Santam's catastrophe and large-loss bill went from R144 million to R1.5 billion in six months, net of what its reinsurers carried. It was mostly fire, and their underwriting margin fell to 8.1% from 11.3%.

The nuance matters. The property book stayed profitable and everyday claims behaved. What blew out was the catastrophe layer, and the reserve buffer thinned from the 91st percentile to the 87th. That pairing is what makes an insurer reprice, and when the market leader moves, the rest will follow.

Expect premiums, excesses and perils terms to shift at your next renewal, hardest on fire and flood in exposed nodes. A higher premium plus a higher excess permanently cuts your income. An under-insured building is worse: that is a balance-sheet event, not an expense line.

The Play:

Build one line per building: sum insured, replacement cost, excess, perils covered, last valuation date, fire-certificate expiry. You already hold all six, scattered across a broker file, a valuation report and a facilities folder. Side by side, two things surface that never show up separately: buildings still insured at a number set years ago, and certificates that have quietly lapsed. Sort by the gap between sum insured and replacement cost and you have your renewal priority list, before your broker sets it for you.

Over the decade to 31 August, the big diversified giants paid almost nothing. Hyprop returned 0.7% a year, Redefine 3.4%, Growthpoint 4.9%. Over the same ten years, according to the SA REIT Chart Book, the specialists ran away with it: Fairvest A returned 17.6%, Stor-Age 13.8%, Vukile 12.6%.

The funds that picked one thing and mastered it beat the own-a-bit-of-everything giants three and four times over. Fairvest in township and rural retail, Stor-Age in self-storage, Equites in logistics. The reflex that a big diversified name is the safe choice is not supported by ten years of its own numbers.

Do not read it as "small is good". Over the same decade, Delta lost 19.7% a year and Accelerate 17.8%. What paid was focus, not size.

The Play:

This applies whether you own shares or buildings. Tag every property you hold with three things: asset class, node, and year bought. Then work out what share of your income comes from your single biggest class-and-node pairing. Most owners have never calculated it, and assume they are more spread or more focused than they are. Under about 40%, you are a generalist competing against people who know their patch better. The number tells you whether to go deeper or sell the outliers.

Every landlord reads a sublet as distress and a fully-let schedule as safety. MetLife's own office portfolios say the opposite: its worst investments held physically empty space, while buildings with subleased space surprised to the upside, as subtenants converted into direct tenants.

That is one firm's experience, not a market study, but the mechanism is measured everywhere. CBRE puts office utilisation at 53%, and organisations acting on their own data cut floor area by 22% on average. Galetti puts the local range at 55% to 65%.

Put a number on it. A tenant in 5,000m² at R150/m² pays you R9 million a year. Hand back 22% at renewal, and R2 million of income disappears. At a 9% yield, that is roughly R22 million off the building's value, taken by a tenant who never missed a payment.

The Play:

Rank your tenants by use, not rent paid. You already hold the inputs: access swipes, parking take-up, after-hours HVAC requests, per-tenant electricity and water. Pull them into one view per tenant so that the quiet ones stand out. Discount swipes by 15% to 25%, because a swipe records someone entering a building, not using a desk. For any tenant far below the rest, run that 22% number against your own income.

SA property valuations just went up 6.4%. Fortress reported valuations up 6.4% and payouts to shareholders up 14.2% to R2.23 billion for the year to June, beating its own guidance, then raised next year's forecast. The share firmed 5.26% on the day. If your own valuations have not moved, ask your valuer why.

A court threw out a municipality's entire valuation roll. The Johannesburg High Court found Lesedi's 2024 to 2029 roll unlawful for failing to follow the Municipal Property Rates Act, then suspended it for six months so the municipality could fix it. If your rates bill looks wrong, the roll itself is challengeable, not just your valuation.

You can store 500,000 litres without building a reservoir. A Cullinan father and son have taken their inflatable bladder reservoir global. It needs no concrete base, sits on any reasonably flat surface, and folds into a bakkie when empty. Closed system, so no evaporation or algae. Backup water for a building without a construction project.

A bank is buying into buildings, not just lending on them. Nedbank Property Partners has put R370 million of equity and mezzanine funding into a joint venture with Stor-Age to build four self-storage facilities. Worth asking about when your development is short of capital rather than short of a loan.

Cheaper money has landed for green buildings. Investec has taken a $200 million (R3.5 billion) loan from the IFC to fund projects that meet recognised green-building standards, across residential, retail, industrial and mixed-use. A green mortgage for buyers of certified units is planned, so the discount may reach your buyers too.

A mining group monitoring a tailings dam over 400+ hectares had to check water levels by manually walking the site over the course of a month. Missing something could cost up to R10 million in environmental fines, plus penalties of R2 million a day.

The fix was activating sensors across the dam to feed into one dashboard with automatic alerts. The month-long blind spot closed to real time while manual effort fell by more than 90%.

The dam did not change. What changed is that they can show, at any moment, exactly what they knew and when.

Built. A newsletter by The Awareness Company.