Metrics & Analysis

The Metrics & Analysis section helps you understand your execution quality through quantitative measurements. This guide explains each metric and how to interpret your results.

---

Key Metrics Overview

Match Rate

Definition: Percentage of alert signals that resulted in successful broker fills.

Match Rate = (Matched Alerts / Total Alerts) × 100

Match Rate	Interpretation
98%+	Excellent — rare misses
95-98%	Good — occasional issues
90-95%	Acceptable — investigate outliers
<90%	Concerning — audit required

Tip

Match rate below 95% usually indicates systematic issues rather than random failures. Investigate the unmatched alerts for patterns.

Average Latency

Definition: Mean time from alert generation to broker fill.

Avg Latency = Sum(fill_time - alert_time) / matched_count

Latency	Interpretation
<200ms	Excellent — near real-time
200-500ms	Good — typical for retail
500ms-2s	Acceptable — some delay
>2s	Concerning — significant delay

Latency components:

• Network (alert → platform): ~50-100ms
• Platform processing: ~10-50ms
• Network (order → broker): ~50-100ms
• Broker processing: ~10-50ms
• Exchange execution: varies by order type

Average Slippage

Definition: Mean price difference between intended (alert) and actual (fill) prices.

Avg Slippage = Sum(fill_price - alert_price) / matched_count

Slippage	Interpretation
<0.05%	Excellent — minimal impact
0.05-0.1%	Good — acceptable costs
0.1-0.5%	Moderate — notable impact
>0.5%	High — significant erosion

Note

Positive vs Negative: Positive slippage means you got a better price than expected. Negative slippage means you paid more (or received less) than intended.

Total Slippage

Definition: Cumulative dollar impact of slippage across all matched trades.

Total Slippage = Sum((fill_price - alert_price) × quantity)

This represents the real-world cost of execution imperfection.

Profit Retention

Definition: How much of your intended profit you actually captured.

Retention = (Actual P&L / Intended P&L) × 100

Retention	Interpretation
>99%	Excellent execution
97-99%	Good — minor leakage
95-97%	Acceptable — some costs
<95%	Concerning — significant erosion

---

Detailed Analysis Components

Signal-to-Execution Analysis

Breaks down the latency into phases:

Phase	Description	Typical Time
Signal	Alert generated in TradingView	0ms (baseline)
Webhook	Alert received by platform	50-200ms
Processing	Order generated and submitted	10-100ms
Submission	Order sent to broker	50-100ms
Execution	Order filled by exchange	10ms-∞

Tip

If most latency is in a specific phase, focus optimization efforts there.

Trade Timing Analysis

Histogram showing distribution of latencies:

What to look for:

• Tight cluster — Consistent execution (good)
• Long tail — Occasional delays (investigate)
• Bimodal — Two distinct behaviors (systematic issue)

Strategy Insights

The system grades your execution (A-F) based on:

• Match rate weighting: 40%
• Latency weighting: 30%
• Slippage weighting: 30%

Grade	Meaning
A	Excellent execution across all metrics
B	Good with minor improvement areas
C	Acceptable but optimization recommended
D	Concerns that should be addressed
F	Significant issues requiring immediate attention

---

Trade Comparison Grid

The detailed grid shows every alert-fill pair:

Grid Columns

Column	Description
Status	Match type badge
Alert Time	When signal was generated
Alert Price	Intended execution price
Fill Time	When broker executed
Fill Price	Actual execution price
Latency	Time difference (ms or seconds)
Slippage	Price difference ($ and %)

Filtering Options

Filter	Purpose
Show Only Problems	Hide fully matched trades
Match Type	Filter by Full/Entry Only/Exit Only/Unmatched
Latency Threshold	Show only high-latency trades
Slippage Threshold	Show only high-slippage trades
Quick Filter	Search across all columns

Sorting

Click column headers to sort:

• By time (chronological analysis)
• By latency (find slowest)
• By slippage (find costliest)
• By status (group by type)

---

Trade Selection & Removal

You can select and remove individual trades from your analysis to focus on specific subsets or exclude outliers.

Selecting Trades

In Comparison Grid:

• Check the box next to any trade row
• Select multiple rows by checking multiple boxes
• Use “Select All” to choose all visible trades

In Alpaca Fills Grid:

• Check boxes next to fills you want to exclude
• Select multiple fills at once
• Selected rows are highlighted

In TradingView Alerts Grid:

• Check boxes next to alerts you want to exclude
• Select multiple alerts at once
• Selected rows are highlighted

Removing Selected Trades

1. Select trades you want to exclude from analysis
2. Click “Remove selected” button
3. Selected trades are hidden from calculations
4. Analysis automatically recalculates without removed trades

What happens:

• Removed trades disappear from grids
• Metrics recalculate without excluded trades
• Charts update to exclude removed markers
• Comparison data excludes removed pairs

Note

Trades are hidden, not deleted. Removed trades can be restored at any time. This is useful for filtering out test trades, outliers, or trades from different strategies.

Resetting Removals

To restore all removed trades:

1. Click the “Reset” button (appears when trades are removed)
2. All previously removed trades reappear
3. Analysis recalculates with full dataset
4. Metrics return to original values

Use cases:

• Compare analysis with/without specific trades
• Test impact of outliers
• Filter out test or manual trades
• Focus on specific time periods

---

Export Functionality

Export your complete analysis data for external analysis, record-keeping, or sharing.

What’s Included in Export

The export is a comprehensive JSON file containing:

Metadata:

• Export timestamp
• Export version
• Configuration used

Configuration:

• Symbol(s) analyzed
• Date range
• Comparison settings (time/price tolerances)
• Selected API key
• Layout mode
• Timeframe

Trade Data:

• Alpaca trade summaries
• TradingView trade summaries
• Full comparison pairs with details
• Unmatched alerts
• Unmatched fills

Analysis Results:

• Match statistics
• Difference summaries (latency, slippage)
• Portfolio balances (start/end)
• Portfolio history
• Removed trade IDs (for reference)

Export Format

• File type: JSON (JavaScript Object Notation)
• Structure: Hierarchical, well-organized
• Size: Varies based on number of trades
• Compatibility: Can be imported into Excel, Python, R, or other analysis tools

How to Export

1. Complete your alerts vs fills analysis
2. Click the “Export” button in the page header
3. File downloads automatically
4. Save the file for future reference

Use Cases

External Analysis:

• Import into Excel for custom calculations
• Analyze in Python/R for statistical analysis
• Create custom visualizations
• Perform advanced data mining

Record Keeping:

• Save analysis results for compliance
• Maintain audit trail
• Document execution quality over time
• Create reports for stakeholders

Sharing:

• Share data with team members
• Send to advisors or consultants
• Provide to broker for review
• Archive for historical reference

Backup:

• Keep local copies of important analyses
• Preserve data before making changes
• Create snapshots for comparison

Tip

Export before changes: Export your analysis before removing trades or changing settings, so you have a record of the original results.

---

Interpreting Results

Healthy Execution Profile

✅ Match Rate: 97%

Most alerts executed successfully.

✅ Avg Latency: 320ms

Within normal retail range.

✅ Avg Slippage: 0.08%

Minimal price impact.

✅ Retention: 98.5%

Keeping most intended profits.

Action: Monitor periodically, no immediate changes needed.

Concerning Execution Profile

⚠️ Match Rate: 88%

12% of signals not executing.

⚠️ Avg Latency: 1.8s

Delays may cause price movement.

⚠️ Avg Slippage: 0.35%

Notable execution costs.

⚠️ Retention: 94%

Losing 6% to execution issues.

Action: Investigate unmatched alerts, review infrastructure, consider order type changes.

---

Common Metric Patterns

High Latency + Low Slippage

Interpretation: Orders are slow but price impact is minimal.

Likely causes:

• Limit orders (intentionally wait for price)
• Network delays without market movement
• Low-volatility instruments

Low Latency + High Slippage

Interpretation: Fast execution but at poor prices.

Likely causes:

• Market orders during volatility
• Low liquidity
• Wide bid-ask spreads

Low Match Rate + Normal Latency/Slippage

Interpretation: When orders execute, they’re fine — but many don’t execute.

Likely causes:

• Order rejections
• Insufficient buying power
• Alert failures

Declining Metrics Over Time

Interpretation: Execution quality degrading.

Likely causes:

• Infrastructure issues
• Market condition changes
• Account limitations

---

Optimization Recommendations

Improving Match Rate

1. Check alert configuration — Ensure all signals fire alerts
2. Verify buying power — Sufficient margin/cash
3. Review order types — Market vs limit behavior
4. Check Alpaca status — API health

Reducing Latency

1. Network optimization — Stable, fast connection
2. Reduce processing — Simplify alert payloads
3. Geographic proximity — Server location matters
4. Order type selection — Market orders execute faster

Minimizing Slippage

1. Use limit orders — Control execution price
2. Trade liquid assets — Tighter spreads
3. Avoid volatility — Don’t trade at market open
4. Size positions — Smaller orders fill better

---

Related Sections

• Scope & Matching — Configuration
• Visual Audit — Chart analysis
• Troubleshooting — Common issues