Algorithmic Trading for Individual Investors in 2026: What Algo Means at Schwab, Alpaca, tastytrade, Tradier, Webull and Kraken, What the DIY Stack Costs, and Three Ways In Without a Quant Degree

Industry estimates put algorithmic trading at somewhere between 60% and 75% of US equity volume, and the SEC's own staff report on the subject calls it "virtually ubiquitous." Retail investors are a real share of the rest: Jefferies estimated retail at more than 20% of US volume in December 2025, and Goldman Sachs put it at 9% in October 2026, down from 15% a year earlier, as the easy summer faded. So the word algo is everywhere and most of the people searching it are not quant funds. They are individuals with a Schwab, Alpaca, tastytrade, Tradier, Webull or Kraken account asking a practical question: what does algorithmic trading mean at my size, what will my broker let me do, what does it cost to build, and is there a way in that does not involve learning Python first. This page answers those four in order.

What algorithmic trading means at retail scale

Strip the word down and it is a rule, written in advance, that software applies without you at the screen. The rule has three parts, and if any part is missing it is not a strategy yet. Entry: what has to be true for a buy, a moving-average cross, a breakout, a momentum reading. Size: how much, as a dollar amount, a share count or a share of the account, with a daily cap. Exit: the stop, the target, the trailing stop, or the time limit that closes it. High-frequency trading is a different animal, measured in microseconds and co-located servers; nothing on this page is about that. At retail scale the edge of automation is not speed, it is consistency: the rule fires the same way on the 200th day as on the first, and it fires while you are at work.

What each broker lets you automate on its own

• Schwab and thinkorswim: study-based conditional orders, where an order is placed or cancelled when a one-plot study condition is met, plus OCO brackets. One order per rule, no re-arming after a fill, and thinkScript strategies are hypothetical by Schwab's own documentation. Beyond that, the Schwab Trader API: whole shares only, no paper sandbox, 120 requests a minute, a 7-day re-login. The API explained and the thinkorswim post have the detail.
• Alpaca: API first. There is no point-and-click automation; you write against the API or connect a platform. A paper-trading environment is free, market data is free on the IEX feed with a 200-calls-a-minute limit, and the full-exchange feed with options data is $99 a month as Algo Trader Plus.
• tastytrade: no native automation. The Open API is the route, and tastytrade's terms make the account holder responsible for orders "placed by an automated or unattended system." Crypto through the API has been disabled since the June 29, 2026 release until further notice.
• Tradier: API first, with a $10-a-month Pro plan that includes API access and commission-free stock and options trading, and a 2026 integration that lets TradingView place options orders directly.
• Webull: one-triggers-OCO and conditional orders with up to six trigger conditions, in the mobile app and during regular hours only, plus an OpenAPI for US stocks, ETFs and options.
• Kraken: "clients are welcome to use trading bots", in Kraken's own words, through REST and WebSocket APIs with sample bots in the documentation. No bots are built into the exchange.

The pattern: every broker on the list will let a rule trade the account, but only two give you a rule-builder inside the app, and both of those are one-shot orders rather than a strategy that keeps running. For a running strategy, the choice is write it or connect something that already runs. Best brokers for automated trading compares the API side.

What the DIY stack costs, at November 2026 list prices

A do-it-yourself algorithm needs four things: data, a place for the signal to run, a way to test it before it is live, and a machine that stays on. Here is what each costs today, from each vendor's own pricing page, and then three ways to assemble them.

• Data: Alpaca's IEX feed is free; its Algo Trader Plus, with every US exchange, options data and unlimited websocket streams, is $99 a month. Massive, the service formerly called Polygon, lists a $29-a-month Starter and a $199 Advanced plan for stocks. Databento's Standard plan moved to $199 a month in June 2026.
• Signals and alerts: TradingView's paid plans are $14.95 (Essential, 20 alerts), $34.95 (Plus, 100 alerts) and $69.95 (Premium, 400 alerts) a month, billed monthly; webhook alerts, the thing that turns a chart signal into an order, are on the paid plans.
• Backtesting: QuantConnect is free to research on and about $84 a month for the plan that includes live-trading nodes. Alpaca's paper environment is free but it is forward-testing, not backtesting.
• Hosting: a small virtual machine is $4 to $6 a month at DigitalOcean, Vultr or Hetzner. Your laptop is free and goes to sleep.

Three builds

• The $0 build: Alpaca's free IEX data, a Python script, Alpaca paper trading, your own laptop. Costs nothing and runs only while the laptop is awake and the script has not crashed.
• The $40 build: TradingView Plus for the signal and the webhook, a $5 machine to receive the webhook and send the order, your broker's free API. The strategy lives in TradingView's alert conditions, which is where its limits are too.
• The $387 build: Algo Trader Plus data, a Databento or Massive Advanced feed for history, QuantConnect for backtests and live nodes, and a machine. This is the serious hobbyist's rig, before any of your time is priced.

None of those numbers include the part that actually costs the most: the hours to write the scheduler, the retry logic, the order reconciliation and the exit path, and the hours to maintain them when a broker changes a field. What automated trading actually costs prices the commissions and the subscriptions on the other side of the ledger.

The rule that changed in June 2026

If the last thing you read about day trading mentioned the $25,000 minimum, it is out of date. FINRA's Regulatory Notice 26-10 replaced the day-trading margin provisions of Rule 4210 "in their entirety, including the day trade count requirements for designating a customer as a 'pattern day trader' and the $25,000 pattern day trader minimum equity requirement." The SEC approved the change on April 14, 2026; it took effect June 4, 2026, with brokers allowed to phase it in until October 20, 2027. What replaces it is an intraday margin standard: on a day when your trading reduces the margin on hand, the broker calculates an intraday deficit and you have to cover it, with a 90-day restriction if you do not. Two things did not change. Cash accounts still settle on T+1 and still produce good-faith violations when you sell something bought with unsettled money; and because of the phase-in, your broker may still be applying the old day-trade count today. Ask before you build a strategy that assumes either regime.

The red flags regulators tell you to watch for

The word algorithm sells, which is why FINRA published an investor insight on auto-trading services on July 29, 2025. Its warning signs: services pitched as "beginner-friendly and/or risk-free" that "provide consistent monthly returns of more than 10 percent"; "vague and unverified claims of impressive past market returns"; claimed partnerships with broker-dealers, sometimes with logos used without permission; vendors that "overstate their AI capabilities," which FINRA calls AI washing; and requests for your credentials, "such as the username and password to your brokerage account." FINRA's advice is to check any firm on BrokerCheck and to verify a claimed partnership by contacting the broker directly. The SEC's 2026 enforcement actions against two fake-bot operations, one for $12.3 million and one for $15 million, are the same story with victims. And FINRA's 2026 oversight report flags, for the industry, the risk of "AI agents acting autonomously without human validation and approval."

Turn those into a test. A legitimate automation never asks for your brokerage password: at Schwab you authorize through Schwab's own login, and at the API brokers you create a key with trade permission and no withdrawal permission, which the safe-connection guide walks through. It never promises a return, because none exists to promise. And it tells you what the rule is, in plain words, before you turn it on.

Three ways in without a quant degree

• Conditional orders at your broker. Free, immediate, and enough for a one-shot rule: buy on a cross, exit at a target or a stop. Schwab's thinkorswim and Webull have them. The limit is that the order does not re-arm, so a repeating strategy means re-entering it.
• Build it. One of the three stacks above, plus Python or a TradingView alert script. You own everything, including the uptime. How to backtest before risking real money is the first thing to read, because the backtest is where most DIY strategies should have stopped.
• Connect a hosted rules-based platform. You set the rule's three parts, entry, size and exit, as settings, and the platform's servers run them against your own brokerage account over the broker's API. No data subscription, no machine, no code. The no-code guide compares the approaches and the beginner's guide is the first-week walkthrough. If a Claude or ChatGPT plan talking to your broker is the version you had in mind, the MCP post prices that route and lists what it leaves out.

Where JorgAI fits

JorgAI is the third way. It connects to Schwab, Alpaca, Tradier, tastytrade or Webull for stocks and to Kraken or the Crypto.com Exchange for crypto, through each one's own API, with the account staying where it is. You set the size and the exits once: daily spend, position size, auto-trades per day, stop and trailing stop. The system scans on its own servers, applies the rules, and places the order at your broker; at Schwab that took a median of 1.0 seconds from signal to fill across 294 measured orders. The free plan shows the same system running on a simulated account before you connect anything, which is the no-cost way to see what a running rule looks like.

Quick answers

Is algorithmic trading legal for individual investors?

Yes. Every broker on this page allows API or conditional-order automation of your own account. What is regulated is the vendor: FINRA tells you to check any firm you entrust with your investments on BrokerCheck and to walk away from guaranteed returns or a request for your brokerage password.

Do you need to know how to code for algorithmic trading?

Not for the first and third ways in. Conditional orders at your broker and a hosted rules-based platform both take settings, not code. Building your own stack does require Python or a scripting language and the time to maintain it.

How much does algorithmic trading cost for an individual?

The DIY stack runs from $0 (free data, your laptop) to about $40 a month (TradingView Plus and a small server) to about $387 a month (full data, history, backtesting and live nodes), before your time. Hosted platforms charge a subscription instead; JorgAI's start at $14.99 a month.

Is the pattern day trader rule gone?

FINRA eliminated the pattern-day-trader designation and the $25,000 minimum effective June 4, 2026, replacing them with intraday margin standards, but brokers may phase the change in until October 20, 2027. Ask your broker which regime it is applying today.

Which broker is best for algorithmic trading?

It depends on what you want to automate and how. Schwab and Webull have conditional orders in the app; Alpaca, tastytrade, Tradier and Kraken are API-first with no in-app rule builder. The broker comparison lays out fees, data and limits side by side.

Volume estimates, prices and rule text are from the sources named in the text as of November 4, 2026; prices change, so check the vendor's page before you buy. The three builds are arithmetic on list prices, not a recommendation. The JorgAI order measurement describes how the connection behaves, not results. Education, not financial advice.

If you want the rule running without the stack, start a free trial: connect the broker you already use and set the entry, the size and the exit once.