If Alice buys a home for $100k and it appreciates $1M over ten years, and Bob rents for ten years for $100k, total. Alice has $1M more than Bob does. That isn't nullified if Alice chooses to buy a $1M home next. She can also go buy another $100k home instead and pocket the $900k. Bob, obviously, cannot pocket anything.

> She can also go buy another $100k home instead and pocket the $900k

No, that's my point. A home that cost $100k when Alice bought hers, now costs $1M, therefore if Alice sells her home for $1M and buys another one for $1M, she makes a profit of $0.

Bob, with a net worth $1M lower than Alice, would probably disagree that she has not profited.

Only if you do fraudulent accounting. Mark-to-market is limited to assets held for trading, as far as I can tell, and for a good reason.

If you sell your home for $1M more than you bought it, it would be fraud if you didn't consider it a gain.

If you sell it and become homeless, yeah. Otherwise you'll be buying a $100k home for $1M, which cancels out all the profits that you made from the sale.

That's not how profits work. Or accounting in general. If you sell for +1M and then buy a different home for $1M, you've recorded a gain of $1M. It doesn't matter if you spend it on another house or don't. Your balance of cash may stay the same, but you've still profited. And the gain is taxable.

That's the accounting profit. Suppose you keep the house, your accounting profit would be zero. If instead you sell the house, and buy it back immediately for the same price, now you have an accounting profit. From this example, it's obvious that the accounting profit can be misleading. What matters is not the accounting profit but the economic profit (aka actual profit).

This assumes you continue to own a home, and own a home in the same market.

You can downsize, rent, or relocate to realize profit.