Valuing crypto
The honest answer is that the standard methods do not apply, and the substitutes are weaker than their users admit. That is worth saying plainly.
Every conventional valuation method estimates future cash flows and discounts them. Shares have earnings, bonds have coupons, property has rent. Most crypto assets have none, so the machinery has nothing to grip.
What that leaves
For an asset with no cash flow, the price is whatever the marginal buyer will pay. That is not a criticism unique to this field — it is equally true of gold, art and collectibles, all of which have held value for a long time. It does mean the tools are different, and weaker.
The approaches in use, and their problems
- Stock-to-flow and scarcity models. Relate price to the ratio of existing supply to new issuance. Fitted to a short history, with no mechanism explaining why the relationship should hold, and they have diverged badly from outcomes since being published. Treat as curve-fitting.
- Network value ratios. Compare market value to on-chain transaction volume, by analogy with a price-to-sales ratio. The denominator is easily inflated by transfers between an owner's own addresses, which is a serious measurement problem rather than a detail.
- Metcalfe-type models. Value proportional to the square of active users. Plausible for a network, and "active address" is not "user", since addresses are free.
- Total addressable market. If it captures x% of gold, or of remittances, it is worth y. The arithmetic is trivial; the assumption is the entire argument and is usually asserted.
- Fee-based valuation. For protocols that earn fees, discount those fees like a business. The most defensible method available, and it applies only to the minority of assets that actually capture revenue. See tokenomics.
Each captures something. None produces an estimate with the reliability that a discounted cash flow on a stable business produces, and none has a track record long enough to evaluate. The honest position is that valuation here is closer to judgement about adoption than to calculation — and the field's habit of dressing that judgement in a model with a chart is what deserves scepticism, not the judgement itself.
The one framework that transfers
The monetary-asset case does not require a cash flow. If an asset is used as a store of value, its price reflects how much wealth people choose to hold in it — and that is a question about adoption and confidence, not earnings. Gold is valued this way and has been for millennia.
It follows that the argument about bitcoin is really an argument about whether it becomes such an asset, which no model resolves. Anyone with a price target derived from a formula has hidden that assumption inside it.
What to do instead
- Separate the categories. A fee-earning protocol can be valued approximately. A monetary asset can only be assessed by adoption. A token with no mechanism is priced by sentiment. Applying one framework to all three is the common error.
- Check the supply side. Whatever you believe about demand, scheduled issuance and unlocks are knowable. See tokenomics.
- Size for being wrong. When valuation cannot give you a floor, position size is the only risk control you have. See position sizing.