> But in a saner world, housing would depreciate in value over time in proportion to its use and maintenance, like other durable goods.

Without any external influencing factors, a house would not typically depreciate. So a depreciating house isn't normal, there is something external causing it (like the city is going broke, lost most jobs, or environmental factors make the area bad to live in, etc).

But if all is well, a house will not depreciate if it is lived in and maintained. A house can last centuries and inflation means building an equivalent house is always more expensive later than it was to build this one.

So it is not natural to expect a house to depreciate.

That would make sense in a world where the surrounding location is completely static, but generally as a town/city grows it's much more desirable - more work for better wages, more business opportunities, economies of scale, better education... A world that doesn't respond to that with increasing price doesn't seem sane. Imagine that you start with a house in the middle of nothing. 20 years later there is a major city around you - and the house should be cheaper than it was when there was nothing? What about new housing around your house, should it have the same nearly zero price too? What even is depreciation in case of housing - my grandmother lives in a 120 year old house and I'm pretty sure my grandchildren will live there too.

Maintaining houses cost time (aka money) and money.

Not so much it'd be visible in the price - maybe 5-10% up or down, more pronounced in remote locations. The location is the important factor. Realtors work with square area pricing set for entire towns or municipalities.

That REALLY depends on the type of housing, not only on location. And it also depends on the interest rate relative to inflation.

In fact, it's really similar to dividends on stocks, just with cost instead of profit.

For a house in a suburb or rural area, maintenance (enough to keep the house in the same condition you bought it in) can easily cost as much per year (on average) as the cost-of-money (interest rate - inflation) for the debt.

And in some case much more than that.

If one such house is twice as expensive to maintain (over time) due to differences in building materials, environmental conditions, size/geometrical factors, etc, the cost of owning the house over a time period may easily be 25-50% higher for the most expensive compared to the least expensive if we assume the same purchase price.

but the land it's built on is an appreciating asset.

Purchasing a single piece of land with the expectation that it will increase in value, is similar in many ways as to do so with single stocks.

Land and stock tend to go up in value. But land also sometimes go to 0, just like stocks.

With the booming American car industry in the 50's and 60's, who would have thought that houses in Detroit could go from having a premium price in 1970 to be sold for $1 40 years later?

Who's to say SF isn't going to be next?

Short of it being on a superfund site most residential land will never have a value of 0. It's called REAL estate for a reason. Partial ownership in a company can evaporate, but 405 park west, manhattan nyc will always be there, at least for the next million years.

Some actually sell for $1 ($0 may be technically more difficult), some may sell for less than 10% of what it was once considered worth, which is close enough to 0 for the difference to not matter much.

A single peice of land may or may not appreciate. but a diversified portfoilio of land will always appreciate if the last 10,000 years has anything to say about it.

The rockefellers and other "Old Money Family's" have the three rules to building multi-generational wealth. Land, Art and Gold.

Even for detroit, land prices have only increased since the 80's[0] and over long time horizons, i'm sure will be back in line with other "Single peice land bets".

[0] https://fred.stlouisfed.org/series/ATNHPIUS19804Q

> A single peice of land may or may not appreciate. but a diversified portfoilio of land will always appreciate if the last 10,000 years has anything to say about it.

I believe there are large parts of Europe that would take centuries if not a millennium to reach the peak value it had during the Roman Empire (measured vs gold).

In general I agree, though. A portfolio is fine.

Most people tend to have most of their "savings" in a single property though.

As an example, what do you think this $2500 property cost when it was new? (Edit: make that 1965, since it's quite old.)

https://www.zillow.com/homedetails/3226-Columbus-St-Detroit-...

a person buying a condo in Vancouver generally isn't purchasing land, and if they are, then yes, the land might appreciate even as the thing built on it degrades over time due to a bunch of natural processes (e.g. wood decays)

someone buying a condo is absolutely buying land. That land just is split up across x number of condos. The appreciation that you get on the Condo, is the fact that the land is usually in a very ideal location. That's when the economics of "Going up" even makes sense.