Smart order routing, algorithms, and analytics – built for institutional crypto
A complete execution layer covering 6 SOR strategies, 8 algorithmic strategies, a proprietary market impact model, and a full transaction cost analysis suite – with safety and risk controls embedded throughout.
One execution layer – from routing to analytics
Talos provides a complete institutional execution suite for digital assets: smart order routing that evaluates liquidity across venues in real time, algorithmic strategies tuned to crypto market microstructure, pre-trade cost modeling before any order is sent, and post-trade TCA that closes the feedback loop. Safety and risk controls are embedded at every layer.
SOR continuously evaluates all connected markets for best average fill price, accounting for maker/taker fees. Orders are re-routed as conditions change. Resting orders move to marketable venues automatically. Designed for orders where immediate or best-price execution is the primary objective.
Eight algorithmic strategies covering passive pegging, interval-based taker execution, volume-weighted and time-weighted schedules, participation-rate control, and multi-level price-range distribution. All passive strategies use anti-gaming protection and blended child order sizing to minimize information leakage.
The Talos Market Impact model provides projected slippage estimates before execution. Post-trade TCA surfaces actual slippage vs. benchmarks, fill rates, dealer latency, and execution markouts at the parent order, venue, and symbol level. Accessible via the Talos UI or REST and WebSocket APIs.
Know your expected execution cost before the order is submitted
Talos Pre-Trade Analytics is built on the Talos Market Impact (TMI) model – a three-component cost decomposition that turns intraday market predictions into quantified slippage estimates, scenario comparisons, and parameter-tuning guidance before a single child order is sent.
The screenshots below show an active pre-trade session for a 100 BTC market VWAP buy order. Four scenarios were generated at different horizons (2h, 1h, 30m, 15m) and compared side-by-side. The 1h scenario is shown in detail. All data reflects live Talos platform outputs against consolidated market conditions at time of capture.
Every time an order crosses the bid-ask spread, a cost is incurred. The spread cost component is a calibrated coefficient multiplied by the predicted bid-ask spread. It is size-independent – it captures the probability of crossing the spread per child order, inclusive of fees. Passive fills (maker) pay zero spread cost; aggressive fills (taker) pay the full spread.
- Driven by the Talos spread prediction alpha for the symbol and markets selected – updated daily.
- The primary argument for passive execution: every resting fill reduces this component to zero for that child.
Physical impact follows a square-root law adjusted with a sigmoid function – where participation rate (π = order size / market volume over the horizon) and intraday volatility jointly determine the impact magnitude. Larger orders relative to market volume create more impact, but the relationship is sub-linear. The sigmoid handles edge cases at extreme participation rates.
- The key input is π – the expected participation rate. Reducing execution speed (longer duration, smaller clips) lowers π and reduces this component.
- Driven by the Talos volume and volatility prediction alphas.
Every dollar of the order scheduled to trade later rather than immediately carries the risk of price moving away from the arrival price. In 24/7 crypto markets with no overnight adjustment period, time risk accumulates continuously. This component is driven by predicted volatility scaled over the execution horizon T – longer durations in volatile conditions are penalized proportionally.
- Unique to crypto: no overnight risk reset. Time risk accumulates 24/7 unlike traditional markets.
- If you have a directional alpha view (expected mean reversion), you can discount this component in strategy selection.
REST endpoint
Synchronous REST interface accepts a hypothetical order and returns time-sliced participation bins, upper/lower execution bounds, and market-level cost metrics: SpreadCost, TemporaryCost, TimeRisk, IntervalVolume, and DailyVolatilityBps per market. Supports historical pre-trade queries for backtesting.
→ kb.talostrading.com/api-docs/principal/analytics/pre-trade-analytics/rest
WebSocket stream
Real-time pre-trade request and response via WebSocket subscription. PreTradeRequest (client → server) submits a hypothetical configuration. PreTradeResponse (server → client) returns the projected trajectory. Link pre-trade sessions to live orders via SessionID / PreTradeSessionID for end-to-end attribution.
→ kb.talostrading.com/api-docs/principal/analytics/pre-trade-analytics/websocket
Six routing strategies that adapt to available liquidity in real time
Talos SOR continuously evaluates liquidity across all connected exchanges and venues to find the optimal child order allocation for best average execution price – factoring in maker/taker fees on each market. Routing is re-evaluated at configurable intervals as conditions change: resting orders are moved to newly marketable venues automatically, without waiting for a fill trigger.
Full Amount waits until a single market can fill the entire order at the limit price before sending any child orders. Until that condition is met, the order rests on the Talos SOR. This prevents partial fills and ensures that size is not fragmented across venues when a complete fill at the target price is achievable from one source.
- Sends an aggressive single-venue order the moment full crossing liquidity is confirmed – minimizing leakage from partial reveals.
- Suitable for orders where all-or-nothing execution at a specific price is more important than time-to-fill.
Iceberg executes large parent orders using smaller visible child quantities, either at a user-specified Show Qty or randomized market-based sizes to blend into the book. For aggressively priced orders it first sweeps available liquidity up to the limit price, then begins posting smaller quantities at the limit price as fills occur.
- Show Qty (optional): the total size made visible across all markets, split evenly across GTC-capable venues. If unspecified, randomized sizes are posted to minimize detectability.
- Child orders replenish continuously as fills arrive – size is maintained at the displayed level throughout the order's life.
For marketable orders, Limit sweeps available liquidity up to the limit price by routing to markets offering the best prices first. The unfilled residual is then rested: split equally across markets that support GTC orders, or held on the Talos SOR for markets that do not. Re-evaluation continuously moves resting size to newly marketable venues.
- Price-ranked routing ensures the most favorable venue fills first – best-execution routing on the sweep leg.
- Residual rest logic is configurable per market – contact Client Services to adjust resting behavior.
Market simulates a market order by sending limit orders at the best available prices across all connected markets simultaneously. If the initial orders result in partial fills, subsequent orders are sent immediately until the full quantity is filled. This provides market-order speed with limit-order price control and SOR venue selection.
- Limit orders are set at the best expected fill price per venue – not a single global limit price.
- Order price protection applies: orders more than 100 bps / $5,000 from the median top-of-book far side are rejected automatically.
Post-Only allocates size to all available markets where the order's price does not cross the top of book – ensuring every fill is passive. Where exchanges support native PostOnly orders, those are used. On exchanges without native PostOnly support, limit orders are placed instead. Size is split evenly across all qualifying markets.
- No aggressive orders are ever sent – the strategy waits for other participants to take the resting quantity.
- Order size is spread evenly across qualifying markets when multiple venues are selected, maximizing passive fill probability.
Stop Limit and Take Profit Limit are conditional order types that place a Limit order in the market once a specified trigger price is reached. Stop Limit activates when price moves against the position (protective stops); Take Profit activates when price moves in favor (profit capture). Both trigger as standard Limit orders and can be combined with other algo strategies.
- Stop Limit: Buy trigger ≥ current price; Sell trigger ≤ current price.
- Take Profit Limit: Buy trigger ≤ current price; Sell trigger ≥ current price.
- Trigger Source (optional): trigger based on selected markets (default) or a reference aggregation mid price.
- Can be combined with Steady Pace and other algos – contact Talos to configure conditional algo activation.
Continuous re-evaluation
Routing is re-evaluated at intervals set by a configurable re-evaluation timer. If resting size on one venue becomes marketable on another, the order is moved automatically. Re-distribution is only triggered when a resting order fills beyond a set threshold – avoiding unnecessary order churn.
Fee-inclusive allocation
Child order allocation accounts for maker and taker fees on each exchange when computing the optimal split. Fee-Aware Allocation (configurable) ensures that post-fee balances match the intended quantity – particularly relevant for exchanges that charge fees in the base asset or where the notional target is in quote currency.
Resting and venue support
Child order resting behavior depends on venue support: markets that support GTC orders receive resting child orders directly. Markets that do not support resting orders hold residual size on the Talos SOR until conditions are met. Max Resting Threshold (in bps from top of book) controls how far off-market child orders are allowed to rest before they are withdrawn.
Eight strategies – from passive pegging to volume-prediction scheduling
Every Talos algo breaks large parent orders into smaller child orders, routes them across multiple venues, and adapts to changing market conditions. Passive strategies use anti-gaming logic and blended order-book sizing to minimize information leakage. Schedule-based strategies reschedule automatically on pause or resumption to avoid aggressive catch-up orders.
| Strategy | Primary objective | Placement | Duration based | Volume pred. | Passive-first |
|---|---|---|---|---|---|
| Pegged | Earn spread via passive resting across multiple book levels | Passive multi-level | — | — | ✓ |
| Steady Pace | Minimize market impact with consistent interval-based clips | Interval taker | ✓ | — | — |
| Time Sliced | Minimize impact via scheduled clip execution; will catch up | Schedule taker | ✓ | — | — |
| VWAP | Minimize impact by front-loading execution into high-volume windows | Passive + active | ✓ | ✓ | ✓ |
| TWAP | Minimize impact and spread cost on a linear time schedule | Passive + active | ✓ | — | ✓ |
| Quantizer | Index replication – guaranteed quantity per fixed interval | Passive + active | ✓ | — | ✓ |
| PoV | Trade a target percentage of real-time market volume passively | Volume prediction | — | ✓ | ✓ |
| Scaled | Passive execution across a price range for better average pricing | Price band passive | — | — | ✓ |
A passive algorithm that tracks the bid (for buys) or offer (for sells) across multiple markets and multiple price levels simultaneously. Pegged improves the probability of fills by posting at multiple venues and book levels, with sizing controlled by the Urgency parameter. Uses anti-gaming logic to protect posted orders from being gamed by other participants.
- Multiple price levels posted simultaneously – higher urgency means more size posted closer to the top of book.
- Anti-gaming features randomize posted schedule and sizes; a proprietary liquidity filter governs order placement decisions.
- Peg Reference Type (API only): Near (same-side best bid/offer) or Mid (mid-price pegging) – Near is default.
| Parameter | Description | Default |
|---|---|---|
| Urgency (1–5) | Controls aggression level. Each level targets a percentage of L2 book depth: 1 ≈ 4.5%, 2 ≈ 10%, 3 ≈ 17.5%, 4 ≈ 35%, 5 ≈ 55% of size per level. | 3 |
| Price Offset | Offset from the reference price. Positive values are more passive (away from top of book). Specified in price units via UI; bps or tick increments available via API. | 0 |
| Peg Reference | Near: track best bid/offer. Mid: track mid-price. API only. | Near |
Steady Pace breaks an order into user-defined clip sizes and executes one clip per interval by routing aggressively to the best available market prices. Each clip is guaranteed to fill for its full size (with error retries within the interval). Critically, if the order becomes non-marketable for a period, Steady Pace does not increase aggressiveness to catch up – it simply misses those intervals, resulting in a partial fill for the overall order.
- Clip Size: quantity per interval – defines the execution pace regardless of market volume.
- Clip Interval: time between each clip send – controls frequency and therefore total execution speed.
- Variance (optional): randomize clip sizes by a specified percentage (e.g., 2% variance) – reduces predictability.
- Sweep Dealer Markets: when enabled, each clip can be filled across dealers/OTC desks in addition to exchanges.
Time Sliced is similar to Steady Pace but calculates its clip interval from the clip size and end time automatically. Unlike Steady Pace, if the order becomes non-marketable, Time Sliced will increase aggressiveness to catch up toward the end time – ensuring the full order quantity is executed by the specified deadline where possible.
- Clip Size: the fixed quantity per execution interval.
- End Time: the order expiry – interval timing is derived automatically from clip size and remaining duration.
- Variance (optional): optional clip size randomization – same as Steady Pace.
- Error retries within each interval ensure the full clip size is filled before moving to the next interval.
VWAP executes child orders in proportion to the Talos Intraday Volume Prediction – placing more size during periods of expected higher volume and less during predicted quiet windows. This reduces per-unit spread cost by trading when natural liquidity is available, rather than distributing uniformly across all time buckets. Volume predictions are re-calibrated daily from the Talos Execution Alphas library.
- Multi-level passive pegging with bounds control – crossing only when filled quantity falls below the lower bound.
- Reschedules execution path automatically if the order is paused or non-marketable, preventing aggressive catch-up orders on resumption.
- Participation Rate Cap (optional): maximum rate of market volume the order may consume – order may not complete when capped. Valid: 0.1–100%.
- Trade Predictions panel in the UI shows predicted spread, volume, volatility, and participation rate for the execution window vs. daily baseline.
TWAP distributes execution evenly across the specified duration regardless of the intraday volume profile. It uses the same passive-first multi-level pegging framework as VWAP with identical bounds control. TWAP is the right choice when the benchmark is the simple time average and volume-weighting would create undesirable concentration – or when intraday volume patterns are unpredictable for the specific pair.
- Passive multi-level pegging with bounds-controlled path – crosses spread only when below the lower bound threshold.
- Reschedules automatically on pause or resumption to prevent large aggressive orders at restart.
- Optional Clip Size and Clip Interval provide granular control over child order frequency within each time bucket (similar to Steady Pace behavior within the TWAP schedule).
- Bounds Control: Narrow ±min(3%, 3 min), Standard ±min(5%, 5 min), Wide ±min(7.5%, 10 min).
Quantizer discretizes execution into independent fixed-size, fixed-time intervals. Unlike TWAP – which allows tracking freedom across the full horizon – Quantizer guarantees that a specified quantity is fully executed by each interval boundary. Within each interval, TWAP-style passive placement minimizes spread crossing while completing the interval quantity.
- Interval Size: the quantity to execute within each interval.
- Interval Time: the duration of each interval.
- End Time (optional): if set, any residual below 20% of an interval size is bundled into the final interval rather than left untraded.
- Key distinction vs. TWAP: Quantizer completes within each interval. TWAP allows catch-up across the full horizon.
PoV aims to trade a user-specified percentage of actual market volume by continuously managing passive and aggressive child orders to stay within the participation rate bounds. Volume prediction allows passive orders to be placed in anticipation of incoming volume rather than purely reacting to it.
- Target Rate (1–50%): the percentage of accessible market volume to target. 5–15% is recommended for most institutional order sizes.
- Bounds Control: Narrow tracks strictly (higher spread-crossing frequency), Standard balances deviation and cost, Wide allows larger passive deviations to maximize passive fill rate.
- Sizes passive child orders based on blended historical and real-time order book moving averages to blend into the book and minimize information leakage.
- Volume prediction places resting orders ahead of where volume is expected to occur – a key differentiator from traditional reactive PoV implementations.
Scaled distributes a parent limit order across multiple post-only child orders at different price levels within a defined range. All child orders are post-only – meaning they will be rejected if any would cross the spread. This provides passive execution across a price band, achieving a potentially better average price than a single limit order at the top of book.
- Size distribution is configurable: a ratio of 2 places twice as much size at the Scale End Price (furthest from market) vs. the Limit Price; a ratio of 0.5 places twice as much near the market. Default is uniform (ratio 1).
- Compatible with Unified Liquidity and Reduce-First logic for position management workflows.
- Up to 20 price levels by default – contact Client Services to increase.
| Parameter | Description | Required |
|---|---|---|
| Scale End Price | The furthest price from market where child orders are placed. Defines the far end of the range. | Yes |
| Num Price Levels | Number of unique price levels within the range (max 20 by default). | Yes |
| Size Distribution | Ratio controlling how size is distributed across levels. 1 = uniform; 2 = more at far end; 0.5 = more at near end. | No (def. 1) |
Note: the Limit Price on the parent order defines the most aggressive (closest-to-market) price level. Child orders are placed from this price outward toward Scale End Price.
Anti-Gaming Protection. Resting agents randomize the schedule and sizes posted to exchanges, preventing other participants from detecting and gaming the order's predictable pattern. A Talos proprietary model filters the liquidity data used to determine order placement – further reducing information leakage.
Deeper Resting Levels. Resting agents can be configured to post orders deeper in the order book to capture more of the spread and minimize toxic fills and adverse price selection. This improves algo slippage vs. the interval VWAP benchmark while maintaining a good overall fill rate.
Multiple Resting Levels. The number of simultaneous resting orders posted to a single market is configurable per strategy type (default: 3 levels). Increasing this improves passive fill probability at the cost of greater order book visibility.
Order Price Protection. Markets differing more than 10% from the Talos Reference Rate are excluded automatically. Aggressive child orders are limited by the Fair Price Model: Aggressive Fair Price = Reference Rate ± (10 × median spread); Passive Fair Price = Reference Rate ± (1 × median spread). SOR orders with potential losses exceeding 100 bps or $5,000 are rejected.
Overfill Protection. If an LP disconnects mid-execution or a child order remains pending beyond timeout, the OEMS waits for resolution before placing new exposure. Child order exposure is validated against the parent order total at all times.
Order State Recovery. On network error or disconnect, gateways reconnect, query all fill statuses, and update parent/child orders. Unmatched fills are still booked for reconciliation and out-trade detection.
Runaway Algo Detection. Order and fill rates are monitored continuously. Exceeding configured thresholds blocks new orders until the rate window clears. Duplicate detection on (Symbol+Side+Price+Qty) prevents runaway algo scenarios.
Measure execution quality from every angle – and use it to improve
Talos Post-Trade Analytics provides transaction cost analysis (TCA) at the strategy, market, and symbol level. Slippage versus multiple benchmarks, fill rates, dealer round-trip latency, and execution markouts are all tracked from your first trade and accessible via dashboard or WebSocket API.
View trading volumes by strategy and slippage at the parent order level versus multiple benchmarks. Focus on the benchmark relevant to each strategy: arrival price for opportunistic, VWAP for volume-sensitive mandates, TWAP for time-scheduled mandates. Drill down to individual parent orders via per-order analytics.
Slippage at the child order level per market, fill rate percentages by order status (filled, rejected, cancelled), and dealer RTT (round-trip time) statistics by LP. BidAskSpreads by size bucket identify structural spread differences across venues for the same underlying pair.
Volume by day decomposed at symbol and market level. Drill-down into quote and spread dynamics at the symbol and market level – click on any chart data point to open the detailed view. Identifies structural differences in liquidity provision across venues for the same underlying pair.
The Execution Markout view measures where the market price moved relative to each child execution price – in 5-second intervals from 5 seconds before to 60 seconds after each fill. All child executions are aggregated by execution type (Active, Passive, Dealer) and market, enabling precise toxicity and adverse selection analysis.
Fills where the algo posted a resting order that was taken by another participant. Passive fills typically show positive markout (price moved favorably) because they do not signal direction to other participants. Target: positive or near-zero spread at all post-fill intervals.
Fills interacting with a dealer (OTC) rather than a public exchange. Inherently lower toxicity than public aggressive fills because they do not appear in the public order book and therefore leak less directional information. Still monitor post-fill markouts – dealer liquidity can still be toxic in some conditions.
Fills where the algo crossed the spread to take liquidity aggressively. Other participants can infer trading direction from aggressive orders, causing adverse price movement post-fill. A persistently negative markout for active fills indicates meaningful information leakage – consider widening bounds or increasing Urgency-down settings.
Sign convention: positive spread = execution price was better than subsequent market price (favorable). Negative = execution was worse. For a buy, negative means price fell after purchase; for a sell, negative means price rose after sale.
| Metric | Definition and use | Level |
|---|---|---|
| Slippage (bps) | Execution price vs. benchmark price (arrival, VWAP, TWAP) in bps. The primary cost metric for each parent order – the number you report against your mandate benchmark. | Parent order |
| Participation rate | Actual percentage of market volume the order traded during its execution window – compare against configured target to assess schedule adherence. | Parent order |
| Maker percentage | Proportion of fills classified as passive (maker). Higher maker % = lower spread cost + lower toxicity. A direct measure of the effectiveness of passive-first execution. | Parent order |
| Order duration | Elapsed time from activation to completion – compare against scheduled duration to identify delays, early fills, or pauses affecting benchmark alignment. | Parent order |
| TWAP slippage | Slippage vs. time-weighted average price for the execution window – relevant for time-schedule benchmark mandates. | Parent order |
| VWAP slippage | Slippage vs. volume-weighted average price for the execution window – relevant for volume-schedule benchmark mandates. | Parent order |
| Dealer RTT | Milliseconds from child order dispatch to dealer response (fill, reject, cancel). A direct measure of LP response quality – outliers indicate dealers costing fills through latency. | Child order |
| Execution markout | Market price spread vs. child execution price at 5-second intervals from −5 to +60 seconds post-fill, by execution type. The primary toxicity and adverse selection measurement. | Child order |
| BidAsk spread | Average bid-ask spread received in bps by size bucket and market. Identifies structural spread differences across venues for the same pair. | Venue |
Data-driven evidence behind every execution decision
Talos's Quant Execution Services team publishes empirical research grounded in production execution data. The findings below are drawn from the Talos Quant Execution Insights Report 2026, which analyzed over 250,000 parent orders across 618 assets throughout 2025 – covering strategy performance, maker-taker dynamics, dealer liquidity, and adverse regime behavior.
| Research area | Key finding | Headline number |
|---|---|---|
| Strategy benchmarks | All Talos execution algos are net positive versus direct sweep. VWAP delivers the strongest sweep savings among trajectory algos, while Pegged outperforms Iceberg significantly on arrival slippage. | 3–22 bps average savings by asset cohort |
| VWAP vs. TWAP | VWAP delivers ~6 bps better arrival slippage than TWAP and ~37 bps more sweep savings. Benchmark tracking: TWAP slippage capped at ~2.6 bps across all size buckets. VWAP slippage vs. market VWAP stays below ~6 bps – a robust outcome given that prediction vs. realization differences are expected. | 37 bps VWAP vs. TWAP sweep savings delta |
| Maker participation | The 50–75% maker participation range delivers the lowest arrival slippage across order sizes. Below 50%, physical impact dominates; above 75%, time risk rises. Mechanically maximizing maker share is not optimal – duration and participation must be calibrated jointly. | ~2 bps arrival slippage at optimal 50–75% maker range |
| Dealer liquidity | Dealer + exchange routing consistently beats exchange-only (6–14 bps of savings on average). For orders above $100k, the size-related slippage ramp is much flatter with dealer access – driven by SOR-based price discovery across a broader liquidity pool. | <2 bps size penalty with dealer routing vs. 3× escalation exchange-only |
| Adverse regimes | Up to 300 bps adverse market move: Talos algos save 45–93 bps vs. sweep. Beyond 300 bps in elevated volatility: immediate sweep outperforms by ~72 bps. Execution strategy must be regime-dependent, calibrated using Pre-Trade Analytics before order submission. | 300 bps adverse move threshold where sweep becomes competitive |
| Market cap breadth | Arrival slippage is stable across large-cap (BTC, ETH) through low-cap assets – worst observed bucket average near 3.5 bps despite materially wider spreads in lower-cap names. Savings vs. sweep are positive across all market cap tiers (5 bps average; 12.3 bps for medium-cap). | 3.5 bps worst-case arrival slippage across any market cap bucket |
Source: Talos Quant Execution Insights Report 2026. Dataset: 250,000+ parent orders, 618 assets, 2025. Left (green): savings vs. direct sweep. Right (orange): slippage vs. arrival.
Talos's execution alphas are re-calibrated daily from production execution data across all connected markets. Volume predictions drive VWAP child order sizing – concentrating execution in windows of expected higher volume. Spread alphas inform Spread Cost estimates in pre-trade. Volatility predictions drive the Time Risk component. All three are updated dynamically as a function of the execution horizon, not fixed to historical averages.
Physical impact follows a square-root law: doubling execution speed more than doubles market impact. Time risk accumulates linearly with horizon in 24/7 crypto markets with no overnight reset. The optimal execution speed is found at the balance between these two forces – and it varies by asset, order size, and market regime. The TMI model makes this optimization explicit and calculable before any order is submitted.
In adverse but typical conditions (up to 300 bps market move against the order), Talos algos add 45–93 bps of value vs. immediate sweep. In extreme adverse moves (above 300 bps in elevated volatility), the sweep is statistically better. This threshold cannot be known ex ante – which is why pre-trade scenario analysis and in-flight monitoring are the correct operating model, not a fixed execution rule.
Across the full 2025 dataset, the 50–75% maker participation range delivers the lowest arrival slippage at every order size. Below 50%, physical impact grows sharply with order size. Above 75%, time risk begins to dominate and erodes the passive cost benefit. Jointly calibrating duration and participation – rather than targeting 100% maker share – produces the best measurable outcomes.
Comprehensive analysis of 250,000+ parent orders: slippage, sweep savings, maker-taker dynamics, dealer liquidity impact, adverse regime behavior. By Sirui Zhang, Eliad Hoch and Kaan Giray.
talos.com/insights
The full TMI model methodology: spread cost, physical impact (square-root law + sigmoid), and time risk. Calibration approach and coefficient derivation from production data. By Eliad Hoch.
talos.com/insights/an-empirical-model-of-market-impact-in-cryptocurrency-trading
Execution insights through transaction cost analysis: benchmark selection (arrival, VWAP, TWAP), slippage interpretation, markouts, and operational best practice for institutional execution teams.
talos.com/insights/execution-insights-through-transaction-cost-analysis-tca-benchmarks-and-slippage
Access the full execution suite – via the platform or the API
All SOR strategies, algorithmic strategies, and analytics features are accessible via the Talos UI. Pre-trade and post-trade analytics are also available via REST and WebSocket APIs for automated workflows and reporting pipelines. Full documentation at kb.talostrading.com.