the free feed vs paid feed decision is bugging me more than expected

I went through the Alpaca connection guide last night, mostly out of curiosity, and honestly the mechanical part is nothing. Make the key, paste two fields, done before my tea cooled. What I keep sitting with is the bit it tells you to read before you trade at all, the part about which market data feed you pick. The guide is pretty direct that this is the choice that actually decides whether the whole thing works for you, not the setup. And I believe it, but I'm not sure I understand it in a practical way for how I trade. My own style is slow. I hold for days, I don't scalp, I've spent years teaching myself to stop reacting to every tick. So my first instinct was that the free feed is obviously fine — I'm not fighting anyone for a fill. But then I thought about it more and I got uneasy for a different reason. If the prices I'm seeing are only part of the picture, then my entry alerts and my stop levels are being measured against something slightly different from what actually happened. Not a big deal on any one trade maybe. But I keep a journal, and a journal built on partial data is a journal that will quietly teach me the wrong lessons about myself. So the question I can't answer from the guide alone: for a slow, low-frequency approach, is the concern about the cheaper feed mainly about fills and slippage, or is it also about the honesty of your own backtest and review? Those feel like different problems to me. One costs you a bit at the point of execution. The other corrupts the thing you use to decide whether you're improving. Has anyone here run the same strategy on both and noticed the difference in their records, rather than in their P&L? That's the bit I'd genuinely like to hear about.
ZenTrader_Ana
12 posts · 0 followers
+ Follow

Related reading

4 replies

L
LurkerLee· Aug 2026 ago
Does the app store which feed a candle came from, or do you just have to remember what you were on at the time?
F
FiveMinFiona· Aug 2026 ago
Your instinct about the journal is the right one and I'd separate it exactly the way you did. For execution, on daily holds you'll likely never feel it. For review, the risk is that your recorded high and low of a bar are narrower than reality, so a stop you think never got touched maybe did. If you're logging "stop held" as evidence your levels are good, that's the lesson getting quietly bent. What I'd do is pick one and stay on it for a full quarter rather than switching mid-way, because comparing months recorded on two different feeds is worse than either feed on its own.
CryptoKarl· Aug 2026 ago
coming from perps this whole conversation is funny to me because over here everyone just assumes the data is a bit cursed and sizes accordingly lol. but yeah for a swing book i'd care about the review side way more than the fill side, you're not getting picked off holding for days.
H
HalfKelly· Aug 2026 ago
Nice framing, Ana. Think of it as two separate error terms. Execution error hits every trade by a small amount and mostly shows up as a shaved edge — it's a drag, it's roughly constant, and you can estimate it. Measurement error is nastier because it isn't symmetric: it tends to make your recorded worst-case excursions look tamer than they were, so any stop distance or size you derive from that history is fitted to a gentler world than the one you trade in. If your sizing comes off observed drawdown per trade, that's the one I'd guard.
Sign in to reply →
← All brokers & apis