yfinance Ticker Info Not Working: Fix Empty Dicts and Names

Why is yfinance ticker info not working for AFT.NZ but fine for MSFT?

yfinance talks to two different Yahoo backends. history() and fast_info use the chart endpoint (v8/finance/chart), which needs only a symbol and returns prices plus a small metadata block. Ticker.info uses quoteSummary (v10/finance/quoteSummary) with modules such as assetProfile, summaryDetail, price, defaultKeyStatistics and financialData.

quoteSummary is populated from Yahoo's fundamentals and profile database. Large US names have full coverage. Smaller or non-US listings (NZX, ASX small caps, some Asian and European lines) often have a quote and chart but little or no profile data. The response then holds a handful of keys, or nothing.

quoteSummary also requires a cookie plus a crumb token, while chart is more lenient. When Yahoo changes the cookie or crumb flow, or throttles your IP, .info breaks first. That is why the same ticker can show a price chart in your code and an empty dict from .info.

What do the different symptoms mean?

The same call gives different failures for different reasons. Treating them all as 'upgrade yfinance' wastes time, so map the symptom to the cause first.

How do I check whether it is my setup or Yahoo's data?

Run these checks in order. They take a few minutes and separate the three failure classes: old library, blocked session and missing data.

1. Upgrade: pip install -U yfinance. Yahoo changes its web flow often, and old releases break on crumb handling.
2. Confirm the version in the interpreter you actually run: import yfinance; print(yfinance.__version__). Notebooks and virtualenvs often differ from the shell.
3. Restart the process, so a stale cookie is discarded, and test MSFT and your ticker back to back in the same session.
4. Open the quote page for the symbol on the Yahoo site and compare it with the Profile and Statistics tabs. If they are empty there, the data is absent at source.
5. Run the control test below. If MSFT also fails, you are blocked or outdated. If only AFT.NZ fails, the data is missing.

import yfinance as yf

for sym in ["MSFT", "AFT.NZ"]:
    t = yf.Ticker(sym)
    try:
        info = t.info
        print(sym, "info keys:", len(info), "longName:", info.get("longName"))
    except Exception as e:
        print(sym, "info error:", type(e).__name__, e)
    print(sym, "fast currency:", t.fast_info["currency"])
    print(sym, "history rows:", len(t.history(period="5d")))

How do I get company name and currency without relying on .info?

Two chart-backed sources avoid quoteSummary. t.fast_info is a lightweight mapping with keys such as currency, exchange and last_price. t.history_metadata returns the chart metadata dict, which typically includes currency, exchangeName, instrumentType, symbol, and often longName and shortName.

history_metadata is populated by a history call, so call t.history(period="1d") first if you get an empty result. Whether longName is present depends on the symbol, so treat every key as optional.

import yfinance as yf

t = yf.Ticker("AFT.NZ")
print(t.fast_info["currency"], t.fast_info["exchange"], t.fast_info["last_price"])

t.history(period="1d")          # ensures metadata is loaded
meta = t.history_metadata or {}
print(meta.get("currency"), meta.get("exchangeName"),
      meta.get("longName"), meta.get("shortName"))

What is a reliable get_name_currency function with a fallback chain?

The layered order is info, then fast_info, then history_metadata, then Search. Each layer only fills fields still missing, every access uses .get(), and each exception is recorded rather than swallowed. A 'source' field shows where each value came from, which matters when you audit a stock list later.

yf.Search(symbol).quotes returns a list of dicts with fields such as symbol, shortname, longname, exchange and exchDisp. It is a separate search endpoint, so it can name a ticker that .info cannot. Match on the exact symbol, not the first result.

import yfinance as yf

def get_name_currency(symbol, use_search=True):
    out = {"symbol": symbol, "name": None, "currency": None,
           "exchange": None, "source": {}, "errors": []}
    t = yf.Ticker(symbol)

    def fill(key, value, src):
        if value and out[key] is None:
            out[key] = value
            out["source"][key] = src

    try:
        info = t.info or {}
        fill("name", info.get("longName") or info.get("shortName"), "info")
        fill("currency", info.get("currency"), "info")
        fill("exchange", info.get("exchange"), "info")
    except Exception as e:
        out["errors"].append(f"info:{type(e).__name__}")

    try:
        fi = t.fast_info
        fill("currency", fi.get("currency"), "fast_info")
        fill("exchange", fi.get("exchange"), "fast_info")
    except Exception as e:
        out["errors"].append(f"fast_info:{type(e).__name__}")

    try:
        t.history(period="1d")
        m = t.history_metadata or {}
        fill("name", m.get("longName") or m.get("shortName"), "history_metadata")
        fill("currency", m.get("currency"), "history_metadata")
        fill("exchange", m.get("exchangeName"), "history_metadata")
    except Exception as e:
        out["errors"].append(f"history:{type(e).__name__}")

    if use_search and out["name"] is None:
        try:
            for q in yf.Search(symbol).quotes:
                if q.get("symbol", "").upper() == symbol.upper():
                    fill("name", q.get("longname") or q.get("shortname"), "search")
                    fill("exchange", q.get("exchDisp") or q.get("exchange"), "search")
                    break
        except Exception as e:
            out["errors"].append(f"search:{type(e).__name__}")
    return out

print(get_name_currency("AFT.NZ"))
# Illustrative shape of a result (values depend on what Yahoo returns that day):
# {'symbol': 'AFT.NZ', 'name': 'AFT Pharmaceuticals Limited', 'currency': 'NZD',
#  'exchange': 'NZE', 'source': {'name': 'history_metadata', 'currency': 'fast_info',
#  'exchange': 'fast_info'}, 'errors': []}

How do I run this for a whole index like the NZX50 without getting blocked?

Yahoo's API is unofficial and its terms restrict automated bulk use, so keep volume low. Every symbol costs several requests in the function above, which is why 429 errors appear in the middle of a 50-symbol loop.

Cache first, throttle second, retry third. Metadata such as name and currency changes rarely, so fetch each symbol once and reuse the file. Retry only transient errors (429, timeouts) with exponential backoff and jitter, and never retry a 404, which is a permanent answer.

import json, os, random, time

CACHE = "names_currency.json"
cache = json.load(open(CACHE)) if os.path.exists(CACHE) else {}

def with_backoff(sym, tries=4):
    for i in range(tries):
        res = get_name_currency(sym)
        limited = any("RateLimit" in e or "429" in e or "TooMany" in e
                      for e in res["errors"])
        if not limited:
            return res
        time.sleep(2 ** i * 5 + random.random())
    return res

symbols = ["AFT.NZ", "AIA.NZ", "FPH.NZ", "MFT.NZ"]   # build from your NZX50 list
for s in symbols:
    if s in cache and cache[s]["name"] and cache[s]["currency"]:
        continue
    cache[s] = with_backoff(s)
    json.dump(cache, open(CACHE, "w"), indent=1)
    time.sleep(1.5 + random.random())

How do I tell missing data from a block, and what if Yahoo lacks it?

Missing data is stable: the same ticker returns the same sparse result across retries, while the MSFT control works in the same session. A block is intermittent or global: MSFT fails too, the error is 429 or Invalid Crumb, and it clears after waiting or restarting. Newer yfinance versions raise YFRateLimitError, which you can catch by name via getattr(yfinance.exceptions, 'YFRateLimitError', ()).

Be careful with a hardcoded suffix map, since it is a last resort and not a source of truth. Cross-listings, dual-currency lines and quote-unit quirks break it. Use it only to fill a currency flagged as 'assumed'.

If Yahoo cannot supply the fields, use a better source for the identity data: the exchange's own listings file (NZX publishes its instrument list), OpenFIGI to map tickers to names and exchange codes, or a paid market-data provider with a licence that permits your use.