onboarding

Algo Orders — your first TWAP without overthinking it

Worked example: buy $1,000 of ETH over 60 minutes via TWAP — when to use it, when not to

6 min · beginner

What you'll have when finished

  • Understand which algo type fits which problem (TWAP vs VWAP vs Iceberg vs Bracket)
  • Place your first TWAP order with a sensible slice count
  • Read the order progress + estimated savings
  • Know when an algo order isn't worth using (smaller trades)

Before you start

  • For trades <$5-10K notional, the slicing overhead isn't worth it. Use Smart Trade or CEX Trading.
  • TWAP/VWAP slices fire as MARKET orders by default. In flash crashes, you'll get filled at bad prices on remaining slices.
  • Cancelling an in-progress TWAP stops pending slices — but slices already filled stay filled.

Walkthrough

  1. Decide if you actually need an algo order

    Before opening the form: **most trades don't need an algo order.** Use TWAP/VWAP when: - **Order size >$5,000 USD equivalent** — below this, slicing overhead exceeds savings - **You don't want to move the market** — entering a position quietly - **Pair has thin order books** — small relative volume = your market order would walk the book Use Iceberg when: - **Order size is huge ($50K+)** and you want to hide your true position from front-runners - **Order book is shallow** — showing full size would scare other participants away Use Bracket when: - **You want entry + TP + SL as ONE linked order set** without manually managing each - **Discipline matters** — pre-committing to TP/SL removes the "I'll move it just this once" temptation Use Smart Trade INSTEAD when: - Trade is <$5K - You want multi-level TPs (Bracket only supports one TP) For our worked example: **buy $1,000 of ETH on Coinbase over 60 minutes via TWAP**. This is the SMALLEST trade where TWAP makes sense — perfect for learning the mechanics without much slicing overhead.

    Success criteria: Identified your order is >$5K (or you're practicing on smaller) · Picked the right algo type

  2. Open the form + pick exchange, pair, side

    Navigate to **CEX Trade → Algo Orders**. Top of the form: 1. **Exchange selector** — pick Coinbase (or wherever you have ETH/USDC available) 2. **Pair** — type ETH in the search; click ETH/USDC 3. **Side** — click Buy (green) The TradingView chart updates to show ETH/USDC on Coinbase. Watch the live price; you'll see it tick during the order.

    Success criteria: Exchange + pair + side selected

  3. Pick TWAP from the algo type cards

    Four colored cards: TWAP / VWAP / Iceberg / Bracket. Click **TWAP** (blue). Why TWAP for this trade: - We're buying $1,000 — small enough that VWAP's volume-weighting doesn't meaningfully help - We're not hiding from anyone (small size, public market) - We don't need TP/SL bundled — we just want to accumulate at a steady rate VWAP would do the same thing but with non-uniform timing (more during peak hours). For 60-minute windows, this difference is small. Iceberg is for >$50K orders where order-book visibility matters. Bracket is for in-and-out trades, not pure accumulation.

    Success criteria: TWAP algo type selected

  4. Switch to QUOTE mode and enter the total

    Two amount-mode buttons: **BASE (ETH)** and **QUOTE (USDC)**. Click **QUOTE (USDC)**. Now you can type dollars directly. **Total amount:** type **1000** The form converts to BASE under the hood (using the live ticker). For $1,000 at ETH = $2,000, that's 0.5 ETH. **Why QUOTE mode:** "I want to spend $1,000" is harder to misread than "I want to buy 0.5 ETH." For algo orders especially, where the same number gets divided across many slices, dollar-thinking is the right default.

    Success criteria: Amount mode = QUOTE · Total = 1000

  5. Set Duration + Slice Count — the two params that matter most

    **Duration (minutes):** how long the total execution takes. For our example, type **60** (1 hour). **Slice count:** how many child orders to break the total into. Default is **10**. Math: $1,000 ÷ 10 slices = $100 per slice. Over 60 minutes, that's one slice every ~6 minutes (with ±5% timing jitter automatically added). **Tuning guidance:** - **More slices** = smaller per-slice market impact + longer effective duration. But more orders = slightly more fees on per-fill venues. - **Fewer slices** = faster execution + bigger per-slice impact. - **Duration matters for trend risk** — over 60 min, ETH might move 1-2%. Over 4 hours, it might move 5%+. The longer you spread, the more directional drift you absorb. **Rule of thumb:** each slice should be **<0.5% of the pair's 24h volume**. For ETH/USDC on Coinbase with billions in daily volume, even $1,000 slices are tiny. For obscure pairs, calculate first. For our example: 60 min / 10 slices = sensible starter values.

    Success criteria: Duration = 60 · Slice count = 10

  6. Submit + monitor in the orders list

    Click **Create TWAP Order** (cyan button at bottom). The order appears in the orders list (right side or below depending on screen size) with: - **Status:** active - **Progress bar:** 0/10 slices filled, 0% complete - **Estimated savings:** vs market-order benchmark (usually 10-40 bps for $50K+ orders, less for $1K orders) - **Next slice in:** countdown to the next slice (varies ±5% jitter) Over the next 60 minutes, you'll see: - Slice 1 fills around minute 6 - Slice 2 fills around minute 12 - ... - Slice 10 fills around minute 60 Each slice executes as a market order at the prevailing ETH price. The total fill price ends up close to the 60-minute average ETH price — that's the whole point of TWAP. **Polling:** the list auto-refreshes every 15 seconds. No need to manually refresh.

    Success criteria: TWAP submitted · First slice filled within ~10 min · Progress bar advancing

  7. What you should see vs. what means something went wrong

    **Normal:** - All 10 slices fill on schedule (with ±5% timing variation) - Average fill price ≈ 60-minute average ETH price (within fees) - Progress hits 100% at ~60 minutes - Status changes to "done" **Concerning:** - Slices keep failing — usually means insufficient USDC balance or hit rate limits. Cancel the order, top up, retry. - Order paused mid-execution — exchange-side issue. Check exchange status; resume if it auto-recovers. - Massive negative estimated savings — your slices are getting bad fills, probably because the pair is too illiquid for this slice size. Reduce slice count next time (each slice will be bigger but maybe still fit the book). **Cancelling mid-flight:** - Click Cancel on the order - Pending slices stop immediately - Slices already filled stay filled (no rollback) - The remaining USDC stays in your account for use elsewhere

    Success criteria: Read the normal vs concerning patterns · Know how to cancel mid-flight

What's next

**TWAP** is your starter algo. After 5-10 TWAP runs, consider: **VWAP** when you want better matching to the market's natural rhythm — useful when you're tracking a benchmark or trying to look like normal market participation. **Iceberg** when your order is $50K+ and you don't want the order book to know your true size. Especially valuable on smaller-cap pairs where one big order moves the price. **Bracket** when you want entry + TP + SL bundled as ONE linked order set — useful for one-shot trades where you're NOT going to micromanage. For multi-TP ladders, use [Smart Trade](/smart-trade) instead. For the next level — pre-trade analysis, fee impact, and venue selection across your algo runs — see the existing intermediate guide [large-order-execution-strategy].