How to export your broker trade history

The short way

Every broker has this, under a menu called Statements, Activity, History or Tax Documents: download a CSV of your fills for a date range, and keep it. That file is the primary record. Every tool below either reads it or reads the same thing over an API, and none of them can reconstruct what the broker did not give you.

Do it before you need it. Brokers age out their download windows, and a closed account frequently stops offering the export at all.

Then pick what reads it. TradesViz is free forever with no card at 3,000 executions a month, stocks only; Tradervue has a free Basic plan capped at 30 imported trades a calendar month, which is not advertised on its pricing page — you pick it at signup instead of the trial.

What the options are

A journal that syncs. Tradervue imports fills from 81 brokers and platforms and is the oldest product in the category. TraderSync syncs directly from 76 brokers and takes CSV uploads for the other 415 on its list, and classifies options spreads on import rather than leaving you four unrelated legs. TradesViz treats stocks, stock options, futures, futures options, forex, crypto and CFDs as first-class instruments once you are off the free plan.

An aggregation API. SnapTrade returns account activities — buys, sells, dividends, fees, transfers, splits, option events — in reverse chronological order by trade date, paginated at a maximum of 1,000 transactions per request, with optional start and end dates. Plaid Investments covers over 3,200 US and Canadian institutions and returns tax lots on holdings, with cost basis, purchase price and purchase date per lot.

Tax software. TradeLog is a different purchase from all of the above: it rebuilds a US trader's tax year from raw broker history and prints Form 8949 detail for Schedule D, Form 6781 for Section 1256 contracts, and Section 475 mark-to-market on its higher tiers. Eight of its 35 supported brokers import through an in-app API connection that runs on SnapTrade; everything else is a file you download yourself.

A tracker rather than a journal. If the question is what you hold and what the return was rather than how you traded, portfolio trackers is the neighbouring listing and Sharesight is the one that produces real tax reports for several jurisdictions.

Where this breaks

The export is not a standard. Every broker's CSV has its own columns, its own date format, its own sign convention and its own idea of what a row is, and any of them can change without notice. This is why an importer's supported-broker list is a real feature rather than marketing: somebody maintains a parser per broker, and a broker not on the list means mapping columns by hand every time.

Partial fills multiply. One order can arrive as a dozen executions. Plan limits are counted in executions and your intuition is counted in trades, and the gap is usually a factor of two or more — TradeLog's own example is that 100 round trips typically come to 200 records or more. Read any "trades per month" cap as executions until the vendor says otherwise.

Multi-leg options arrive as legs. A four-leg spread that imports as four unrelated trades gives you four wrong P&L figures and no record of the position you actually had. This is the single thing to test with a sample file before paying for anything, and it is why TraderSync leads with spread detection.

The API lookback is short. Plaid's investments transactions endpoint is documented as up to 24 months. SnapTrade will go back to the earliest available transaction, but "available" means whatever the brokerage exposes, which is frequently less than your account's life. Anything older than the window lives only in the broker's own archive.

The 1099-B is not your trade history. It is a summary with the broker's own wash-sale adjustments already applied and no statement of which cost basis figures were changed, covering only positions closed in the year. Reconstructing a tax year from it is not possible, which is why the software built for that job starts from raw fills and reconciles gross proceeds against the form rather than importing it.

Two accounts do not reconcile themselves. Transfers in kind, corporate actions and anything moved between brokers arrive as an opening position with no history behind it. Someone has to supply the cost basis, and that someone is you, from the statement of the account it came from.

If you outgrow this

When you are filing rather than reviewing, the tool changes. A journal answers "was that setup any good"; tax software answers "what does Schedule D say", and the two are not substitutable. TradeLog is the US-specific answer here at $219 a year for up to 1,500 records.

When you want it in a program, move from CSV to brokerage account aggregation. One API, one auth flow, and history that refreshes instead of being re-downloaded — at the cost of a per-connection bill and a lookback window.

When several accounts have to agree, the problem is reconciliation rather than import, and a tracker built around multiple brokers and currencies is the right shape: Sharesight, Portseido and Kubera are the cards for that.

When the journal's own export matters, check it before you commit. Getting data into a product is always easier than getting it out again, and Tradervue's free Basic plan, for one, has no CSV export at all.

The rest of the category is trading journal software.