TauricResearch/TradingAgents · warning · NoMarketDataError

Yahoo Finance returned no rows

Error message

Yahoo Finance returned no rows

What it means

Raised in tradingagents/dataflows/stockstats_utils.py after a yfinance download for the symbol: the returned frame is empty or lacks a 'Close' column, so nothing is cached and a NoMarketDataError (symbol, canonical, 'Yahoo Finance returned no rows') is raised. The router turns it into the NO_DATA_AVAILABLE sentinel; empty frames are deliberately never persisted to the cache.

Source

Thrown at tradingagents/dataflows/stockstats_utils.py:206

            not cached.empty
            and "Close" in cached.columns
            and not _needs_same_day_refresh(data_file, curr_date_dt, today_date)
        ):
            data = cached

    if data is None:
        downloaded = yf_retry(lambda: yf.download(
            canonical,
            start=start_str,
            end=end_str,
            multi_level_index=False,
            progress=False,
            auto_adjust=True,
        ))
        downloaded = _ensure_date_column(downloaded.reset_index())
        # Only cache real data — never persist an empty frame.
        if downloaded.empty or "Close" not in downloaded.columns:
            raise NoMarketDataError(
                symbol, canonical, "Yahoo Finance returned no rows"
            )
        downloaded.to_csv(data_file, index=False, encoding="utf-8")
        data = downloaded

    data = _clean_dataframe(data)

    # Filter to curr_date to prevent look-ahead bias in backtesting
    data = data[data["Date"] <= curr_date_dt]

    # Reject a stale frame (latest row far older than curr_date) rather than
    # feeding year-old prices into indicators (#1021).
    _assert_ohlcv_not_stale(data, curr_date, symbol, canonical)

    return data


def filter_financials_by_date(data: pd.DataFrame, curr_date: str) -> pd.DataFrame:

View on GitHub (pinned to a33fd4c0f1)

Solutions

1. Verify the ticker exists on Yahoo: yf.Ticker(sym).history(period='5d') — if that is empty too, the symbol/coverage is the problem
2. Widen the requested date range so it actually spans trading days
3. Add a second vendor in config (data_vendors="yfinance,alpha_vantage") so uncovered symbols fall through to another source
4. Catch NoMarketDataError (or check for the NO_DATA_AVAILABLE sentinel string) and report the symbol as unavailable rather than estimating

Example fix

# before
get_stock_data_indicators_window_sma("ZZZZZZ", "2025-06-10", 10, 10)
# -> NoMarketDataError: ... Yahoo Finance returned no rows

# after
from tradingagents.dataflows.errors import NoMarketDataError
try:
    out = get_stock_data_indicators_window_sma(symbol, "2025-06-10", 10, 10)
except NoMarketDataError:
    out = f"No data for {symbol}; skip analysis"   # never fabricate prices

import yfinance as yf

def yahoo_covers(symbol: str) -> bool:
    try:
        return not yf.Ticker(symbol).history(period="5d").empty
    except Exception:
        return False

from tradingagents.dataflows.errors import NoMarketDataError

try:
    out = get_stock_data_indicators_window_sma(sym, curr_date, 10, 10)
except NoMarketDataError as e:
    out = f"NO_DATA: {e.symbol} — report unavailable, do not estimate"

When it happens

Trigger: yf.download for an invalid/delisted ticker; symbols Yahoo does not cover (some exchanges/OTC); an empty date window (start >= end); rate-limit/login walls on Yahoo returning zero rows; canonical symbol resolution producing a Yahoo-unrecognized form.

Common situations: LLM-generated or user-typo'd tickers; delisted stocks in historical runs; region-restricted symbols on Yahoo; too-narrow date ranges that exclude all trading days; Yahoo's periodic unauthenticated-access throttling.

Related errors

• No OHLCV data available for {symbol}.
• No OHLCV rows on or before {curr_date} for {symbol}.
• latest row is {latest.date()}, {stale_days} days before the
• ticker must be a non-empty string, got {value!r}
• ticker exceeds {max_len} chars: {value!r}

AI-assisted analysis of TauricResearch/TradingAgents@a33fd4c0f1 (2026-08-14). Data as JSON: /api/errors/48bd0786851cfe89. Report an issue: GitHub.