates age fast. A roof cost from 2022 is not a roof cost in 2026. Use the PCA's costs as a starting point and update them with your current vendor quotes annually.

The PCA review at takeover

In the first 30 to 60 days of ownership, walk the property with the PCA in hand. This walk is your real-life calibration of every assumption you bought the deal on.

1. Verify each immediate repair item. Confirm what was closed by the seller, what's still open, and what was missed entirely.
2. Visually check every "fair" or "poor" rated system. The engineer's "fair" sometimes means "borderline poor." Your eyes matter more than the rating.
3. Confirm the reserve table years. If the PCA says roof in 2029, look at the roof and ask: do I believe 2029, or am I really looking at 2027? Annotate the table.
4. Build your initial 5-year plan from the annotated reserve table. This is the document you live with.

Your plan, not theirs

By year 1, your 5-year plan should be the document you actually work from, not the PCA. The PCA stays in the file as the original engineer's reference. Your plan reflects what you've done, what you've spent, what you've learned, and what the quarterly visits keep telling you. Items get added, items get moved, items come off, all based on ground truth. By year 3, the PCA is mostly archived. Your plan is the document everyone refers to.

• Which systems did you actually inspect vs. infer from age and visible condition?
• Where are you on the conservative end of remaining useful life, and where on the optimistic end?
• What didn't you see that you wish you had?
• If you owned this property for 5 years, where would you spend your first dollar?
• What's the one issue most owners miss on a property like this?

There's a version of this job where you run it from a laptop, get monthly reports, and only show up when something is on fire. That version is cheaper for your time. It also produces a predictable result: the property drifts. The drift is invisible at the monthly level and obvious at the annual level, and by then you're spending real money to recover ground you should never have lost.

The cadence

What the quarterly walk actually covers

Tool 1 at the end of this module is the fillable template. The zones it covers:

Where to look that PM won't show you

PM's instinct is to walk you through the property they're proud of. Your job is to also see the property they hope you don't notice. The places worth a few extra minutes:

• The roof. Every roof you can safely reach.
• The mechanical room behind the office that nobody opens.
• The dumpster enclosure. The dumpster pad. Behind the dumpster.
• The last building from the entrance. Furthest from where prospects park.
• The crawl space access. The attic hatch.
• The vacant units that have been vacant for more than 30 days. There's a reason.
• The make-ready board. Compare it to what the rent roll says is vacant.

Document discipline

A walk without a written record is half a walk. Three habits worth installing:

1. Photos with context. Every issue gets a photo with location noted. "South elevation, Building 4, second floor exterior wall, sealant failure at window head." Date stamped. These photos become the comparison set for the next quarterly visit.
2. The actionable list. Categorized as: PM to address (operating expense), Schedule for capex (project), Watch and reassess next quarter. The third category is where most things land. That's fine.
3. Same-day debrief with PM. Before you leave the site. Walk through the list, agree on what each item is, who owns it, when you expect it closed or addressed. Written confirmation within 48 hours.

Red flags on the walk

• "We just had that fixed" said about an obvious recurring issue.
• Mechanical rooms with no inspection tags or expired ones.
• Fresh paint in one specific area of a building. Investigate what's under it.
• Multiple work orders open more than 30 days on the same unit.
• The make-ready board doesn't match what's actually being worked on.
• Caulk and sealant failures around windows, especially on the upper floors.
• Standing water anywhere it shouldn't be. Foundation, slab, parking lot low spots.
• The property manager rushing you past a specific zone.

Curb appeal drives leasing. Property health protects the asset. They both matter and they compete for the same dollars. It's easy to default to one of them based on what kind of pressure you're under in a given month. Leasing slipping? Curb appeal. Owner asking about reserves? Property health. The default is not a strategy.

Trap 1: All curb, no bones

You repaint the monument sign, redo the landscaping at the entrance, refresh the leasing office, and the property looks great when you pull in. Three months later, the chiller fails in July and you spend the curb appeal budget twice. Six months later, the parking lot finally cracks open because you never did the seal coat. The leasing team is now showing units to prospects in a property that looks polished from the lot and dated from the second floor up.

Symptom: your annual capex is mostly cosmetic. Less than 30% goes to systems with more than 10-year useful life.

Trap 2: Solid bones, ugly face

The roofs are new. The HVAC is humming. The parking lot was sealed last year. The trash compactor is replaced. But the entrance sign hasn't been repainted in five years. The landscaping at the front is overgrown. The leasing office furniture came from the seller. Your prospect-to-application ratio is 30% below comp set. You're sitting on a healthy property that nobody wants to lease.

Symptom: your operational metrics (occupancy, traffic, application volume) keep slipping while your physical condition reports keep improving. You can defend this property to the engineer. You can't defend it to a prospect.

The 30-second test and the 30-minute test

Two checks worth running every quarter, both during the walk:

When budget is tight: how to choose

When the capex budget is genuinely constrained, the framework I use:

"Defer with intent" is the important phrase. It means: written into next year's plan, with a specific re-evaluation point, not silently kicked. The day you stop tracking deferrals is the day they compound on you.

The replacement tracker (Tool 3 at the end of this module) is the version I use. It's stripped down on purpose. Categories, install year, useful life, current condition, cost estimate, projected replacement year. That's it. A more complex tracker is a tracker that doesn't get updated.

Useful life vs actual life

Useful life is the manufacturer-rated lifespan under good conditions. Actual life is what your property delivers. The gap between them depends on:

• Climate. A flat roof in Phoenix and a flat roof in Boston live different lives. Same install. Same warranty. Different actual life.
• Original install quality. The roof someone rushed in November under a tight deadline doesn't last 20 years no matter what the warranty says.
• Maintenance discipline. An HVAC system that gets a real spring inspection every year outlives one that doesn't, by 3 to 5 years.
• Use intensity. A trash compactor at a 300-unit property runs through twice as many cycles as one at a 150-unit property. Same model, half the actual life.

After 2 years in a property, you'll know which way your actual life cuts vs. the engineer's estimate. Track it. Adjust the replacement year in the plan. Don't pretend the engineer's table is gospel forever.

The categories that matter

The plan tracks every major item that is expected to need replacement in the 10-year window. The list varies by property type, but for most garden and mid-rise multifamily, the spine is:

Why 5 years, not 3 or 10

Three reasons the 5-year window is the right one to manage:

• It matches a typical hold. Most value-add and core-plus holds are 5 to 7 years. A 5-year plan covers your hold either fully or substantially.
• It's long enough to see big items coming. Roofs, parking lot overlays, major HVAC replacements. Things you can't fund out of one year's operating cash.
• It's short enough to forecast credibly. The 10-year reserve study is an exercise in compounding uncertainty. The 5-year plan you can defend with current quotes and known condition.

The annual refresh

The plan gets touched throughout the year as items get done or things get spotted on the quarterly walks. Once a year, usually heading into budget season, you sit down with the full plan and do a real refresh. Four moves:

1. Items completed. Mark them done. Update the install year, reset the next-replacement clock.
2. Items that aged faster than expected. Move their replacement year forward. If you find a roof that's 12 years old and looking like 16, your replacement year is sooner.
3. Items that aged slower than expected. Push them out. Document why. ("Roof inspection clean, no granule loss observed, push to 2030.")
4. New items found. Things the PCA missed. Things that emerged. Add them with your own condition rating and cost estimate.

Reserve adequacy

If you have a funded reserve account, the plan tells you whether the per-unit-per-month reserve contribution is enough. Sum your 5-year capex forecast, then compare it to current reserves plus 5 years of contributions. The gap (or surplus) is your reserve adequacy. If you're short, the conversation with the owner about increasing reserves needs to happen before the year you actually need the money. If you're surplus, you have flexibility to pull a planned 2027 item into 2026.

The conceptual line

How they behave in returns

The hybrid trap

Here's the trap: calling deferred maintenance a value-add to justify it to the owner. "We're going to replace the roofs and pick up $X in rent because of better energy efficiency." The roofs needed to be replaced. Period. The rent lift, if it materializes, is a small bonus. Framing the roof replacement as a value-add does three damaging things:

• It distorts the value-add returns math, because the roof cost shouldn't be in the value-add denominator.
• It sets a fake expectation with the owner about rent lift that probably doesn't show up.
• It under-protects future deferred maintenance, because once you've spent the "value-add" budget on a defensive item, you don't have the budget for the actual value-add project.

Call the roof what it is. Replace it because the property needs it. Defend it with the 5-year plan, not with a rent assumption.

When they actually overlap

There are real cases where one capital event closes a deferred item and creates a value-add opportunity at the same time. The kitchen needs new appliances anyway (deferred), and you also reface the cabinets and replace countertops (value-add). The right move:

1. Track the costs separately. Appliances at the like-for-like replacement cost goes to deferred. Cabinet refacing and countertop go to value-add.
2. Underwrite the value-add portion against the rent lift. The deferred portion gets no rent credit.
3. Report both in their own buckets to the owner. Don't bundle them in the project recap.

Above-the-line vs below-the-line

First, the accounting fork worth being precise about:

The line between them is in your accounting policies and in tax law (capitalization thresholds, useful life rules). PM software typically codes them differently. If your team is putting capital items in R&M to make a tight month look better, you have a different problem (Module 02 covers it). The capex budget conversation assumes you've already drawn the line correctly.

Funded reserves vs cash flow capex

Most multifamily deals fund capex two ways:

• Funded reserves. A per-unit-per-month contribution (often required by the lender) sits in a reserve account. Capital items get drawn from it. Common range: $250 to $400 per unit per year, but it varies widely.
• Cash flow capex. Items funded out of operating cash, typically because they're smaller, more predictable, or didn't qualify for reserve draw under the lender's rules. Often things like turn-related items, smaller parking lot repairs, exterior painting.

The annual capex budget splits items by funding source. Big items go to reserves (when there's headroom). Smaller predictable items go to cash flow. The split affects what shows up in your distributable cash, which is what the owner actually cares about.

Timing relative to the operating budget

Module 03 walked the 7-month operating budget cycle. The capex budget runs parallel to it but on a different rhythm:

Owner approval thresholds

Most owners want eyes on capital items above a threshold, separately from the annual capex budget approval. Typical thresholds I've seen:

• Items above $25,000 or $50,000: individual approval before commitment.
• Anything over 110% of the budgeted amount on a single item: re-approval at the new number.
• Net-new items not in the approved annual plan, regardless of size: pre-approval.

Confirm your thresholds with the owner in writing. The middle of an emergency is the wrong time to discover the threshold was different than you thought.

The capex budget conversation

Different from the operating budget conversation in a specific way: owner pushback on operating expense tends to be on principle. Pushback on capex tends to be on selection. Two different conversations.

• Operating: "Why is R&M up 8% next year?" The conversation is about discipline and trend.
• CapEx: "Do we really need to do the roof on Building 3 next year, or can it wait?" The conversation is about timing and judgment on specific items.

Walk into the capex conversation prepared to defend each line individually. The 5-year plan and the quarterly visit notes are your defense. The PCA is your bedrock. "Here's what the engineer said. Here's what I saw last quarter. Here's the quote." That's the structure.

The seven standards

These are the behaviors that separate a portfolio with capex discipline from one without it. Use them as the bar for your own work and the work of anyone on your team.

Common AM failures

• The 5-year plan only gets touched at annual budget time.
• "We'll address that next year" with no documented decision and no flag in the plan.
• Treating the lender-required reserve study as the actual capital plan. (It's not. It's a financing document.)
• Approving capex items individually without checking against the 5-year plan first.
• Letting PM lead the capex conversation. PM brings scope. You bring sequencing and budget logic.
• Calling deferred maintenance a value-add to justify it to ownership.
• Spending the entire reserve in year 1 and 2 of the hold, then asking for capital calls in years 4 and 5.
• Skipping the quarterly walk because "PM said everything's fine."

Red flags from the field

• R&M expense trending up year-over-year while capex stays flat. You're paying twice for the same problem.
• Multiple emergency or "unplanned" capex events in a single year.
• The reserve balance is dropping faster than the 5-year plan projected.
• The annual capex budget is built from "what's leftover after operating" instead of from the plan.
• PM reports the same deferred maintenance item every quarter without a path to resolution.
• The property's condition rating drifts down between annual walks even though the plan is being executed.
• Vendor proposals on file are more than 12 months old but still being used for budgeting.

The conversation with PM

PM is your eyes on the ground between visits. The capex conversation with PM is recurring, not annual. Three rhythms:

• Monthly (in the weekly call, Module 01). Flag new capex candidates surfaced from work orders. PM raises them. You catalog.
• Quarterly (after the walk). Joint review of what the walk surfaced and where it changes the plan. Sequence next 12 months. Identify items needing vendor quotes.
• Annually (during budget cycle). Full capex budget conversation with sequencing, funding source, and owner-approval items flagged.

The conversation with the owner

Owners pay more attention to capex than operating expense in two situations: when they're writing a capital call, and when they're projecting exit proceeds. Your job between those moments is to make sure they know what's coming, why, and how it ties to the plan they approved.

A simple cadence works. Monthly capex line in the owner narrative (one paragraph: what got done, what's coming, anything material). Quarterly capex status with the owner: what got done last quarter, what's coming next quarter, any items pushed or pulled. Annual approval of the capex budget. Material item approvals as they come up. Nothing fancy. Just predictable.

What you'll need

• The Claude desktop app with Cowork mode.
• Your walk photos, saved together in one folder on your computer.
• Your walk notes: a simple list of what you observed and the action item for each.

Step-by-step

1. Take usable photos during your walk. Snap a photo of anything you'll want to show in the recap: damage, deferred maintenance, curb-appeal items, amenity issues. Get close enough that the problem is obvious. Don't worry about phone rotation; Claude straightens photos automatically.
2. Write your notes as a simple list. One line per item: the observation plus what you want done. Plain language is fine. Example: "Tile missing from entrance to the amenity area. See if construction has any extra."
3. Put all the photos in one folder. Drop every photo from the walk into a single folder (for example, a folder named for the property). Keeping them together is what lets Claude match them to your items.
4. Open Cowork and share the folder. Start a new chat in the Cowork app. When Claude asks for access to your files, point it to the folder you just made. You only need to do this once per session.
5. Paste your notes and make the request. Tell Claude the property name, paste your list of notes, and say what you want. Use the prompt template below. Mention that the photos are in the shared folder.
6. Answer the quick setup questions. Claude will ask a couple of questions up front: what format you want (Word doc, email, or both) and how to organize the items (a simple list, by category, or as a tracker with owner and due-date columns). The standard is a Word doc plus email text as a simple numbered list.
7. Review the photo-to-item matching. Claude places each photo under the item it belongs to and writes a short caption. Skim the draft and confirm everything landed in the right place. If a photo is matched to the wrong item, just say so.
8. Ask for any edits. Treat it like a draft. You can split one note into two items, reword something, flag a true priority, or add a note Claude should call out. For example, separating a cosmetic mildew stain from an actual water-intrusion issue.
9. Finalize and send. Fill in the walk date and the recipients (these are left blank on purpose), give it a final read, and send the email with the Word doc attached, or paste the email text directly.

Copy-and-paste prompt

Start your request with something like this, then paste your notes underneath.

Tips for the best result

• One folder per walk so photos don't get mixed between properties.
• Write the action, not just the problem. "Replace," "re-secure," "get a vendor out" tells the team what to do.
• Flag diagnosis-first items. For things like water intrusion, ask Claude to frame it as find-the-source-then-repair rather than just cosmetic.
• Ask for a tracker version if you want Owner / Due Date / Status columns to manage items over time.
• Reuse the format. Keep one finished file as your template and reuse the structure for every property.

Troubleshooting

• Claude says it can't see the photos. Make sure they're actually in the shared folder, then tell Claude they've been added. Re-dropping the files usually fixes it.
• A photo is sideways in the source. No action needed. Claude auto-rotates photos when it embeds them.
• Wrong photo under an item. Tell Claude which photo belongs where and it will swap them.

How the score is built

A weighted score that puts the heaviest weight on items where the risk of deferral is catastrophic, then on urgency, then on return:

Score = (Risk × 3) + (Urgency × 2) + Return weight

Return weight: High ROI = 3, Medium = 2, Low = 1, Defensive (no ROI) = 0

Tier cutoffs:
  Score ≥ 22 → Do now
  Score 16–21 → This year
  Score 10–15 → Next year
  Score < 10  → Defer (with documented re-evaluation)