How are you forecasting reserves beyond the generic 1-2% rule?

Question

The standard advice is set aside 1-2% of property value per year for maintenance and capex. It's a fine starting point but it treats a property with a brand new roof and HVAC the same as one where everything's 15 years old.

I switched to forecasting reserves system by system, roof, HVAC, water heater, appliances, based on actual age and typical lifespan, and inflating replacement costs forward. It's more work upfront but the number is grounded in something real instead of a rule of thumb that was probably written for a different market and decade.

the part that surprised me most was how uneven the "healthy" number turned out to be year to year. Flat percentage budgeting smooths everything out, real forecasting shows the lumps, a big year when three systems are aging out at once, followed by quieter years. curious how others here are handling this, especially anyone with a bigger portfolio where doing this per-property gets tedious. Is there a shortcut that still respects the real condition of each asset, or does it just come down to spreadsheet discipline?

Dan Handford · Answer

System-by-system forecasting is much more useful than a flat percentage. I would add two layers to it. First, maintain a minimum liquidity floor for vacancy, insurance deductibles, and surprises that are not tied to a component's expected life. Second, stress test correlated expenses. A roof, HVAC unit, and turnover can all hit in the same year even if the average annual forecast looks comfortable. For a portfolio, I would forecast each property's components, then compare the combined five-year schedule to the cash reserve available at the portfolio level. That helps separate an adequate long-term reserve plan from an actual near-term liquidity plan.

Chris Seveney · Answer

This is how we forecast as well.Anticipated life expectancy and cost to repair/replace and model it out.

Victor Patel · Answer

I think you’re looking at it the right way. The 1–2% rule is useful for a quick back-of-the-envelope analysis, but it can be misleading when you’re actually underwriting a property.

I like the idea of treating reserves more like a sinking fund by major system. Roof, HVAC, plumbing, parking lot, appliances, etc. Then you can estimate remaining useful life, replacement cost, and inflation rather than assuming every property has the same needs.

For a larger portfolio, I think the shortcut is to standardize the inputs rather than standardize the reserve percentage. Have a simple schedule for each property showing the age and expected replacement year for the major components. Then roll those numbers up across the portfolio.

You still get the uneven years you mentioned, but you can see the capital requirements coming instead of being surprised by them. In my experience, the biggest mistake isn't having the "wrong" reserve percentage—it's not knowing which major expense is coming next.