Data Accuracy

Targeted answers when a specific number Alphio AI shows you doesn't match what you expect.

---

Stock data looks wrong. Most "wrong stock data" reports trace to one of a few sources:

• Quote staleness — quotes refresh on intervals; during fast moves the most recent on-screen quote may lag the broker's tape by seconds.
• Pre / post-market vs regular session — extended-hours prints can show very different prices from the consolidated tape's last regular print.
• Adjusted vs unadjusted — historical prices may be split / dividend adjusted; the "wrong" number may be the correct adjusted price.
• Wrong ticker disambiguation — same ticker on different venues (US ADR vs home listing) shows different prices.

Verify against the broker's app for the same session and ticker. If they still disagree, file a data report (see below).

Crypto price doesn't match my exchange. Crypto prices vary across venues:

• Different venue, different price — the same token on Binance, Coinbase, Uniswap, and the chain's native DEX can trade at slightly different prices because each is a separate book.
• Index vs single venue — Alphio may display a composite index price; your exchange shows that exchange's local price.
• Liquidity-driven divergence — on low-cap tokens the price can diverge meaningfully across venues.

For trading decisions, the price that matters is the venue you're actually executing on. Alphio's price is a reference — confirm at the Trade Modal what you'll actually pay.

Earnings prediction was way off. Earnings predictions are probabilistic, not deterministic. The published number is a point estimate of a distribution, and surprises (the actual print being far from consensus and from Alphio's estimate) are common across the industry.

• Calibrate expectations using the historical accuracy view on the AI Earnings Calendar — hit rate, mean absolute error, and surprise distribution are published per ticker.
• For volatile names (small caps, IPOs, post-pandemic earners), per-ticker error tends to be larger.
• Use the prediction as one input alongside consensus, guidance, and your own model — not as a verdict.

If a specific ticker is repeatedly far off across multiple quarters, report it via the data-issue flow so the team can audit the model for that name.

Whales tracker data is delayed. Whales Auto Tracker draws from upstream filings and on-chain indexers, each with its own lag:

• 13F (institutional) — filed quarterly with up to a 45-day delay after quarter-end. The data is fundamentally lagged; there is no real-time feed.
• Form 4 / congressional disclosures — filed within statutory windows (typically 2 days for Form 4, 30-45 days for congressional). Alphio surfaces them shortly after they hit the SEC / disclosure portal.
• On-chain wallets — near-real-time, limited by indexer cadence (typically minutes).

If the on-chain view is hours stale (not just minutes), report it via the data-issue flow. If the 13F view feels stale, that's the upstream filing cadence — not a delay on Alphio's side.

Where does Alphio's data come from?

warning

The exact data-provider identities (which vendor supplies stock quotes, fundamentals, on-chain data) are still being finalized for public documentation. Alphio aggregates from multiple upstream providers and picks the best available source per field.

In general: regulated market data for stocks comes from licensed vendors; on-chain data comes from a mix of node providers and indexers; alternative data (sentiment, news, filings) comes from specialized providers. Specific names will be published as data partnerships mature.

Reporting a data issue. When a data point looks wrong, give the data team the minimum needed to reproduce:

• Ticker / token / address of the affected asset.
• Field (price, market cap, holders count, earnings estimate, etc.).
• Value shown on Alphio at the time you observed.
• Value you expected and the source of that expected value (broker name, exchange URL, on-chain tx).
• Timestamp (your time zone is fine).
• Screenshot if convenient.

Submit via Settings → Feedback at /account/feedback. The data team triages provider mismatches on a weekly cadence.