Every car looks good when it's new. Every guitar is shiny hanging on the wall in the store. Every new TV looks spectacular when somebody first plugs it in.

I'm much more interested in what happens later. Does the car still start every morning at 100,000 miles? Does the guitar stay in tune, or does every piece of hardware slowly start giving up? When the TV develops a problem, can somebody repair it - or is it now a very thin, very expensive piece of garbage?

Things age. Things break. Things stop performing at 100%. Real estate does too.

I was looking at the numbers on a small multifamily property recently. At 100% occupancy, it looked great. Then we removed one tenant. Nothing catastrophic. No hurricane. No lawsuit. No fire. One empty apartment. And suddenly the deal looked very different.

That's when I started wondering whether we're asking the wrong question when we underwrite these things. We usually ask: "How much does this property make?" Maybe we should first ask: "How much does this property have to work?" Because 100% output, 100% of the time, isn't a realistic operating plan for almost anything.

If I buy four units and my deal only works when four checks arrive every month, what exactly did I buy? A good investment? Or a machine that has to operate at 100% capacity just to keep itself alive?

Of course we account for vacancy. We build reserves. We estimate repairs. Anybody with experience knows that. But I'm talking about something a little broader. How gracefully does the investment age? One tenant leaves. Another pays late. The AC dies in August. Insurance goes up. Property taxes get reassessed. The roof you thought had seven years left has three. None of those things is some bizarre black-swan event. That's just Tuesday.

A good car doesn't have to be a car that never breaks. Maybe it's a car that can lose an alternator without destroying the engine, can be fixed by a normal mechanic, and has parts that don't require importing something from Stuttgart by helicopter. Maybe a good property should be judged the same way. Not: "Can it operate perfectly?" But: "How badly can it operate and still be okay?"

Now, there's a danger on the other side. Stress-test a deal hard enough and you can kill anything. What if one unit is vacant? What if two are vacant? What if the roof fails? What if insurance jumps 30%? What if rents fall? What if interest rates rise? What if the neighborhood turns?

Keep going and eventually every spreadsheet tells you to put your money under the mattress. Risk is part of investing. You don't get paid for eliminating it. So I'm not looking for a property that survives the apocalypse. I'm trying to figure out what percentage of normal performance should be enough.

If the property needs 100% performance to work, that scares me. If it works at 90%, I'm more interested. If it can take a vacancy, an unexpected repair and a bad month without forcing me to reach for my personal checkbook, now I'm listening.

Maybe that's one of the better definitions of a durable investment: it doesn't have to work perfectly in order to work. Because sooner or later, something will break. Somebody will leave. Somebody won't pay. Something will cost more than we expected. The question isn't whether that will happen. The question is: "How much can this thing stop working before the investment stops working?"