Lesson 2 — Token burns: what they actually do

**The mechanism — what a burn actually does.** A burn moves tokens from a wallet to an address whose private key is provably impossible to derive (typically `0x000...000` or similar 'dead' addresses). The tokens still exist on-chain; they just can't be spent. The effect on circulating supply depends entirely on the size of the burn relative to total supply and the rate at which new tokens are created.

**Three categories of burn that you'll see announced.** (1) **Protocol-level fee burns.** Ethereum's EIP-1559 burns a portion of every transaction's base fee. BNB's quarterly burns are tied to BNB Chain transaction volume. These are real supply removal mechanisms — they happen automatically, the size is determined by activity, and they meaningfully reduce supply over time. (2) **Buyback-and-burn programmes.** A project uses treasury or revenue to buy tokens from the market and burn them. The size is discretionary and the effect depends on whether the buy is large relative to daily volume. Marketing-burn variants buy modestly and announce loudly. (3) **One-time supply burns of pre-existing project allocations.** A team burns part of its own holdings to claim 'committed to long-term value.' These are often the biggest individual burn announcements but the smallest real-world impact — burning tokens you control is functionally equivalent to never having released them, and the announcement is mostly signalling, not actual supply reduction.

**The right diagnostic questions.** When you see a burn announcement, ask: What percentage of total supply is being burned? (Anything under 1% is rarely market-moving even if announced loudly.) Who held the burned tokens before the burn? (Tokens burned from a team wallet are functionally equivalent to never-released tokens; tokens burned from circulating supply are real reduction.) Is this a one-off or part of a programmatic mechanism? (Programmatic burns compound; one-offs don't.) Does the project's tokenomics include corresponding emission? (A 5% burn paired with 8% annual emission is net-inflationary.)

**The 'percentage of supply burned' math that misleads.** Marketing material often headlines a percentage that sounds enormous. 'Project X has burned 50% of its supply!' sounds dramatic. But many projects launched with enormous initial supplies precisely to allow large-percentage burn announcements later — burning half of 1 quadrillion tokens still leaves 500 trillion tokens, and the price impact of the burn is determined by what fraction of *liquid circulating supply* (not total mint) was actually removed.

**The deflationary narrative tested against returns.** Ethereum's EIP-1559 produced real, sustained deflationary pressure during high-activity periods (more ETH burned than issued for stretches in 2022–2023). ETH's price didn't simply track deflation; it tracked broader macro and crypto-specific cycles. Burn programmes can produce structural supply changes; they don't reliably produce price-action because price depends on the full demand side, which burns don't control. Treat 'this project burns tokens, so it must go up' as the categorical fallacy it is.

**The honest test.** Real burn programmes are codified in protocol-level mechanics that don't need to be announced because they happen automatically (you can see them on-chain). Burn programmes that require a marketing announcement to be noticed are usually doing the work of marketing rather than supply mechanics.