Financial Modeling Prep (FMP) API Guide

Financial Modeling Prep (FMP) API Guide

Overview

Financial Modeling Prep provides comprehensive financial data APIs for stocks, forex, cryptocurrencies, and more. This guide focuses on endpoints used for dividend growth stock screening with RSI technical indicators.

Two-Stage Screening Approach

This screener supports two screening modes:

1. Two-Stage (FINVIZ + FMP) - RECOMMENDED

• Stage 1: FINVIZ Elite API pre-screens with RSI filter (1 API call → 10-50 candidates)
• Stage 2: FMP API performs detailed fundamental analysis on pre-screened candidates
• Advantage: Dramatically reduces FMP API calls by leveraging FINVIZ's technical filters
• Cost: FINVIZ Elite $40/month + FMP free tier

2. FMP-Only (Original Method)

• Single Stage: FMP stock-screener + detailed analysis
• Limitation: FMP free tier (250 requests/day) limits to ~40 stocks
• Cost: FMP free tier (upgrades available)

Recommended: Two-stage approach for regular screening (daily/weekly) to maximize coverage while minimizing costs.

API Key Setup

Obtaining an API Key

1. Visit https://financialmodelingprep.com/developer/docs
2. Sign up for a free account
3. Navigate to Dashboard → API Keys
4. Copy your API key

Free Tier Limits

• 250 requests per day
• Rate limit: ~5 requests per second
• No credit card required
• Sufficient for daily/weekly screening runs

Paid Tiers (Optional)

• Starter ($14/month): 500 requests/day
• Professional ($29/month): 1,000 requests/day
• Enterprise ($99/month): 10,000 requests/day

Setting API Key

Method 1: Environment Variable (Recommended)

Linux/macOS:

export FMP_API_KEY=your_api_key_here

Windows (Command Prompt):

set FMP_API_KEY=your_api_key_here

Windows (PowerShell):

$env:FMP_API_KEY="your_api_key_here"

Persistent (add to shell profile):

# Add to ~/.bashrc or ~/.zshrc
echo 'export FMP_API_KEY=your_api_key_here' >> ~/.bashrc
source ~/.bashrc

Method 2: Command-Line Argument

python3 scripts/screen_dividend_growth_rsi.py --fmp-api-key your_api_key_here

FINVIZ API Setup (Optional - For Two-Stage Screening)

Obtaining FINVIZ Elite API Key

1. Visit https://elite.finviz.com
2. Subscribe to FINVIZ Elite ($40/month or $400/year)
3. Navigate to Settings → API
4. Copy your API key

Setting FINVIZ API Key

Environment Variable (Recommended):

export FINVIZ_API_KEY=your_finviz_key_here

Command-Line Argument:

python3 screen_dividend_growth_rsi.py --use-finviz --finviz-api-key YOUR_KEY

FINVIZ Pre-Screening Filters

When using --use-finviz, the screener applies these filters via FINVIZ Elite API:

• Market Cap: Mid-cap or higher (≥$2B)
• Dividend Yield: 0.5-3% (captures dividend growers, excludes high-yield REITs/utilities)
• Dividend Growth (3Y): 10%+ (FMP verifies 12%+)
• EPS Growth (3Y): 5%+ (positive earnings momentum)
• Sales Growth (3Y): 5%+ (positive revenue momentum)
• RSI (14-period): Under 40 (oversold/pullback)
• Geography: USA

Output: Set of stock symbols (typically 30-50 stocks) passed to FMP for detailed analysis.

Rationale for balanced filters:

• 10%+ dividend growth ensures high-quality dividend compounders
• 5%+ EPS/sales growth captures growing businesses (not just mature dividend payers)
• 0.5-3% yield range excludes mature high-yielders (>4%) and focuses on growth stocks
• Reduces FINVIZ candidates from ~90 to ~30-50 stocks
• Stays within FMP free tier limits with efficient analysis (250 requests/day)

Key Endpoints Used

1. Stock Screener

Endpoint: /v3/stock-screener

Purpose: Initial filtering by market cap and exchange

Parameters:

• marketCapMoreThan: Minimum market cap (e.g., 2000000000 = $2B)
• exchange: Exchanges to include (e.g., "NASDAQ,NYSE")
• limit: Max results (default: 1000)

Note: This screener does NOT pre-filter by dividend yield (unlike value-dividend-screener). We retrieve a broader universe and calculate actual yields from dividend history to ensure accuracy.

Example Request:

https://financialmodelingprep.com/api/v3/stock-screener?
  marketCapMoreThan=2000000000&
  exchange=NASDAQ,NYSE&
  limit=1000&
  apikey=YOUR_API_KEY

Response Format:

[
  {
    "symbol": "AAPL",
    "companyName": "Apple Inc.",
    "marketCap": 2800000000000,
    "sector": "Technology",
    "industry": "Consumer Electronics",
    "price": 185.50,
    "exchange": "NASDAQ",
    "isActivelyTrading": true
  }
]

2. Historical Dividend

Endpoint: /v3/historical-price-full/stock_dividend/{symbol}

Purpose: Dividend history for growth rate calculation and yield verification

Example Request:

https://financialmodelingprep.com/api/v3/historical-price-full/stock_dividend/AAPL?
  apikey=YOUR_API_KEY

Response Format:

{
  "symbol": "AAPL",
  "historical": [
    {
      "date": "2024-11-08",
      "label": "November 08, 24",
      "adjDividend": 0.25,
      "dividend": 0.25,
      "recordDate": "2024-11-11",
      "paymentDate": "2024-11-14",
      "declarationDate": "2024-10-31"
    }
  ]
}

Usage in Script:

• Aggregate dividends by calendar year (sum all payments in each year)
• Calculate 3-year dividend CAGR: ((Div_Year3 / Div_Year0) ^ (1/3) - 1) × 100
• Extract latest annual dividend for yield calculation
• Verify consistency (no significant cuts year-over-year)

3. Historical Prices (NEW for RSI)

Endpoint: /v3/historical-price-full/{symbol}

Purpose: Daily price data for RSI calculation

Parameters:

• symbol: Stock ticker
• timeseries: Number of days to retrieve (e.g., 30)

Example Request:

https://financialmodelingprep.com/api/v3/historical-price-full/AAPL?
  timeseries=30&
  apikey=YOUR_API_KEY

Response Format:

{
  "symbol": "AAPL",
  "historical": [
    {
      "date": "2024-11-01",
      "open": 184.50,
      "high": 186.20,
      "low": 183.80,
      "close": 185.50,
      "adjClose": 185.50,
      "volume": 45000000,
      "unadjustedVolume": 45000000,
      "change": 1.00,
      "changePercent": 0.54,
      "vwap": 185.10,
      "label": "November 01, 24",
      "changeOverTime": 0.0054
    }
  ]
}

Usage in Script:

• Extract close prices from last 30 days
• Sort chronologically (oldest first)
• Calculate 14-period RSI:
  1. Compute price changes (close[i] - close[i-1])
  2. Separate gains and losses
  3. Calculate average gain and average loss over 14 periods
  4. RS = Average Gain / Average Loss
  5. RSI = 100 - (100 / (1 + RS))

RSI Filter: Stocks with RSI > 40 are excluded (not oversold/pullback)

4. Income Statement

Endpoint: /v3/income-statement/{symbol}

Purpose: Revenue, EPS, net income analysis

Parameters:

• symbol: Stock ticker (e.g., "AAPL")
• limit: Number of periods (e.g., 5 for 5 years)
• period: "annual" (default)

Example Request:

https://financialmodelingprep.com/api/v3/income-statement/AAPL?
  limit=5&
  apikey=YOUR_API_KEY

Key Fields Used:

• revenue: Total revenue
• eps: Earnings per share
• netIncome: Net income (for payout ratio calculation)
• date: Fiscal period end date

Usage in Script:

• Calculate 3-year revenue CAGR (must be positive for qualification)
• Calculate 3-year EPS CAGR (must be positive)
• Extract net income for payout ratio calculation

5. Balance Sheet Statement

Endpoint: /v3/balance-sheet-statement/{symbol}

Purpose: Debt, equity, liquidity analysis

Parameters:

• symbol: Stock ticker
• limit: Number of periods (typically 5)

Key Fields Used:

• totalDebt: Total debt (short-term + long-term)
• totalStockholdersEquity: Shareholders' equity
• totalCurrentAssets: Current assets
• totalCurrentLiabilities: Current liabilities

Usage in Script:

• Debt-to-Equity: totalDebt / totalStockholdersEquity (must be < 2.0)
• Current Ratio: totalCurrentAssets / totalCurrentLiabilities (must be > 1.0)
• Financial health check (both ratios must pass)

6. Cash Flow Statement

Endpoint: /v3/cash-flow-statement/{symbol}

Purpose: Free cash flow analysis for dividend sustainability

Parameters:

• symbol: Stock ticker
• limit: Number of periods

Key Fields Used:

• freeCashFlow: Free cash flow (OCF - Capex)
• dividendsPaid: Actual dividends paid (negative value)

Usage in Script:

• FCF Payout Ratio: dividendsPaid / freeCashFlow
• Validates dividend is covered by cash generation (< 100% is sustainable)

7. Key Metrics

Endpoint: /v3/key-metrics/{symbol}

Purpose: ROE, profit margins, valuation ratios

Parameters:

• symbol: Stock ticker
• limit: Number of periods (typically 1 for latest)

Key Fields Used:

• roe: Return on Equity (decimal, e.g., 0.15 = 15%)
• netProfitMargin: Net profit margin (decimal)
• peRatio: Price-to-Earnings ratio
• pbRatio: Price-to-Book ratio
• numberOfShares: Shares outstanding (for payout ratio calculation)

Usage in Script:

• ROE: Quality metric for composite scoring (higher is better)
• Profit Margin: Profitability metric for scoring
• P/E and P/B: Context only (no exclusionary limits in this screener)
• Number of Shares: Used to calculate total dividend payout

Rate Limiting Strategy

Built-in Protection

The screening script includes rate limiting:

• 0.3 second delay between requests (~3 requests/second)
• Automatic retry on 429 (rate limit exceeded) with 60-second backoff
• Timeout: 30 seconds per request
• Graceful degradation: Stops analysis if rate limit reached, returns partial results

Managing Request Budget

For free tier (250 requests/day):

Requests per stock analyzed:

• Stock Screener: 1 request (returns 100-1000 stocks)
• Dividend History: 1 request per symbol
• Historical Prices (RSI): 1 request per symbol
• Income Statement: 1 request per symbol
• Balance Sheet: 1 request per symbol
• Cash Flow: 1 request per symbol
• Key Metrics: 1 request per symbol

Total: 6 requests per symbol + 1 screener request

Budget allocation:

• Initial screener: 1 request
• Detailed analysis: 6 × N stocks = 6N requests
• Maximum stocks per run: (250 - 1) / 6 = ~41 stocks

FMP-Only Mode Optimization: Use --max-candidates parameter to limit analysis:

python3 screen_dividend_growth_rsi.py --max-candidates 40

Two-Stage Mode (RECOMMENDED):

When using FINVIZ pre-screening (--use-finviz):

Request breakdown:

• FINVIZ pre-screen: 1 FINVIZ API call → 10-50 symbols
• FMP quote fetching: 1 request per symbol (to get current price)
• FMP detailed analysis: 6 requests per symbol (dividend, prices, income, balance, cashflow, metrics)

Total FMP requests:

• 10 symbols: 10 + (6 × 10) = 70 requests
• 30 symbols: 30 + (6 × 30) = 210 requests
• 50 symbols: 50 + (6 × 50) = 350 requests (exceeds free tier)

Advantage: FINVIZ's RSI filter typically returns 10-30 stocks (not 1000), making FMP analysis feasible within free tier limits.

Cost comparison:

• FMP Starter Plan ($14/month, 500 requests): ~80 stocks/day
• FINVIZ Elite + FMP Free ($40/month, 250 FMP requests): ~30 stocks/day with RSI pre-filtering
• Result: FINVIZ approach provides higher-quality candidates (RSI pre-filtered) despite lower volume

Best Practices

1. Use FINVIZ two-stage for regular screening: Maximizes candidate quality, stays within FMP free tier
2. Run FMP-only for one-time analysis: Good for testing or when FINVIZ subscription not available
3. Run during off-peak hours: Lower chance of rate limits
4. Space out runs: Once daily or weekly, not multiple times per hour
5. Cache results: Save JSON output and analyze locally
6. Upgrade if needed: If screening >30 stocks daily, consider FMP Starter ($14/month)

Error Handling

Common Errors

1. Invalid API Key

{
  "Error Message": "Invalid API KEY. Please retry or visit our documentation."
}

Solution: Check API key, verify it's active in FMP dashboard

2. Rate Limit Exceeded (429)

{
  "Error Message": "You have exceeded the rate limit. Please wait."
}

Solution: Script automatically retries once after 60 seconds. If persistent, wait 24 hours for rate limit reset.

3. Symbol Not Found

{
  "Error Message": "Invalid ticker symbol"
}

Solution: Script skips symbol and continues (expected for delisted/invalid tickers)

4. Insufficient Price Data for RSI

• Empty array or < 20 days of price data Solution: Script skips symbol (common for newly listed stocks, low-volume stocks, or data gaps)

5. Insufficient Dividend Data

• Empty dividend history or < 4 years of data Solution: Script skips symbol (requires 4+ years to calculate 3-year CAGR)

Debugging

Check request count:

# Count API calls in script output
python3 scripts/screen_dividend_growth_rsi.py 2>&1 | grep "Analyzing" | wc -l

Monitor rate limit status: Script outputs warning when approaching limit:

⚠️  API rate limit reached after analyzing 41 stocks.
Returning results collected so far: 3 qualified stocks

Verbose debugging (if needed): Add at top of script:

import logging
logging.basicConfig(level=logging.DEBUG)

Data Quality Considerations

Data Freshness

• Annual statements: Updated after fiscal year end (delays possible)
• Real-time prices: Updated during market hours
• Dividend history: Updated after declaration/payment
• RSI calculation: Based on most recent 30 days of price data

Data Gaps

Some stocks may have:

• Incomplete price history: < 30 days (newly public, suspended trading)
• Missing dividends: Not all dividend-paying stocks report via API
• Inconsistent metrics: Different accounting standards, restatements
• Adjusted prices: Stock splits, dividends affect historical prices

Script behavior: Skips stocks with insufficient data (requires 4+ years dividends, 30+ days prices)

Data Accuracy

FMP sources data from:

• SEC EDGAR filings (US companies)
• Exchange data feeds (real-time prices)
• Company investor relations

Note: Always verify critical investment decisions with:

• Company filings (10-K, 10-Q) at sec.gov
• Company investor relations websites
• Multiple data sources for confirmation

RSI Calculation Accuracy

Inputs:

• 30 days of closing prices (ensures 14-period RSI + buffer)
• Standard 14-period RSI formula

Potential issues:

• Data gaps: Weekends, holidays, halted trading (script uses available data)
• Split adjustments: FMP provides adjusted prices (correct behavior)
• Intraday volatility: RSI uses only closing prices (not intraday highs/lows)

Validation: Compare calculated RSI with other sources (TradingView, Yahoo Finance) to verify accuracy.

Screening Workflow

Request Sequence

1. Stock Screener (1 request)
   ↓ Returns 100-1000 candidates

For each candidate (up to max-candidates limit):

2. Dividend History (1 request)
   ↓ Calculate dividend yield and CAGR
   ↓ If yield < 1.5% or CAGR < 12% → Skip

3. Historical Prices (1 request)
   ↓ Calculate RSI
   ↓ If RSI > 40 → Skip

4. Income Statement (1 request)
   ↓ Calculate revenue/EPS growth
   ↓ If negative growth → Skip

5. Balance Sheet (1 request)
   ↓ Check financial health
   ↓ If unhealthy ratios → Skip

6. Cash Flow (1 request)
   ↓ Calculate FCF payout ratio

7. Key Metrics (1 request)
   ↓ Extract ROE, margins, valuation

8. Composite Scoring
   ↓ Rank qualified stocks

9. Output JSON + Markdown Report

Optimization: Early exits reduce API calls. If a stock fails dividend or RSI checks (steps 2-3), remaining 4 API calls are skipped.

API Documentation

Official Docs: https://financialmodelingprep.com/developer/docs

Key Sections:

• Stock Fundamentals API
• Stock Screener API
• Historical Dividend API
• Historical Price API
• Ratios API

Support: support@financialmodelingprep.com

Alternative Data Sources

If FMP limits are restrictive:

1. Alpha Vantage: Free tier, 500 requests/day, includes RSI endpoint
2. Yahoo Finance (yfinance): Free, unlimited, Python library available
3. Quandl/Nasdaq Data Link: Free tier available for fundamental data
4. IEX Cloud: Free tier, 50k messages/month

Implementation Notes:

• Alternative sources require script modifications for different data formats
• Yahoo Finance (yfinance) provides built-in RSI calculation
• Alpha Vantage has dedicated technical indicators endpoint

yfinance Example (Free Alternative):

import yfinance as yf

# Get stock data
ticker = yf.Ticker("AAPL")

# Dividend history
dividends = ticker.dividends

# Historical prices
prices = ticker.history(period="1mo")

# Calculate RSI manually or use TA-Lib

Trade-offs:

• FMP: Structured, comprehensive, rate-limited
• yfinance: Free, unlimited, less reliable (Yahoo API changes frequently)
• Alpha Vantage: Good for technical indicators, limited fundamental data

Troubleshooting

"No Results Found"

Possible Causes:

1. All stocks failed RSI check (market not oversold)
2. High dividend growth (12%+) is rare
3. API rate limit reached before finding qualified stocks

Solutions:

• Relax RSI: --rsi-max 45
• Lower dividend growth: --min-div-growth 10.0
• Reduce candidates: --max-candidates 30 to avoid rate limit
• Check during market corrections (more oversold opportunities)

"Rate Limit Reached Quickly"

Causes:

• Used API for other purposes earlier today
• Script analyzed many stocks before finding qualified ones
• API limit reset not yet occurred (resets at UTC midnight)

Solutions:

• Wait 24 hours for rate limit reset
• Use --max-candidates 30 to conserve requests
• Upgrade to paid tier ($14/month for 500 requests/day)
• Check current usage in FMP dashboard

"RSI Calculation Errors"

Causes:

• Stock has < 20 days of trading history
• Data gaps (suspended trading, newly listed)
• API returned incomplete price data

Solution: Script automatically skips these stocks with warning:

⚠️  Insufficient price data for RSI calculation

No action needed - this is expected behavior.

Last Updated: November 2025 Script Version: 1.0 FMP API Version: v3