Trading journal with Alpaca integration in 2026
Alpaca's OAuth grant carries trading authority, as three of the other direct connections' grants do, rather than being broker-enforced read-only. What that means, and what the sync reads.
Published by TradeFlow Quantum. Filed under alpaca, broker, how-to.
Alpaca built its business on the idea that a brokerage should be an API first and a screen second. For anyone connecting a trading journal, that shows up immediately: the OAuth flow is a plain redirect, the token comes back without ceremony, and the first sync works. Of the ten brokers TradeFlow Quantum connects to directly, Alpaca is the one that gives us the least trouble.
It is also the one with the scope you should read most carefully before you click approve, and those are not separate observations. The same developer-first design that makes the connection painless is what puts trading authority in the token.
account:write trading data, which includes trading authority. TradeFlow Quantum contains no order-placement code, but that limit is ours, not Alpaca's — and that is a weaker guarantee than the one Schwab gives.Read-only as intent versus read-only as capability
Every journal in this category says "read-only". The phrase hides a distinction that matters more than any feature comparison, and Alpaca is the clearest place to show it.
- Read-only enforced by the broker. Schwab's
AccountAccessgrant will not accept an order no matter what a connected tool sends. A bug in the tool cannot become a trade in your account, because the rejection happens at the broker. - Read-only as a property of the tool. Alpaca's OAuth scopes are coarser. The grant that lets us read your fill history is the same grant that would let us place an order. Nothing on Alpaca's side stops it; the only thing stopping it is that the code to do it does not exist.
Alpaca is the second kind. We state it plainly on the Alpaca page in our supported-brokers list rather than letting "read-only" do quiet work, because a claim about our own restraint is not the same as a claim about what is possible, and you are entitled to know which one you are being given.
What narrows it in practice is how little the integration asks for. TradeFlow Quantum makes exactly one call against your Alpaca account: /v2/account/activities, filtered to fills and the three option events: expiration, assignment and exercise. It does not read your positions, your balances, or your open orders — not because it cannot, but because nothing in the journal needs them. Fills are the whole input. Everything on the analytics side is computed from those and from what you type.
Setting it up
Four steps, no CSV, and the slowest part is Alpaca's login:
- System → Brokers → Connect Alpaca.
- You are redirected to Alpaca to authorize TradeFlow Quantum. Your password goes to Alpaca; we never see it.
- Approve the scope on Alpaca's authorization page — read the permission list here, because it includes trading.
- You land back on the brokers page, and clicking Sync runs the first pull, a year of fills.
There is no institutional disclosure screen the way there is with Schwab, and no seven-day refresh-token clock. This is the connection that tends to keep working without you thinking about it.
Sync runs nightly, and the button is there for sooner
Since November 8, 2026 the direct connections, Alpaca included, are pulled once a night at 09:20 UTC for paid accounts, trial included, with the token refreshed first, and every five minutes during US market hours while a prop evaluation is active. The Sync button on the brokers page runs the same pull on demand. Before that date this section said no scheduled job existed, and it was right then.
If you journal at the end of a session or on the weekend, the session is already there when you open the app. What the nightly pull cannot give you is the fill from an hour ago — press Sync for that, and it takes seconds.
What the pull gives you and where it stops
Alpaca returns fills: symbol, side, quantity, price, timestamp. The importer pairs those into round-trip trades, which is the step that makes a scaled-in position read as one trade rather than four rows. A year is the reach of the first pull, not a cap on what the journal keeps — each pull after it reads the last 90 days, and every fill read is kept.
Two limits to know before you connect:
- Equities and options only. Alpaca's crypto product sits behind a separate OAuth flow that TradeFlow Quantum does not support today. Crypto fills taken there will not appear, and CSV import is the workaround.
- One Alpaca environment per deployment. Paper and live are different Alpaca API hosts, and which one this app talks to is a server-level setting rather than a per-account choice you make at connect time. If you run both and expect to see both side by side in one journal, that is not how it currently works.
What you get once the trades are in
The integration is plumbing, and it is worth being clear about what it plumbs toward. Alpaca can give the journal what you did. It cannot give it why — the setup you thought you were taking, what you expected, what you were feeling at entry. That half is typed by you, and it is the half the analytics actually need.
With both present, the cohort view pivots your trades by setup, hour of day, day of week, side, psychology score, R-multiple bucket or symbol, and reports P&L, win rate, average R and trade count per cell, average R only over trades that carry one — a broker sync or CSV import supplies none. The compare view runs two filter sets head to head. The setup breakdown flags setups whose recent win rate has dropped 15 percentage points or more below their all-time figure — descriptive, not a verdict, but usually the first place a real problem surfaces.
The argument for connecting a broker at all is not the seconds saved per trade. It is that hand-entered history is a biased sample: the trades you feel like revisiting are the ones that get logged. A fill feed does not have an opinion about which trades were embarrassing.
Honest disqualifier
If the trade-capable scope is the part you cannot get comfortable with, do not talk yourself into it — use CSV import, which asks for nothing but a file and loses you none of the analytics. If you trade crypto on Alpaca, the connection will look broken to you because the half you care about is missing. And if you take a handful of trades a month, a direct sync is convenience rather than insight: the cohort views need a sample before they say anything, and connecting a broker does not manufacture one.
Related posts
Not financial advice. This post reflects the author’s opinion based on publicly-available information at the time of writing. Mention of third-party products is not an endorsement; product features and prices change over time. Past performance does not guarantee future results.