What is Earnings Call?

In Plain English

Four times a year, the CEO and CFO of every major public company get on a phone call with Wall Street. For 45-90 minutes, they explain the quarter's results, discuss what they expect going forward, and answer questions from analysts at Goldman Sachs, Morgan Stanley, and hundreds of other institutions. These calls are recorded, transcribed, and filed publicly. They are among the richest sources of information about a company that exists anywhere.

The format is consistent: prepared remarks (15-30 minutes, scripted), followed by a question-and-answer session (30-60 minutes, unscripted). The prepared remarks are polished and carefully lawyer-reviewed, often revealing what management wants to emphasize. The Q&A is where things get interesting — analysts probe for specifics, push back on guidance, and occasionally force management to address uncomfortable topics. The unscripted responses are frequently more revealing than the prepared text.

What makes earnings calls so valuable for NLP is the volume, consistency, and density of meaningful information. You have the same executives, talking about the same company, once per quarter, for years. Changes in language patterns become detectable. A CEO who has said "we're confident in our pipeline" every quarter for two years who suddenly says "we're encouraged by early indicators" has said something subtly different — a shift that might not trigger a keyword alert but would be detected by comparing embedding vectors.

Financial analysts have known for decades that how management says things matters almost as much as what they say. The challenge was extracting that qualitative signal at scale. Transformer-based embeddings now make this practical across thousands of companies.

Technical Definition

An earnings call transcript has a defined structure that VectorFin's pipeline uses for intelligent chunking:

1. Operator introduction: boilerplate, minimal signal 2. Safe harbor / forward-looking statement disclaimer: legal preamble 3. CEO prepared remarks: strategic context, highlights, narrative 4. CFO prepared remarks: detailed financial results, guidance 5. Q&A session: analyst name, question, management response 6. Operator closing: boilerplate

VectorFin segments transcripts into 512-token chunks with 64-token overlap. Speaker identification is preserved in the chunk metadata — Q&A chunks include the speaker's name and role, enabling filtering for CFO-only commentary or specific analyst questions.

Fiscal period format used in VectorFin: {YYYY}-Q{N} where N is the company's fiscal quarter number (1-4). Note: a company with a June 30 fiscal year end will have its Q1 = July-September, which corresponds to calendar Q3. VectorFin stores both the fiscal period label and the calendar quarter for cross-company comparison.

How VectorFin Uses This

VectorFin embeds every earnings call transcript for 5,000+ tickers back to 2018, totaling approximately 200,000+ calls and 50M+ embedding chunks stored in:

gs://vectorfinancials-data/warehouse/embeddings/transcripts/

Each chunk row: ticker, fiscal_period, chunk_idx, section (prepared/qa), speaker, effective_ts (fiscal period end), knowledge_ts (ingestion time), embedding (float[768]).

GET https://api.vectorfinancials.com/v1/embeddings/{ticker}?period=2024-Q4
GET https://api.vectorfinancials.com/v1/embeddings/{ticker}?period=2024-Q4&section=qa

The Q&A section filter is particularly useful for detecting management evasion — when management consistently gives vague or deflecting answers to analyst questions about a specific topic.

Code Example

import requests

API_BASE = "https://api.vectorfinancials.com"
API_KEY = "vf_your_api_key_here"

# Analyze the Q&A section separately from prepared remarks
# Q&A contains more candid, unscripted management language

ticker = "TSLA"
period = "2024-Q3"

# Fetch all chunks
all_resp = requests.get(
    f"{API_BASE}/v1/embeddings/{ticker}",
    params={"period": period},
    headers={"X-API-Key": API_KEY},
)
all_chunks = all_resp.json()["chunks"]

# Separate prepared remarks from Q&A
prepared = [c for c in all_chunks if c.get("section") == "prepared"]
qa = [c for c in all_chunks if c.get("section") == "qa"]

print(f"{ticker} {period}: {len(prepared)} prepared chunks, {len(qa)} Q&A chunks")

# Search the Q&A for analyst questions about specific topics
qa_search = requests.post(
    f"{API_BASE}/v1/embeddings/search",
    json={
        "query": "production capacity constraints supply chain bottlenecks delivery challenges",
        "tickers": [ticker],
        "period_start": period,
        "period_end": period,
        "section": "qa",
        "top_k": 5,
    },
    headers={"X-API-Key": API_KEY},
)

print(f"\nTop Q&A passages about production challenges:")
for r in qa_search.json()["results"]:
    speaker = r.get("speaker", "Unknown")
    print(f"\n[{speaker}] (similarity: {r['score']:.3f})")
    print(f"{r['text'][:250]}")