Investment Alternative Data MCP Server

Investment alternative data intelligence — institutional 13F convergence, congressional trading alpha, macro regime classification, and innovation momentum scoring — delivered via the Model Context Protocol. Built for hedge funds, quant analysts, corporate strategy teams, and family offices that need signals beyond price and volume.

This server wraps 13 Apify actors spanning SEC filings, Federal Reserve economic data, congressional disclosures, USPTO patents, and consumer sentiment into 8 structured MCP tools. Each tool fetches data in parallel, runs a scoring algorithm, and returns structured JSON with scores, classifications, raw data, and actionable findings in one call.

What investment data can you access?

Why use Investment Alternative Data MCP Server?

Building an edge in public markets is expensive. A Bloomberg Terminal costs $24,000/year. FactSet runs $12,000+. Quandl alternative data subscriptions start at $500/month per dataset — and most only cover one data type. For portfolio managers, analysts, and systematic traders who need breadth without enterprise contracts, the cost barrier is prohibitive.

This server eliminates that barrier. For $0.045 per query you get parallel access to 13 data sources — SEC filings, Fed data, congressional disclosures, patent records, consumer stress indicators, and cross-asset signals — synthesized into a single scored response.

Beyond the data, you get the full Apify platform behind every query:

• Scheduling — run daily congressional trade monitoring or weekly macro regime checks via cron triggers
• API access — call any tool from Python, JavaScript, TypeScript, or any HTTP client using standard MCP protocol
• Parallel execution — each tool dispatches up to 5 actor calls simultaneously, completing in the time of a single serial request
• Monitoring — get Slack or email alerts when spending limits are reached or tools return unexpected results
• Integrations — connect outputs to Zapier, Make, Google Sheets, webhooks, or trading platform APIs

Features

• Smart Money Convergence Score (0-100) — counts distinct institutional filers in 13F-HR data: 10+ filers scores 85, 5-9 scores 60, 2-4 scores 35, adjusted by insider buy/sell ratio (+10 for BULLISH, -10 for BEARISH)
• Insider sentiment classification — classifies Form 4 transactions as BULLISH (buy ratio >0.65), BEARISH (sell ratio >0.65), or NEUTRAL, then cross-validates against institutional convergence
• Congressional Alpha Score (-100 to +100) — net score derived from congressional buy/sell transaction counts; identifies top 10 traders by volume with direction
• Macro regime classification — EXPANSION / PEAK / CONTRACTION / TROUGH with confidence score, driven by FRED GDP, UNRATE, T10Y2Y yield curve inversion detection, BLS CPI, IMF forecasts, and OECD composite leading indicators
• Innovation Momentum Score (0-100) — patent filing velocity thresholds (50+ = ACCELERATING, 20-50 = STEADY) combined with R&D hiring signal (20+ engineering/research/scientist postings = STRONG)
• Consumer Stress Index (0-100) with 5-tier severity — LOW / MODERATE / ELEVATED / HIGH / SEVERE, derived from CFPB complaint volume, BLS CPI, FRED savings rate (PSAVERT), and University of Michigan consumer sentiment (UMCSENT)
• VIX-based risk regime detection — classifies market as RISK-ON (<20), CAUTIOUS (20-30), or RISK-OFF (>30) from FRED VIXCLS series
• Sector rotation recommendations — regime-to-sector mapping: EXPANSION favors cyclicals, PEAK triggers defensive rotation, CONTRACTION positions for treasuries and healthcare
• Parallel data fetching — runActorsParallel() dispatches up to 5 simultaneous actor calls per tool using Promise.all() to minimize latency
• Inverted yield curve detection — T10Y2Y < 0 triggers a double-weight contraction signal (counts as 2 contraction signals), reflecting its historical predictive power
• Company fundamental scan — combines SEC EDGAR analysis, Finnhub market data, insider trading, and patent data in a single rapid screening call
• Cross-asset correlation — simultaneously pulls equities (Finnhub), crypto (CoinGecko), FX (Exchange Rate Tracker), VIX, 10Y yield, and yield curve for tail risk analysis
• Pay-per-event pricing — charged only when a tool executes; no standby compute charges between calls
• Stateless MCP transport — each POST to /mcp creates a fresh StreamableHTTPServerTransport session, eliminating state management overhead for LLM clients

Use cases for investment alternative data

Hedge fund research and idea generation

Fundamental analysts at long/short equity funds use smart_money_tracker to validate investment theses before taking a position. If 12 institutional filers are increasing exposure to a mid-cap biotech and insiders are net buying, it corroborates independent research without requiring access to private channel checks. Pair with company_fundamental_scan for a rapid first-pass screen of a watch list.

Political and legislative risk monitoring

Policy-sensitive portfolios — energy, healthcare, defense, financials — benefit from congressional_alpha_signals. Members of Congress must disclose trades within 45 days under the STOCK Act. Net congressional buying in a sector ahead of major legislation is a documented alpha source. Use it as a complement to traditional lobbying data.

Macro regime-based asset allocation

Multi-asset allocators and CIOs use macro_regime_report to time tactical tilts. The tool classifies the current economic phase (Expansion / Peak / Contraction / Trough) with a confidence percentage and recommends sector positioning. Run it weekly alongside an IMF/OECD update cycle to track regime transitions before they are broadly priced in.

R&D-intensive growth stock screening

Quantitative analysts screen for innovation-driven alpha using innovation_momentum_score. Companies with accelerating patent filings combined with aggressive R&D hiring represent early-stage investment in future revenue — before it appears in earnings. Apply this to technology, biotech, and semiconductor sectors where patent velocity leads financial performance by 12-24 months.

Consumer credit and sector risk assessment

Credit analysts and consumer sector PMs use consumer_stress_indicator to anticipate rising default rates and consumer discretionary weakness. The tool cross-references CFPB complaint volumes (an early warning of financial distress), BLS inflation data, and FRED savings rate and consumer sentiment. Elevated stress predates deteriorating retail earnings and rising credit card charge-off rates.

Cross-asset risk and portfolio correlation monitoring

Risk managers and multi-strategy funds use cross_asset_correlation to detect regime shifts between risk-on and risk-off states. When equities, crypto, and high-beta FX are moving together, it signals a risk-on regime. When they diverge with VIX spiking above 30, it triggers defensive repositioning. The tool retrieves equities, crypto, FX, VIX, 10Y yield, and yield curve in a single parallel call.

How to connect Investment Alternative Data MCP Server to your AI client

Step 1: Get your Apify API token

Log in to the Apify Console, navigate to Settings > Integrations, and copy your API token.

Step 2: Add the MCP server to your client

The MCP endpoint URL is:

https://investment-alternative-data-mcp.apify.actor/mcp

Pass your Apify token as a Bearer token in the Authorization header, or via the token query parameter.

Step 3: Start querying

Call any of the 8 tools with a natural language query from your AI assistant. Each tool accepts structured arguments described in the tools section below.

Step 4: Retrieve results

Results are returned as structured JSON directly in the MCP response with scores, classifications, raw data, and findings strings explaining each result.

MCP tools

Tool parameters

smart_money_tracker

congressional_alpha_signals

macro_regime_report

innovation_momentum_score

consumer_stress_indicator

sector_rotation_signals

company_fundamental_scan

cross_asset_correlation

Connection examples

Claude Desktop — add to claude_desktop_config.json:

{
  "mcpServers": {
    "investment-alternative-data": {
      "url": "https://investment-alternative-data-mcp.apify.actor/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_APIFY_TOKEN"
      }
    }
  }
}

Cursor / Windsurf / Cline — use the same URL in your MCP server configuration panel with the Authorization header.

Python (direct HTTP):

import requests

response = requests.post(
    "https://investment-alternative-data-mcp.apify.actor/mcp",
    headers={
        "Content-Type": "application/json",
        "Authorization": "Bearer YOUR_APIFY_TOKEN",
    },
    json={
        "jsonrpc": "2.0",
        "method": "tools/call",
        "params": {
            "name": "smart_money_tracker",
            "arguments": {
                "company_name": "NVIDIA",
                "cik": "0001045810"
            }
        },
        "id": 1
    }
)
result = response.json()
print(result["result"]["content"][0]["text"])

JavaScript:

const response = await fetch(
  "https://investment-alternative-data-mcp.apify.actor/mcp",
  {
    method: "POST",
    headers: {
      "Content-Type": "application/json",
      "Authorization": "Bearer YOUR_APIFY_TOKEN",
    },
    body: JSON.stringify({
      jsonrpc: "2.0",
      method: "tools/call",
      params: {
        name: "macro_regime_report",
        arguments: {
          fred_series: ["GDP", "UNRATE", "T10Y2Y", "CPIAUCSL", "PSAVERT", "UMCSENT"],
          country: "US"
        }
      },
      id: 1,
    }),
  }
);
const data = await response.json();
console.log(data.result.content[0].text);

cURL:

curl -X POST "https://investment-alternative-data-mcp.apify.actor/mcp" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_APIFY_TOKEN" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {
      "name": "congressional_alpha_signals",
      "arguments": { "query": "NVIDIA" }
    },
    "id": 1
  }'

Output examples

smart_money_tracker response for NVIDIA:

{
  "company": "NVIDIA",
  "smartMoneyScore": 95,
  "netDirection": "ACCUMULATING",
  "insiderSentiment": "BULLISH",
  "findings": [
    "14 institutional filers holding positions — strong convergence",
    "Insider sentiment BULLISH: 8 buys vs 2 sells"
  ],
  "data": {
    "thirteenFFilings": 14,
    "insiderTrades": 10,
    "edgarFilings": 5
  },
  "rawInsiderTrades": [
    {
      "name": "Jensen Huang",
      "transactionType": "Purchase",
      "shares": 25000,
      "pricePerShare": 875.40,
      "date": "2025-10-14"
    }
  ]
}

macro_regime_report response:

{
  "regime": "PEAK",
  "confidence": 61,
  "indicators": [
    { "name": "GDP Growth", "signal": "EXPANSION", "value": 2.4 },
    { "name": "Unemployment Rate", "signal": "EXPANSION", "value": 3.9 },
    { "name": "Yield Curve (10Y-2Y)", "signal": "CONTRACTION", "value": -0.18 },
    { "name": "CPI/Inflation", "signal": "OVERHEATING", "value": 3.2 }
  ],
  "findings": [
    "Inverted yield curve — historically strongest recession predictor",
    "Inflation at 3.2% — overheating risk",
    "Macro regime: PEAK (3 expansion signals, 2 contraction signals)"
  ],
  "rawData": {
    "fred": [{ "seriesId": "GDP", "value": "2.4", "date": "2025-09-30" }]
  }
}

consumer_stress_indicator response for mortgages:

{
  "stressIndex": 48,
  "level": "ELEVATED",
  "components": [
    {
      "name": "Consumer Complaints",
      "contribution": 20,
      "detail": "87 CFPB mortgage complaints — elevated distress"
    },
    {
      "name": "Inflation (CPI)",
      "contribution": 13,
      "detail": "CPI at 4.3% — eroding purchasing power"
    },
    {
      "name": "Savings Rate",
      "contribution": 15,
      "detail": "Personal savings rate at 3.8% — consumers not saving"
    }
  ],
  "sector": "mortgages",
  "data": {
    "cfpbComplaints": 87,
    "topComplaintTypes": [
      { "type": "Mortgage", "count": 52 },
      { "type": "Loan servicing", "count": 21 }
    ]
  }
}

Output fields

How much does investment alternative data cost?

This server uses pay-per-event pricing — you pay $0.045 per tool call. Platform compute costs are included. There is no subscription, no monthly minimum, and no charge when the server is idle.

You can set a maximum spending limit per run in the Apify Console to control costs. The server stops charging when your budget is reached and returns a structured error response.

Compare this to Bloomberg Terminal at $2,000/month, FactSet at $1,000+/month, or Quandl alternative data bundles at $500+/month per dataset. Most active analyst users spend $5-25/month with no subscription commitment.

How Investment Alternative Data MCP Server works

Data collection phase

Each tool call triggers runActorsParallel(), which wraps the Apify client's actor().call() in a Promise.all() array. The macro regime tool, for example, dispatches FRED, BLS, IMF, OECD, and exchange rate actors simultaneously with a 120-second timeout each. Actors run on Apify's infrastructure at 256MB memory allocation per sub-call. Results are collected from each actor's default dataset via dataset.listItems() before scoring begins.

Scoring and classification phase

Raw API results flow into five purpose-built scoring functions in scoring.ts. The Smart Money function iterates 13F filers into a Set for unique-filer counting, then applies convergence thresholds (10+ = 85 score). The Macro Regime classifier counts expansion vs. contraction signals across all FRED series, with yield curve inversion (T10Y2Y < 0) weighted double. The Consumer Stress function applies weighted contributions: CFPB volume (up to 30 points), CPI (up to 25 points), savings rate (15 points), and University of Michigan sentiment (15 points).

MCP transport layer

The server uses @modelcontextprotocol/sdk v1.17.5 with StreamableHTTPServerTransport in stateless mode (sessionIdGenerator: undefined). Each POST to /mcp creates a new transport and server instance, making the server horizontally scalable and compatible with any MCP client that supports HTTP transport — Claude Desktop, Cursor, Windsurf, Cline, and custom integrations. The /mcp endpoint returns 405 for GET and DELETE to enforce stateless operation.

Standby mode operation

The actor runs in Apify Standby mode (usesStandbyMode: true), meaning it stays alive between calls on the Apify platform, serving MCP requests at https://investment-alternative-data-mcp.apify.actor/mcp without cold-start latency. In non-standby context (direct actor run), it starts a brief health-check server on port 8080 and exits cleanly after 1 second.

Tips for best results

1. Use company_fundamental_scan before smart_money_tracker — the fundamental scan fetches EDGAR, Finnhub, insider, and patent data in one call. Use it first to confirm the company is worth deeper analysis, then run smart_money_tracker for the institutional convergence picture.

2. Combine macro_regime_report with sector_rotation_signals — macro_regime_report gives you the regime classification and confidence; sector_rotation_signals maps that regime to specific sectors and overlays congressional trading patterns. Run both for a complete allocation signal.

3. Filter congressional_alpha_signals by sector, not just company — leave query blank and scan all recent trades to identify which sectors congress members are most active in. Sector-level patterns are often more actionable than single-stock signals.

4. Set include_job_data: false for innovation_momentum_score when speed matters — the job market actor adds one parallel call. When screening a long list rapidly, disable it to cut per-call time roughly in half, then re-enable for your highest-conviction candidates.

5. Monitor consumer_stress_indicator by sector — pass specific values like "credit cards", "mortgages", or "student loans" to the sector parameter to get CFPB complaint data focused on the financial product category most relevant to your exposure.

6. Schedule macro_regime_report weekly — macro regimes shift slowly. A weekly Apify schedule run costs $0.045/week ($2.34/year) and keeps your regime classification current for allocation decisions. Use a webhook to push results to Slack or a Google Sheet.

7. Use cross_asset_correlation before earnings season — VIX elevation above 25 combined with yield curve inversion and crypto-equity correlation breaking down is a historically reliable pre-earnings risk-off pattern. Run this tool in the 2 weeks before major earnings windows.

Combine with other MCP servers and actors

Limitations

• 13F data is delayed by 45 days — quarterly filings are due 45 days after quarter end. The smart money tracker shows the most recent available positions, which can be 1-4 months old. Real-time institutional flow data requires prime brokerage access.
• Congressional disclosures have a 45-day filing window — some congressional members file close to the deadline. Very recent trades may not yet appear. Use this tool for pattern detection, not real-time monitoring.
• Macro regime confidence degrades without FRED data — if the FRED actor returns no results for the requested series, the classifier defaults to UNCERTAIN with 0% confidence. This can occur if a series ID is mistyped or temporarily unavailable.
• Patent data reflects filings, not grants — USPTO data includes applications. A company with 50 patent applications may have fewer granted patents. Patent velocity measures R&D activity, not IP quality.
• Innovation Momentum does not control for company size — 10 patents from a 50-person startup signals different intensity than 10 patents from a Fortune 500. Normalize by employee count for cross-company comparisons.
• Consumer Stress Index is US-centric — CFPB complaints, BLS data, and FRED series are primarily US data. For international consumer stress analysis, supplement with IMF and OECD tools directly.
• Cross-asset correlation is observational — the tool reports asset data and VIX levels but does not compute a mathematical correlation coefficient. Quantitative correlation analysis requires a time-series database with historical data.
• Scoring models use threshold-based logic — the scoring functions use defined thresholds (e.g., 10+ filers = 85 score) rather than machine learning models trained on historical data. Treat scores as screening signals, not precise predictions.

Integrations

• Claude Desktop — add the MCP URL to claude_desktop_config.json to query investment data in natural language from the Claude chat interface
• Cursor — configure as an MCP server in Cursor settings to query market intelligence while writing investment research or analysis code
• Apify API — call the actor directly via the Apify REST API for custom integrations with trading platforms or research tools
• Zapier — trigger weekly macro regime reports and push results to Google Sheets, Notion, or email
• Make — build automated workflows that run congressional trade monitoring on a schedule and alert on net direction changes
• Webhooks — receive POST notifications when a macro regime change or consumer stress threshold is breached
• LangChain / LlamaIndex — embed investment alternative data queries into AI research agent pipelines for automated investment memo generation

Troubleshooting

Tool returns empty findings or zero scores despite valid input — The sub-actors may have returned no results for the query. Check that the company name is spelled correctly and matches SEC filing records. For EDGAR-based tools, try providing the CIK number directly instead of a name string. If the FRED actor is returning empty data, verify the series IDs match the format in the FRED series browser.

Spending limit reached error on first call — Your Apify account may have a low default per-run spending limit. Navigate to your Apify Console run settings and increase the maximum cost per run. The free tier includes $5/month, which covers approximately 111 tool calls.

Macro regime classifies as UNCERTAIN — This occurs when the FRED actor returns no data matching the specified series IDs. The default series list (GDP,UNRATE,T10Y2Y,CPIAUCSL,PSAVERT,UMCSENT) should return data reliably. If you have customized fred_series, verify each ID against the FRED database.

MCP client cannot connect — Ensure you are sending requests to the correct URL (https://investment-alternative-data-mcp.apify.actor/mcp) with a valid Apify API token in the Authorization header. The endpoint only accepts POST requests. GET requests return 405.

Congressional alpha tool returns NO_DATA — The Congressional Stock Tracker actor may return no records matching your query. Try broader queries or leave both query and member blank to retrieve all recent congressional trades, then filter the response manually.

Responsible use

• This server accesses only publicly available government and regulatory filings: SEC EDGAR, CFPB database, STOCK Act disclosures, USPTO, FRED, BLS, IMF, and OECD data.
• Congressional trading data is public information mandated by the STOCK Act (2012) and is legally available for research and analysis.
• Do not use investment signals derived from this tool as the sole basis for investment decisions. Past signal patterns are not a guarantee of future performance.
• Comply with applicable securities laws when acting on investment intelligence. Consult a licensed investment adviser before making financial decisions.
• For guidance on web scraping legality, see Apify's guide.

FAQ

How current is the 13F institutional holdings data for smart money tracking? 13F-HR filings are quarterly with a 45-day reporting deadline. The smart money tracker reflects the most recently filed quarter, which can be 1-4 months old depending on when you query. Form 4 insider trading data is more timely, with a 2-business-day filing requirement after the transaction.

How does investment alternative data differ from a Bloomberg Terminal? Bloomberg provides real-time Level 2 market data, news feeds, and chat. This server provides alternative data signals that Bloomberg does not: congressional trading patterns, patent filing velocity, CFPB consumer stress scoring, and institutional convergence analysis. Use them as complementary tools, not substitutes.

Is it legal to analyze congressional stock trading data? Yes. Congressional stock trading disclosures are public records mandated by the STOCK Act (2012). Every trade must be disclosed within 45 days. Analyzing this public data for investment research is entirely legal. See Apify's web scraping legality guide.

How accurate is the macro regime classification? The classifier uses FRED GDP, UNRATE, T10Y2Y, CPI, savings rate, and consumer sentiment with IMF/OECD leading indicators. The confidence score reflects the ratio of expansion to contraction signals. With default series, expect 60-85% confidence in non-transitional periods. At regime inflection points, confidence drops to 40-60%, signaling genuine uncertainty.

What does a Smart Money Score of 85 actually mean? It means 10 or more distinct institutional filers appear in 13F-HR filings for that company, adjusted upward if insiders are net buying. A score of 85+ indicates strong institutional consensus. It does not indicate future price direction — high convergence can also signal crowded trade risk if sentiment shifts.

Can I use investment alternative data signals to automatically execute trades? You can integrate tool outputs into automated workflows via webhooks or the Apify API. However, this server does not connect to brokerage APIs or execute trades. All outputs are analytical signals for human review or further processing by your own trading infrastructure.

How many alternative data queries can I run per month on the free tier? Apify's free plan includes $5 of monthly credits. At $0.045 per tool call, that covers approximately 111 queries per month at no cost. The first call is typically free as part of the Apify trial.

Does the Congressional Alpha Score detect illegal insider trading by congress members? No. The score measures the net direction of disclosed trades (legally required filings). It cannot detect undisclosed trades. Research papers documenting congressional trading alpha refer to the pattern of disclosed trades, not illegal activity.

How does Innovation Momentum Score compare to Patsnap or Derwent? Patsnap and Derwent provide deep patent analytics with citation networks, litigation history, and technology classification. Innovation Momentum Score is a lightweight velocity signal using USPTO filing counts and hiring data — sufficient for screening, not for IP due diligence.

Can I schedule this MCP server to run automatically without an AI client? Yes. Trigger any tool via the Apify Actor API directly using Python, JavaScript, or cURL. The Apify platform has built-in cron scheduling to run macro regime reports or congressional alerts daily and push results to a webhook, Slack, or Google Sheet.

What happens if one of the 13 underlying actors fails during a parallel call? runActorsParallel() wraps each actor call in a try/catch. A failing actor returns an empty array. The scoring function receives partial data and produces a lower-confidence result; findings strings indicate when data was insufficient.

How is this different from Quandl or YipitData? Quandl and YipitData provide curated, backtested datasets with historical depth. This server provides real-time query access to public government data sources with scoring overlays at $0.045/query with no subscription — lower cost for ad-hoc research; institutional vendors are better for systematic backtesting.

Help us improve

If you encounter issues, you can help us debug faster by enabling run sharing in your Apify account:

1. Go to Account Settings > Privacy
2. Enable Share runs with public Actor creators

This lets us see your run details when something goes wrong, so we can fix issues faster. Your data is only visible to the actor developer, not publicly.

Support

Found a bug or have a feature request? Open an issue in the Issues tab on this actor's page. For custom solutions or enterprise integrations, reach out through the Apify platform.