Best broker checklist for algo and automated F&O trading in India

Most “top 10 algo trading platforms” articles are ranked lists written for someone else’s trading style. They tell you Broker A is number one, Broker B is number three, and leave you to work out whether that ranking has anything to do with the way you actually trade Nifty weeklies or run a set-and-forget options selling strategy.

This piece takes a different route. Below is a 12-point scorecard you can tick off yourself, weighted for what genuinely decides whether a systematic F&O strategy survives contact with the market: regulatory standing, automation route, execution tooling, cost stack, margin behaviour, and reliability on expiry day. Score your shortlist, add up the points, and pick.

Investments in the securities market are subject to market risks. F&O trading carries a high level of risk.

Why a scorecard beats a “top 10” list when you trade F&O

A ranked list assumes every trader wants the same thing. An algo trading broker that is brilliant for a Python developer running 40 instruments through a REST API may be a poor fit for a working professional who wants two rules-based strategies running with a stop-loss and no code at all.

Three reasons a scorecard works better:

• Your weights are not my weights. If you sell options on expiry day with four legs, margin efficiency and order slicing matter far more than backtest length. If you run a slow monthly rebalance, brokerage per order barely registers.
• Rankings age badly. API pricing, order-per-second caps, and algo approval status change. A checklist you re-run in six months stays useful. A ranking published last year does not.
• Disqualifiers are binary, not scored. Some things are pass/fail. A broker without valid SEBI registration does not get “3 out of 10”. It gets removed.

The scorecard also forces you to look at what marketing pages skip: what happens when a leg fails, what the exit tooling looks like at 3:15 pm on expiry, and what the total charge per round trip actually is once statutory levies are added.

How to use this checklist

Run each shortlisted broker through all 12 checks. Use this structure:

• Must-pass (Checks 1, 2, 7, 12). Score 0 or 10. Anything scoring 0 is out, regardless of how well it does elsewhere.
• Heavily weighted (Checks 3, 4, 6, 8, 9). Score 0–10 and multiply by 2. These decide whether your strategy is practical and cost-effective at the margin.
• Nice-to-have (Checks 5, 10, 11). Score 0–10 at face value. Important, but rarely the deciding factor for a retail book.

Maximum score is 190. Below 120, you are compromising on something that will cost you later. Instant disqualifiers are listed near the end. There are five of them, and any one should end the evaluation on the spot.

Check 1: Regulatory status

The first check is the least glamorous and the most important. Confirm four things, in writing, from the broker’s own disclosures:

• SEBI stock broking registration number. Every legitimate Indian broker publishes it. Lemonn’s broking entity, NU Investors Technologies Private Limited, holds SEBI stock broking registration INZ000304837.
• Exchange membership. NSE and BSE membership means your orders route through the recognised exchange infrastructure and you get investor grievance recourse.
• Depository participant status. CDSL or NSDL. Your shares and pledged holdings sit here, not with the broker.
• Research Analyst registration, if the broker publishes any research, signals, or strategy commentary. Lemonn holds RA registration INH000016764.

That last one is worth dwelling on. If a platform is pushing trade ideas into your app without an RA registration behind it, you have found a compliance problem, not a feature.

Score 10 if all four are published and verifiable. Score 0 if any is missing.

Check 2: Exchange-approved algo offering

SEBI’s 2025 framework for retail algorithmic trading reshaped what brokers may offer. The core ideas retail traders need to understand:

• Algo strategies offered to clients must go through exchange approval, and approved algos carry a unique identifier so every automated order can be traced back to a registered strategy.
• Brokers are responsible for the algos they distribute, including those built by third-party providers and vendors plugged into their APIs.
• Orders above a specified per-second threshold from a retail account are treated as algorithmic and must be tagged accordingly.

The practical question for you: is the broker offering algo access under a framework that is registered and tagged, or through a grey-market arrangement with an unaffiliated tool? A broker that has done the work will say so plainly and will explain how strategy IDs and order tagging function on their platform.

Score 10 for a clearly disclosed, exchange-aligned algo offering. Score 0 for vagueness.

Check 3: Which automation route you actually get

There are three realistic paths for a retail trader in India, and they suit very different people.

Route A: No-code strategy builder. You select or configure rules-based strategies with built-in stop-losses and risk controls, and the platform handles execution logic. This is the route for the large majority of retail traders who have a view on the market but no interest in maintaining code. Lemonn’s SmartInvest sits here: algorithmic, rules-based, designed for investors without programming skills, with risk controls built into the strategies rather than bolted on afterwards.

Route B: API access. You write your own logic in Python, Node, or whatever you like, and place orders programmatically. Maximum flexibility, maximum responsibility. You own the uptime, the reconnection logic, the error handling, and the consequences of a bug at 9:20 am.

Route C: Broker-native structured signals with manual confirmation. The platform generates signals from named strategies with a predefined entry, target, and stop-loss, and you tap to confirm each order. Lemonn’s BOLT works this way, with strategies such as Scalping Pulse, Prime EMA Scalper, Swing King, Traffic Light, Inside Candle, Mean Reversion Bollinger, and EMA Cross, each with Auto TP/SL preset, all inside the same app you trade from. One distinction matters here. BOLT suggests, you confirm. It is structured signal discovery with predefined risk levels, not personalized investment advice, and not hands-off automation.

Be honest about which route you want before you shortlist. Traders who pick Route B because it sounds more serious, then never build anything, would have been better served by Route A or C.

Score against the route you actually need, not the one that sounds most advanced.

Check 4: API access and the real meaning of “free”

If you are going the API route, “free API” is a phrase that needs unpacking. Indian brokers structure API access very differently, and the pricing page is only half the story. Verify each of these against the broker’s live documentation before you commit:

• Subscription fee for order placement. Some brokers charge a flat monthly fee per API app. Others waive it entirely. Both models exist in the Indian market today.
• Market data feed cost. Order APIs and live data feeds are often priced separately. A “free” order API with a paid WebSocket feed is not free for anyone running an intraday strategy.
• Instrument and symbol limits on the data socket. How many contracts can you stream at once? This bites hardest when you scan an option chain.
• Rate limits. Orders per second, requests per minute, and daily caps.
• Login and session mechanics. Daily OAuth token generation versus persistent sessions changes how much babysitting your system needs. Some brokers also require a static IP for certain integrations, which means extra infrastructure cost.
• Historical data availability. Needed for any credible backtest.

Compare the popular free-API options in India and the honest answer is that the “cheapest” broker depends entirely on whether you need streaming data, how many symbols you track, and whether you can tolerate a daily re-login. Check the live pricing and documentation pages yourself. These terms change more often than blog posts get updated.

Score 10 only if the total cost of the API stack, including data, is documented and matches your usage.

Check 5: Order-per-second limits and algo tagging

Two things retail traders routinely get wrong:

• Your manual clicking can cross into algo territory. Under the current framework, order flow above a defined per-second threshold from a retail account is treated as algorithmic, whether or not you think of it that way. Rapid-fire scalping through an API can trip this.
• Tagging is not optional. Algo orders need to carry the strategy identifier. If your broker cannot explain how tagging happens on your account, you inherit that ambiguity.

Ask the broker directly: what is the order-per-second cap on my account, how are algo orders tagged, and what happens if I breach the threshold? A confident, specific answer tells you a lot about the whole operation.

Check 6: F&O execution tooling

This is where systematic strategies quietly leak money. Look for:

• Order slicing. Large F&O orders that exceed exchange freeze quantities need splitting. Doing it manually costs seconds you do not have. Lemonn’s Slicing feature auto-splits large F&O orders for smoother execution.
• Smart price assistance. Getting a better entry than a blind market order compounds across hundreds of trades. Lemonn’s Dash is a smart price assistant for entries.
• GTD orders. Limit orders valid up to a year mean set-and-forget levels genuinely stay set. Lemonn supports GTD orders valid up to a year.
• Fast exits. On expiry day, the difference between exiting one leg at a time and hitting a single control matters. Lemonn has Exit Now and Exit All for rapid F&O exits.
• Desktop-grade charting. A TradingView-powered Web Terminal, as Lemonn offers, means you are not squinting at a phone while managing four legs.

Weight this check heavily if you trade multi-leg structures or size above a lot or two.

Check 7: Built-in risk controls

Automation without risk controls is just a faster way to lose money. Must-pass items:

• Predefined stop-loss and target on every strategy, set before entry rather than improvised after.
• Maximum loss caps at the strategy or day level.
• A kill switch that actually works. One action that stops new orders and lets you flatten open positions.
• Position and exposure visibility in real time, not on an end-of-day report.

Lemonn’s approach reflects this. BOLT signals ship with Auto TP/SL preset on every trade, and SmartInvest strategies carry built-in stop-losses and risk controls. The point is not that the controls guarantee anything, because they do not. The point is that defined risk is part of the design instead of an afterthought.

Score 0 and disqualify if there is no reliable way to halt everything quickly.

Check 8: The cost stack

A systematic strategy that trades often lives or dies on total cost per round trip. Model all of it:

• Brokerage. Flat per-order pricing is far easier to model than percentage-based pricing. Lemonn charges a flat Rs 20 per executed order across equity delivery, intraday, futures, and options. One number, no tiers to reverse-engineer.
• Securities Transaction Tax. Different rates for futures and options, and charged on premium for options sells.
• Exchange transaction charges. Vary by segment.
• SEBI turnover fees and stamp duty.
• GST on brokerage and transaction charges.
• DP charges on equity delivery sells, if your strategy touches the cash segment.

Build a simple spreadsheet: cost per round trip multiplied by expected trades per month. If that number is a meaningful fraction of your expected edge, the strategy does not work, regardless of how good the backtest looked.

Check 9: Margin efficiency

Multi-leg options strategies live on margin benefit. Check:

• How SPAN and exposure margins are computed and displayed before you place the order, not after.
• Whether hedge benefit is recognised at order time. A broker that blocks full naked margin on leg one and only releases benefit after leg two goes through will strand your capital.
• Margin Trading Facility terms, if you use leverage in the cash segment. Lemonn’s MTF, branded Boost, offers up to 4x leverage on 1400+ stocks with interest charged only for the days a position is held. The current published rate is around 10.95% per annum, roughly 0.03% per day. Rates are subject to change, so verify the live figure.
• Pledge mechanics. Using existing holdings as collateral frees cash for F&O margin. Lemonn supports pledging holdings.

Weight this heavily if you sell options. Margin drag is the silent killer of premium-selling strategies.

Check 10: Backtesting, paper trading and honest disclosure

Look for three things and treat the third as the most revealing:

• Backtesting depth. How much history, and does it account for the actual bid-ask spread rather than mid prices?
• Paper trading or observation mode. Running a strategy live-but-unfunded for a few weeks tells you more than any backtest.
• How performance is disclosed. This is the tell. A credible platform shows performance metrics with clear caveats and never implies assured returns. Any platform showing only winning trades, or advertising a monthly return percentage, has told you exactly what it is.

Past performance never indicates future results. A platform that blurs that line should worry you more than a modest feature set.

Check 11: Reliability under load

The market does not care that your algo was correct if the order never reached the exchange.

• Outage history. Search for the broker’s name alongside expiry dates from the last year.
• Expiry-day capacity. Peak-load behaviour on Thursday afternoons is the real stress test.
• Exchange penalty and technical glitch disclosures. Brokers publish these. Read them.
• Fallback path. If the app is down, can you reach support and get positions squared off?

Check 12: Support and security

• 24×7 human support, not a chatbot loop. Lemonn offers 24×7 customer support.
• ISO/IEC 27001:2022 certification for information security management. Lemonn holds this.
• Two-factor authentication on login and on order placement.
• FIU compliance and bank-grade encryption on data in transit and at rest.

Must-pass. If you cannot reach a human while a leg is stuck, nothing else on this list matters.

Worked example: scoring three shortlisted brokers

Here is how the arithmetic looks for a hypothetical trader who sells weekly index options, three to four legs, five to eight trades a week, no coding background.

| Check | Weight | Broker A (API-first) | Broker B (full-service) | Broker C (no-code + signals) |

|—|—|—|—|—|

| 1. Regulatory | Pass/fail | Pass | Pass | Pass |

| 2. Algo framework | Pass/fail | Pass | Pass | Pass |

| 3. Automation route | ×2 | 6 (API only) | 5 | 9 (no-code fits) |

| 4. API cost | ×2 | 9 | 5 | 6 |

| 5. OPS and tagging | ×1 | 8 | 7 | 8 |

| 6. Execution tooling | ×2 | 6 | 7 | 9 |

| 7. Risk controls | Pass/fail | Pass | Pass | Pass |

| 8. Cost stack | ×2 | 8 | 4 | 9 |

| 9. Margin efficiency | ×2 | 7 | 8 | 8 |

| 10. Backtest/paper | ×1 | 8 | 5 | 6 |

| 11. Reliability | ×1 | 7 | 8 | 7 |

| 12. Support/security | Pass/fail | Pass | Pass | Pass |

| Weighted total | | 95 | 80 | 107 |

Broker A wins on API economics and loses on the checks this particular trader actually depends on. That is the whole argument for a weighted scorecard over a ranking. The same three brokers reorder completely if the trader is a Python developer instead.

Where Lemonn scores on this checklist

Running Lemonn through the same 12 checks, honestly:

• Regulation (Check 1): SEBI-registered broking through NU Investors Technologies Private Limited (INZ000304837), Research Analyst registration (INH000016764), NSE and BSE member, CDSL depository participant.
• Automation route (Check 3): Both no-code and structured-signal routes. SmartInvest gives rules-based algo strategies with built-in stop-losses for non-coders. BOLT gives real-time, strategy-led F&O signals with predefined entry, target, and stop-loss, plus Auto TP/SL, and every order needs manual confirmation, so the trader stays in control.
• Execution tooling (Check 6): Slicing for large F&O orders, Dash for smarter entries, GTD orders valid up to a year, Exit Now and Exit All for fast expiry-day exits, and a TradingView-powered Web Terminal for desktop.
• Cost stack (Check 8): Flat Rs 20 per executed order across equity delivery, intraday, futures, and options. One number to model.
• Margin (Check 9): MTF up to 4x on 1400+ stocks with day-count interest, plus Pledge to unlock capital from existing holdings.
• Support and security (Check 12): 24×7 customer support, ISO/IEC 27001:2022 certification, FIU compliance, bank-grade encryption, 2M+ users.

Where Lemonn is a weaker fit: if your requirement is a raw, self-hosted API with custom historical data pulls and you want to write every line of the logic yourself, an API-first broker will serve you better. Use the scorecard, not the brand.

Five red flags that should end an evaluation on the spot

• No published SEBI registration number. Non-negotiable. Walk away.
• Any promise, implication, or screenshot suggesting assured returns or a guaranteed win rate. SEBI-regulated entities do not make return promises. Anyone who does is either unregulated or non-compliant.
• Automation offered through an unaffiliated third-party tool with no explanation of algo approval or order tagging. You inherit the regulatory exposure.
• No kill switch or no clear way to flatten all positions quickly. One bad morning is all it takes.
• Opaque pricing. Brokerage that requires a call with a relationship manager to understand, or “free” claims that unravel when you read the data-feed terms.

A realistic risk note

SEBI’s own studies on the equity derivatives segment have consistently shown that a large majority of individual F&O traders lose money, and that losses tend to grow with trading frequency. Automation does not repeal that finding. What automation can do is remove some of the discretionary errors, such as moving stop-losses, doubling down, or skipping the exit, that make outcomes worse than the strategy itself.

Build your automation plan around three honest assumptions: the strategy will have losing stretches, the cost stack is a fixed drag you cannot negotiate away, and the risk controls exist to be triggered rather than overridden. Size positions accordingly. Treat the checklist as a way to eliminate avoidable failure points, and accept that no platform, signal, or algorithm changes the underlying risk of the derivatives market. This article is general information, not individualized investment advice.

FAQs