Which cryptocurrency wins in the long run? That question frames most crypto debate, but it might be asking the wrong thing entirely.
Nonco analyst Caue Teixeri argues that infrastructure—the pipes, not the tokens—is what actually accrues value as the space matures. That's a shift in how people think about winners and losers. Instead of betting on Bitcoin or Ethereum or the next Layer 2, the real edge sits in the systems that let any of those assets settle, verify, bridge between chains, or get indexed for search.
The infrastructure thesis tracks. Exchanges, custody providers, and cross-chain bridges survive bear markets because they're not tokens with speculative demand—they're services with genuine use. A developer building on Ethereum doesn't stop needing block explorers or bridging tools when the price tanks. A fund holding digital assets doesn't stop needing settlement and verification layers. These layers generate fees or provide cost advantages that don't vanish when sentiment does.
That doesn't mean infrastructure tokens always outperform or that picking them is straightforward. It means the category is less purely speculative than individual coin bets, because the demand curve is tied to actual usage, not just narrative rotation.
When forced selling peaks before the bottom
Timing buy signals during forced selling is notoriously hard. Alen Pavlović at Liquibit Capital used CoinDesk's liquidation feed to track June's margin call cascade, finding liquidations peaked near $68,000 in bitcoin terms—several days before BTC actually bottomed. That gap is useful context: when liquidations spike, the market hasn't yet flushed out all the leverage, so the pain trade often extends beyond the peak-chaos moment.
The pattern also suggests that watching liquidation volume and price together can flag when forced selling is likely exhausted, even if it's not a bulletproof reversal signal. Pavlović's data points to the messier reality of crash bottoms: they don't announce themselves at peak drama. They arrive in the quiet after.