> It's indeed a flawed comparison but marketcap is still the best we have now.

I disagree. Many exchanges have public order books, so you can see exactly how much the price would slip if you sold, e.g., $1m worth of a coin. This is what is relevant for something functioning like money, not the last traded price multiplied by number of coins in existence.

> I disagree. Many exchanges have public order books, so you can see exactly how much the price would slip if you sold, e.g., $1m worth of a coin.

Yeah, but if you wanted to sell $1M worth of coins you'd probably use an auction at Gemini or elsewhere so you don't move through more than 2 or 3 price levels of the book.

You'd also probably enable some sort of smart selling algo for the same reason were you try and move $1M worth of BTC not at auction.

You could make the same argument for Gold or Platinum being over valued based on slippage from open interest at a point in time.

> Yeah, but if you wanted to sell $1M worth of coins you'd probably use an auction at Gemini or elsewhere so you don't move through more than 2 or 3 price levels of the book.

Selling $1m worth of bitcoins on GDAX, Bitstamp and Bitfinex results in a slippage of 0.2864%, 0.3527%, and 0.2910%, respectively.

As far as I can gather from Gemini's fee schedule[1], they take a >0.40% fee even if you've traded for more than $10m in the past 30 days, so I'm not sure it's worth it for a $1m market sell.

[1] https://gemini.com/fee-schedule/#fee-schedule

> You could make the same argument for Gold or Platinum being over valued based on slippage from open interest at a point in time.

I disagree. I'm not arguing that anything is over-valued. Simply that market cap as a measure of value of commodities does not make sense, because of the great difference in marginal utility between them.

Ah I see, I don't have a background in finance, how do you calculate that price slip?

You look at the public order books. So, for example, the order book for a particular exchange's BTCUSD market might contain the following buy orders (from various people who've deposited USD and wish to buy BTC):

    BUY 10 BTC @ 15,000 USD
    BUY 10 BTC @ 14,900 USD
    BUY 50 BTC @ 14,800 USD

If you sell 67.5 BTC into this market, you will sell the first 10 BTC for 15,000 USD each, the next 10 BTC for 14,900 USD each, and the remaining 47.5 BTC for 14,800 each, netting you a total of 150,000+149,000+703,000=1,002,000 USD. The average price would then be 1,002,000/67.5=14,844.4444, which means you would incur a slippage of (15,000-14,844.4444)/15,000=0.0103703733=~1% (which is the difference between your average selling price (14,844.4444) and the price of the first buy order that your sell order matched against (15,000)).

i m interested in this as i m trying to use historical data to calculate something indicative of "real invested value". do you know if there are appropriate models ?