Where a Free Tier Breaks on Expiry Day, and How to Find Out on Tuesday
A free tool can work well for eleven trading days out of twelve. You place a few orders, watch a few positions, and the limits never show up. Then comes a weekly expiry morning. You enter four legs in a hurry, spot moves 80 points in five minutes, three stops fire together, and the tool does something you have never seen it do.
Expiry day is not the same as other days. It compresses activity. Orders bunch up at the open, stops cluster during sharp moves, and exits bunch up again in the last hour. Any limit that sits quietly on a normal day can bite on this one. Free plans have limits more often than paid ones, and they are usually set for casual use, not for a busy expiry session.
Why expiry day is the day to test, not the day to learn
The weekly contracts that remain are NIFTY on NSE and SENSEX on BSE. SEBI's rule of one weekly expiry per exchange ended the Bank Nifty weeklies, so a trader who sells weekly index options now has these two to work with. Each has its own expiry day, and both days are busy for option sellers.
The first reason expiry is demanding is gamma. On the last day, an at-the-money option moves quickly with spot, so a short straddle's loss grows fast on a trending move. That means stops fire sooner and more often than on a mid-week day. The second reason is the clock. EliteAlgo's desk study of 163 NIFTY and SENSEX weekly expiry sessions between January 2025 and July 2026 shows that the average ATM straddle premium falls from 100% of its 09:16 level to about 88.6% by 11:16. So most of the premium action happens in the first two hours, which is also when most entries, adjustments and early stop-outs happen. Your tool's heaviest load lands in that window.
The third reason is consequence. On a quiet day, a missed order costs you an opportunity. On expiry day, a missed stop costs you real money fast. A limit that blocks a stop order is much worse than a limit that blocks an entry. So the test is simple: find out which limits can stop a protective order, not just an entry order.
The four limits that bite
The table below names each limit, what it does when you hit it, and the signal that tells you it is happening. These are general mechanics. Each platform handles them in its own way, so confirm with the vendor.
LimitWhat breaks on expiry dayHow you noticeOrder rate (orders per second or minute)Orders queue, get delayed, or are rejected during bursts such as a 4-leg entry or several stops firing togetherTime gap between signal and fill; reject messages; legs filled seconds apartLeg or position capThe tool refuses a fifth or sixth leg, or an adjustment that would exceed the capHedge or adjustment legs never appear; strategy runs "naked" without you realisingSession or token lifeLog-in expires, API token lapses, or the app is closed by the platform mid-sessionSilent stop of strategy updates; positions open with nothing watchingSupport hours and channelNo human reachable at 09:20 on the day it mattersTicket sits unread until after market close
If you want to see how a comparison of free options reads when it is organised around limits rather than features, a published page on free algo trading software is one place to look, and the same limit-by-limit questions can then be asked of every name on it. Whichever list you use, treat a limit that is not documented as unknown, and test unknown limits.
The next four sections take each limit in turn, with a way to measure it.
Order rate: do the arithmetic before the exchange does
An order-rate limit is a ceiling on how many orders can go out in a given time. It can be set by the software vendor, by your broker, or by the exchange framework. The actual figures differ and are revised from time to time, so look up the current ones. What you can do today is count how many orders your strategy needs in a bad minute.
Here is an illustrative count for one NIFTY short strangle with stop-losses, on one lot. Entry places two sell orders. Each leg needs a protective stop, so two more orders. That is four orders to get set up. If your stops are placed as separate orders and your tool also moves them (a trailing stop), each move is another order, called a modification. If the tool trails every few seconds, a single leg can generate dozens of modifications in a trending hour. Two legs doubles it.
Suppose the trail logic fires a modification every 5 seconds on each of two legs while price is trending. That is 24 modifications a minute. Over a 20-minute trend, 480 modifications, plus the four setup orders. A vendor limit expressed as a daily order cap could be eaten by one hour of trailing. A per-minute cap might be hit the moment you add a second strategy to the same account. The numbers here are made up to show the method, and the method is what you should reuse: list every order type your strategy sends, multiply by how often it sends them in the worst plausible minute, and compare with the documented limit.
Pay attention to which orders are throttled first. In some set-ups, a throttled queue treats all orders equally. That means a trailing-stop modification can sit in front of an urgent exit. Ask the vendor, or test it, whether protective and exit orders get priority. If you cannot find out, reduce the chatter: trail less often, trail in larger steps, or use a fixed stop on the exchange side.
There is also the regulatory side. SEBI has issued a framework for retail participation in algorithmic trading, which covers how strategies are identified and how order flow is controlled between broker and trader. Details have been updated more than once, so read the current text on sebi.gov.in and ask your broker how it applies to the tool you use. A free tool that was built before a rule change may behave differently from one built after.
Leg caps: what a half-built structure costs
Many tools limit the number of legs per strategy or the number of live strategies. A simple short strangle needs two legs. Add protective wings and you have four. Add an adjustment leg during the day and you may need five or six. If the cap is four, the adjustment will fail, and it will usually fail on the very day when you need it.
The sharper danger is partial building. Suppose your structure has four legs: two sold, two bought as hedges. The sensible order is to buy the hedges first, then sell. If the tool sends all four at once and a cap or rate limit lets only two through, which two? If the sells go first and the hedges are rejected, you hold an unhedged short straddle that you designed to be an iron condor. Margin requirement is higher, and the risk is no longer what you planned.
Test this on purpose. Build your full structure in the tool, including the adjustment leg you might add, and see whether it accepts all of it. Then check the order in which the tool sends the legs. A well-designed tool lets you set the sequence, with buys before sells. If your tool does not, then you must either enter hedges by hand first or use a different structure.
Also check what the tool does when one leg is rejected by the broker. Some platforms retry. Some cancel the rest. Some leave the partial position and notify you, if you are lucky enough to look at the notification. For a margin-sensitive structure this matters a great deal. Any one of those behaviours might be fine, but you must know which one you have before the day.
Session life: the silent stop
This is the one that surprises people most. Many broker APIs use access tokens that expire daily, so a log-in is needed each morning. Some desktop and web tools time out after a period of inactivity or when the machine sleeps. A free tool hosted in a shared environment may restart for maintenance. None of these will necessarily tell you.
The failure looks calm. The screen still shows your positions from the last update. The strategy status says "running". But the feed behind it has stopped, so the stops are not being evaluated. You notice when you look at your broker app and the numbers disagree.
There are three ways to defend. The first is a heartbeat check: confirm that the last-tick time in the tool keeps moving, and write down a maximum acceptable age, such as 10 seconds. The second is to place stop orders on the exchange side, through your broker, so they do not depend on the tool being alive. The third is a manual look at the broker app at fixed times, say 09:30, 11:00 and 13:00, and a note of the open positions. This is not elegant, but it is cheap and it works.
Support gaps: who answers at 09:20?
Support is the least technical limit and often the most costly. Free tiers commonly offer community forums, email with slow replies, or a ticket queue. That is reasonable for a free product. The mismatch is with your need: on expiry day, your problem has a half-life of minutes.
Plan for no support. That means writing down, before the day, what you will do for each type of fault. If the tool stops updating, you flatten at the broker app. If an order is stuck, you cancel at the broker app. If a leg is missing, you add it by hand. The broker's own trading app should be installed, logged in and tested on the phone that is with you. It is the tool of last resort, and it needs to work when everything else does not.
Also learn the broker's support path. Many brokers have a dealing desk or a priority line for order issues during market hours. Save the number. A free algo tool cannot fix a broker-side rejection, but the broker can.
Finally, check the tool's status page or community feed for the days when others had problems. If you see a pattern of complaints on expiry mornings, take it seriously. Ask in the forum how people handled it. The answers will tell you how the community behaves under stress, which is a good guide to how support will behave under yours.
The pre-expiry test, step by step
The best time to run this test is the day before your expiry, in the late morning, when markets are open and the tool is under normal load. Paper or sandbox mode is ideal. If your tool has none, use the smallest live size possible, one lot, and far out-of-the-money strikes so the risk is low. The goal is to see behaviour, not to make money.
Step 1: The full-structure build
Create the exact structure you plan to trade, including hedges and the adjustment leg. Note whether the tool accepts it, and in what order it sends orders. Record any warning or silent change.
Step 2: The burst test
Fire the entry, then immediately place and modify stops on all legs several times in a minute. Count orders sent and orders accepted. Compare with the order count you estimated earlier. If any are delayed or rejected, note the time gap.
Step 3: The priority test
While the burst is running, send an exit order for one leg. See how long it takes to reach the broker compared with a normal exit. This tells you whether protective orders are queued behind trailing modifications.
Step 4: The idle test
Leave the tool alone for the longest stretch you are likely to leave it, such as 45 minutes, with a position open. Come back and check the last-tick time, the strategy status and the broker app. If the tool needed a log-in or showed stale data, you have found a session limit.
Step 5: The kill test
Close everything from the broker app while the tool is still running. See how the tool reacts. Does it try to re-enter? Does it show a phantom position? A tool that fights your manual exit is dangerous on expiry day.
Step 6: The support test
Send a short, specific question to the support channel and note the reply time. You are not trying to get an answer. You are measuring how long an answer takes.
Write the results in a one-page note: limit found, how it showed up, workaround. Keep it with your trading log. Repeat after any update to the tool or any change in broker terms.
Reading the results: three outcomes
After the test you will be in one of three positions. In the first, nothing broke and you have a short list of documented numbers. Good. Note the margins: how close the burst test came to any ceiling. If you used 80% of an order budget in the test, you have no room for a worse day.
In the second, something broke but you have a workaround. For example, the tool caps legs at four, so you will trade a four-leg structure and handle adjustments by hand at the broker. That is a legitimate way to use a free tool, as long as the manual step is written down and practised.
In the third, something broke in a way that touches protective orders, and you have no workaround. That is a reason to change the strategy, the tool, or the size. Smaller size fixes some limit problems directly, because fewer lots does not reduce the order count, but fewer legs and less trailing do. A simpler strategy with fixed stops on the exchange side may run safely on a tool that would fail a complex one.
The desk data is a reminder of why protective orders matter so much. In a study of 159 usable sessions for a short ATM straddle entered at 09:20, the median give-back from peak profit to the close was 42.8% of the peak, and 48 of the 159 sessions (about 30%) showed a profit at some point and finished with a loss. A position that had no working exit during the swing from gain to loss has no way to avoid that outcome. The exit is the part of the system you cannot afford to have throttled.
Free is not the problem; unmeasured is
None of this is an argument against free tools. They lower the cost of learning, and many traders run small, simple strategies on them without trouble. A one-leg or two-leg strategy with fixed stops generates few orders and fits inside most caps. The trouble starts when a strategy grows in complexity while the tool stays the same, and nobody re-measures.
The rule is to match the strategy to the tool you have measured. If the strategy needs 6 legs, 3 trailing stops and a second index in the same account, test that exact load. If it passes with room to spare, fine. If it passes only barely, treat the tool as being at its limit, and plan to move before the strategy grows further.
Cost still matters. Paying for a tool does not remove limits either, since every product has some, and a paid plan with a lower cap than your need is no better than a free one. The only way to know is to test with your own load. That is the general lesson, and it is why the test above does not mention any brand.
Frequently Asked Questions
Can I run an expiry-day short straddle on a free algo trading platform?
Often yes, for a simple two-leg structure with fixed stops. The risks come from order-rate caps, leg caps, session timeouts and slow support. Run the pre-expiry test with your exact structure to find out which of them applies to your tool.
What happens if my algo stops running in the middle of an expiry day?
Your open positions stay open at the broker. Any stop that lived only inside the tool stops working. That is why a manual check of the broker app, and exchange-side stop orders where your set-up supports them, matter.
How many orders per second can I send to the exchange?
The ceiling depends on your broker, the exchange and the current regulatory framework, and it has been revised more than once. Ask your broker for the current figure in writing and check the regulator's pages.
Is a paid platform always safer on expiry day?
No. Paid plans also have caps, and a plan may suit casual use better than heavy use. What makes a tool safe is that you have measured its limits against your own load, written down the workarounds, and practised them.
Should I use a free tool for hedged and unhedged strategies equally?
Be more careful with unhedged strategies. A missed or delayed exit on an unhedged short option has a much larger downside than on a hedged one. If the tool cannot guarantee the exit path, hedges or smaller size are the cheaper insurance.
Where do I report problems if the vendor does not answer?
Start with your broker, because they can see the order trail and can act on rejections. Keep screenshots and timestamps. The exchange and the regulator have their own grievance channels, described on their websites.
Conclusion
A free tier is a product with limits, and expiry day is when limits show. Four are worth checking: order rate, leg count, session life and support. Each can be measured with a simple rehearsal on the day before expiry.
The aim of the rehearsal is a one-page note: what broke, how it showed up and what you will do about it. With that note you can trade a simple strategy on a modest tool with open eyes. Without it, you find out the limits at 09:20 on the busiest morning of your week.
Algo trading carries a real risk of loss, and a working tool does not remove it. EliteAlgo is analytical software and not an investment adviser, and nothing in this article is investment advice. The order counts above are illustrative, and the session figures come from a past-data study that does not predict future results.
