Everything in this series transfers to digital assets - and almost everything needs adjustment. Crypto market structure looks like equities rewound to a wilder, more fragmented era: dozens of venues, no consolidated tape, no central clearing, and a market that never closes. This final part maps the differences and what they mean for each layer of the execution stack.

What’s Structurally Different

No Sessions, No Auctions

Crypto trades 24/7. There is no opening print, no closing auction, and therefore none of the liquidity anchor points that equity execution schedules around (Part 2). Volume still has seasonality - it follows the sun through Asia, Europe, and US hours, clusters around US macro releases, and spikes at perpetual-futures funding timestamps and monthly expiries - but the pattern is flatter and less reliable than an equity U-curve. VWAP profiles (Part 5) must be built on these rhythms, and “end of day” becomes a choice the desk makes, not a fact of the market.

Fragmentation Without a Tape

Liquidity for a major pair is split across many centralized exchanges - plus decentralized venues - and no regulator consolidates the quotes. There is no NBBO; each venue is its own price. Cross-venue price discipline comes only from arbitrageurs, and in fast markets venues visibly desynchronize. For execution this has two consequences: building your own consolidated book is prerequisite infrastructure, and every benchmark must name its venue (or define an explicit composite) or the TCA numbers of Part 12 are meaningless.

Prefunding and Settlement

The deepest routing difference: without central clearing, you can generally only trade on a venue where you already hold assets. A crypto smart router (Part 9) therefore optimizes over funded venues, and inventory management - how much of which asset sits where, and when to rebalance across venues given transfer times and withdrawal fees - becomes part of execution itself. Counterparty risk joins the routing objective too: the best price on a venue you wouldn’t trust overnight is not the best price.

Fee Structures That Actually Bind

Tiered maker/taker schedules, volume-based discounts, and occasional maker rebates create fee differences across venues and accounts that frequently exceed the visible price differences. Router logic must compare fee-adjusted prices, and passive-versus-aggressive decisions (Part 11) shift meaningfully with the account’s tier.

Data You Can’t Fully Trust

Reported volume on loosely surveilled venues can be inflated by wash trading. Any component that consumes volume - POV denominators (Part 6), VWAP profiles, impact model calibration (Part 13) - needs a venue whitelist and plausibility filters. Displayed order-book depth, for all its spoofing caveats, is often a more honest liquidity measure than printed volume.

Strategy by Strategy

• TWAP (Part 4) earns a bigger role than in equities: for long-tail alts and new listings there is no volume profile worth trusting, and time-slicing with heavy randomization is the defensible default. Thin books make display caps and size jitter non-negotiable.

• VWAP works for majors on high-quality venues, built on sun-cycle and funding-clock seasonality rather than an open/close U-curve.

• POV demands filtered volume: participate against trusted-venue prints only, or a wash-inflated tape will pull the algorithm into aggression precisely on the noisiest venues.

• Opportunistic styles (Part 7) thrive: crypto’s frequent dislocations - liquidation cascades, cross-venue desyncs - are exactly the “unusually favorable moments” the strike logic was built for. Impact in thin books skews temporary and reverts hard (Part 13), which rewards the patient, liquidity-taking-on-dips posture.

• Derivatives-aware execution: perpetual futures are often deeper than spot, so large programs frequently execute the perp first for immediacy and migrate into spot over time - a leader/follower structure straight out of Part 8, with the basis and funding rate as the spread being managed.

A Crypto Execution Checklist

1. Build (or buy) a consolidated, fee-adjusted view of every venue you can trade.

2. Whitelist venues for data as well as for execution - they are different lists.

3. Treat inventory placement and rebalancing as part of the execution problem.

4. Randomize aggressively; thin books make patterns cheap to spot.

5. Define benchmarks explicitly - venue, composite method, timestamp source - before trading, not after.

6. Size orders against depth-based liquidity measures, not reported volume.

That closes the series. The through-line from Part 1 to here: execution is a single optimization - cost against risk against alpha (Part 3) - implemented as a stack of schedulers, routers, and placement tactics, and kept honest by measurement (Part 12). Master the framework and the market-specific details, crypto included, become configuration rather than mystery.