What if the next wave of market participants never sleeps, signs its own transactions, and arbitrages basis spreads in milliseconds? According to Coinbase’s institutional strategy lead John D’Agostino, that’s where we’re headed—and he argues those AI agents will need crypto rails to function at scale. If machines are going to act on our behalf, they’ll demand real-time settlement, verifiable data, and programmatic money—three things legacy finance struggles to deliver.

What’s happening

AI-driven agents are increasingly building, trading, and interacting across Web3. D’Agostino told CNBC that traditional rails weren’t built for machine-to-machine, high-frequency value transfer, while blockchains provide globally available, programmable, and near-instant “money rails.” He also pushed back on the constant Bitcoin vs. gold debate, noting BTC’s portability and programmability, and cautioned that institutional adoption will continue—but in steady, deliberate steps rather than a sudden “wave.”

Why it matters to traders

If autonomous agents scale, order flow will migrate toward venues that offer low latency, finality, and composability. That tilts activity toward chains and infrastructure with high throughput and robust developer tooling. Macro still matters: with the Fed cutting rates for the first time this year on Sept. 17 (and debate ongoing about further cuts), the “cash on the sidelines” in money markets could start rotating—D’Agostino argues a slice may flow into Bitcoin and crypto risk. Positioning ahead of structural flow shifts can be a decisive edge.

Opportunities on the radar

Traders can look beyond headline narratives to the plumbing agents will use. Watch sectors that benefit from machine-native activity: stablecoin settlement, oracles, execution engines, and L2 scaling. Bitcoin’s relative appeal also improves if real yields compress; keep an eye on the BTC/gold ratio as a proxy for risk rotation into programmable “digital collateral.”

Risks you must price in

Machine-native markets don’t erase risk—they remix it. Key threats include MEV and latency arbitrage affecting fills, oracle manipulation during thin liquidity windows, poor data quality for AI models, and regulatory uncertainty around autonomous agents. Macro risk remains: if inflation proves sticky and rate cuts stall, the “cash-to-risk” rotation slows.

Actionable game plan

• Track on-chain “agent heat”: rising stablecoin velocity, transaction throughput, and gas utilization on high-performance chains can front-run flow.
• Monitor the BTC/gold ratio; a sustained breakout above its 200D average can signal renewed preference for programmable collateral.
• Build a watchlist of “agent rails”: leading oracles, perp DEXs with deep liquidity, and scaling solutions. Trade strength on pullbacks rather than chasing green candles.
• Use execution discipline: limit orders around liquidity pools, avoid low-liquidity hours prone to MEV spikes, and split size to reduce footprint.
• Tie entries to macro: align risk adds with dovish shifts in Fed expectations and falling real yields; fade overextensions when rate-cut odds compress.

The institutional tempo: slow, steady, consequential

Don’t wait for a mythical “wave.” Pensions and endowments move methodically—but when they do, the footprint is durable. Expect incremental participation across custody, basis trades, and BTC allocation frameworks, not sudden all-in behavior. That favors strategies that compound via consistent positioning rather than one-off event bets.

Bottom line

If AI agents become meaningful market actors, liquidity will reward chains and venues that clear value at machine speed with verifiable truth. Align your scans with that future, position for macro tailwinds, and respect the new microstructure risks that come with it.

If you don't want to miss any crypto news, follow my account on X.