Flows & data

How to track Bitcoin ETF flows, daily

The primary sources, the free trackers worth using, a five-minute daily routine, and how to calculate flows yourself from shares outstanding and NAV.

Flow data is free, published daily, and one of the few genuinely useful public datasets in this market. It is also easy to consume badly — refreshing a dashboard hourly for a number that only updates once a day, or reacting to a single session of noise.

This page covers where the data comes from, which free sources are worth your time, and a routine that takes five minutes and produces a better read than an hour of scrolling. What the numbers mean is covered separately in flows explained.

Where the data comes from

Every flow figure in existence is derived from two inputs: how many shares a fund has outstanding, and what its net asset value per share was that day. Both come from the issuer. Trackers collect, multiply and aggregate — they do not observe flows directly, because nobody does.

Primary sources: the issuer and EDGAR

For any single fund, the issuer's own page is authoritative. It publishes net assets, NAV per share, shares outstanding and basket composition with an explicit as-of date — as an example, IBIT reported net assets of $60,018,939,295 and a NAV of $43.7352 as of 2 September 2026, with a basket of 22.65 bitcoins valued at $1,749,211.32.

For verification over longer periods, SEC EDGAR holds each trust's periodic filings. Forms 10-Q and 10-K contain reviewed or audited statements of bitcoin held, shares outstanding and expenses; Form 8-K covers material events. It is slower than any tracker and it is the record that settles disagreements.

Free trackers worth using

Farside Investors publishes the daily table most commonly quoted in the press — net flow by fund in US dollar millions, with the category total. Its 27 August 2026 table showed a total net flow of $242.3 million with IBIT at +$277.6m, FBTC at −$83.6m and GBTC at −$27.2m. Clean, consistent, and enough for most purposes.

CoinGlass maintains flow history and holdings charts for both Bitcoin and Ether funds, useful when you want to see a run of weeks rather than a single day.

What none of them provide is attribution. No public dataset tells you whether an inflow came from a retirement contribution, an adviser rebalance or a hedge fund running a delta-neutral basis trade against CME futures. That limitation is inherent, not a gap in coverage.

A five-minute daily routine

1. Check after the US close, once

   NAV is struck daily, so data lands in the evening. There is nothing to see at lunchtime.

2. Read the category total first

   Then the two or three largest fund contributors. Ignore funds at zero — that simply means no baskets moved.

3. Compare to the week

   Week-to-date against the prior week filters most of the noise. A single $200 million day means little on a category holding $84 billion.

4. Net out migration

   GBTC outflows against Grayscale Mini Trust inflows are the same investors switching fee tiers. Subtract that before concluding anything about demand.

5. Cross-check the surprises

   Anything that would change your mind gets verified on the issuer page before you act on it.

Calculating flows yourself

Worth doing once, because it removes any mystery about what the number is.

1. Record shares outstanding for each fund from the issuer page, at the same time each day.
2. Take the day-on-day change in shares.
3. Multiply by that day's NAV per share.
4. Sum across funds for the category total.

Worked example: a fund's shares outstanding rise by 6,000,000 and NAV is $43.7352. That is 6,000,000 × $43.7352 ≈ $262.4 million of inflow. That is the entire calculation. Any difference between your figure and a tracker's is timing or price-source convention, not a different methodology.

Timing and lag, precisely

Mistakes to avoid

• Treating volume as flow. Investors trading with each other generates volume and no flow at all.
• Reading one day as a trend. The series is extremely noisy at daily frequency.
• Reading GBTC outflows as bearishness. A years-long fee migration, not a view. See GBTC.
• Confusing flows with assets. Different measurements — see AUM explained.
• Mixing sources. Comparing Monday from one tracker to Tuesday from another manufactures a trend out of methodology.
• Assuming flows predict price. The 2026 record argues strongly against it — see flows and price.

Tracking ether and altcoin fund flows

The same method transfers directly to the other categories, with two practical differences. Ether fund coverage is nearly as good as bitcoin's — both Farside and CoinGlass publish daily ETH tables, and the August 2026 series was detailed enough to show a nine-to-ten session streak adding $1.42 billion from 17 August, a single-day figure of $225.8 million on 27 August, and ETHA leading one week with $567 million against FETH's $96.5 million.

Coverage of XRP, Solana and Dogecoin funds is thinner and less consistent. Aggregators sometimes quote cumulative inflows rather than daily flows for these — Solana funds passing $1.16 billion in cumulative inflows by mid-2026, XRP funds reaching roughly $2.0 billion in combined assets across seven products — because the daily numbers on small funds are frequently zero and not worth tabulating. For those categories, go to the issuer page and track shares outstanding directly.

One extra source becomes useful at that end of the market. Institutional holders filing Form 13F disclose fund positions quarterly, with a lag of up to 45 days after quarter end. It is far too slow for anything tactical, and it is the only public window into who owns these products rather than how much money moved. Treat it as archaeology, not a signal.