Risk Disclosure Statement

1. What this statement covers

triStack AlgoTrader is managed hosting for trading software you operate yourself. We provision and run an instance for you: your own database, your own dedicated egress address, your own engine. You write the algorithms. You schedule them, start them, stop them, and set every limit that constrains them. The software executes what you configured, and it does so literally.

triStack AlgoTrader is hosting infrastructure, not an investment adviser and not a signal provider. Tristack Technologies LLP is not registered with the Securities and Exchange Board of India (SEBI) in any capacity: not as an Investment Adviser, not as a Research Analyst, not as a Portfolio Manager, and not as a stock broker. It does not hold, and does not claim, any SEBI registration.

This statement describes risk. It sits alongside the Terms of Service, which govern your use of the service, the Disclaimer, and our compliance position. Nothing here is exhaustive. No written document can list every way a market position can go against you, and no list of failure modes is ever complete.

Where this statement says “you”, it means every person whose broker account is connected to an instance, whether the account belongs to you or to someone who has authorised you to operate it.

2. Market risk

The price of any security can fall as easily as it can rise, and it can do so faster than any order-routing system can react. The following are ordinary market conditions, not exceptional ones:

• Volatility. Prices can move several percent within seconds around results, policy announcements, index rebalancing or global news. A position that looked small when you sized it can become large in its effect on your capital inside a single session.
• Gap openings. Markets close overnight and over weekends and holidays; news does not. A stock can open far below the previous close, past any level at which you intended to exit. A stop level is a trigger, not a guarantee of the price you will get.
• Circuit limits and price bands. When a security hits an upper or lower circuit, or a market-wide breaker halts trading, you may be unable to exit at any price until the band is revised or trading resumes. Your exposure continues while your orders sit unexecuted.
• Corporate actions and contract changes. Splits, bonuses, mergers, expiries and lot-size revisions change what a quantity means. A strategy written months ago keeps doing exactly what it was told to do, including after the underlying contract specification changed underneath it.
• Regime change. A strategy that worked in one volatility regime can lose steadily in the next without anything visibly breaking. Nothing in the software detects this, and nothing warns you about it.

3. Leverage and margin

Intraday equity, futures and options positions are leveraged: you control an exposure much larger than the margin you deposited. Leverage is symmetric. It magnifies adverse moves exactly as much as favourable ones.

• Losses on leveraged positions can exceed the margin deposited. Your broker can call for additional margin at short notice and can square off your positions without further reference to you if it is not met.
• Writing (selling) options carries risk that is theoretically unlimited on the upside and very large on the downside, against a premium that is capped. An automated option-writing strategy can accumulate that exposure faster than you would by hand.
• Margin requirements are set by the exchange, the clearing corporation and your broker, and can be raised intraday. A position that was fully margined in the morning can be short of margin by the afternoon without you doing anything.
• A strategy that sizes positions as a fraction of available margin will size up as your account grows and will keep sizing up after the edge has stopped working. Position sizing is a decision you make in code, and it is the decision that determines whether a bad week is survivable.

4. Your own code is the largest risk

These are ordinary defects, seen repeatedly, each of which places real orders with real money:

• A sign error or an inverted condition. Buy where you meant sell, or greater-than where you meant less-than. The strategy is not confused by this and will trade the inverse of your idea all day.
• Quantity and lot-size mistakes. Trading lots where you meant units, or units where you meant lots, changes your exposure by the lot multiple. So does a hard-coded lot size that the exchange has since revised.
• An exit path that is never reached. Entries are the easy half. An exception thrown between entry and exit, a condition that can never become true, or an exit that depends on a variable reset at the top of the loop, all leave you holding a position the strategy believes it has closed.
• No kill switch and no loss limit. If nothing in your code stops it, nothing stops it. There is no platform-level daily loss cap watching over your account, and you should not assume there is one.
• Retry loops without a bound. A rejected order retried in a tight loop can place hundreds of orders, exhaust your broker’s rate limit, or fill far more size than you intended if some of the retries succeed.
• Duplicate orders after a restart. A strategy that keeps its state in memory and is restarted mid-position can re-enter a position it already holds. State that matters must survive a restart, and making it do so is your responsibility.
• Timezone and schedule errors. Schedules run on IST. A strategy scheduled in the wrong timezone, or one that assumes every weekday is a trading day, will act at times you did not intend, including on exchange holidays.
• Untested edge cases. Expiry day, a symbol change, a missing tick, an empty order book, a partial fill, a zero quantity after rounding. Code that was never exercised on these paths is code whose behaviour on them is unknown to you.
• A strategy someone else wrote. Code you bought, copied or were given carries every one of the above, and you cannot see which. Running it makes its behaviour yours.

The software provides a paper-trading sandbox and an order-approval mode in which every order must be confirmed by hand before it is sent. Both exist for precisely this risk. Whether you use them is your choice, and the consequence of not using them is also yours.

5. What automation changes

Automation does not create new market risks. It changes the speed, the scale and the supervision, and each of those changes the size of a loss.

• Speed. A mistake that would have cost one bad trade by hand can cost fifty in a minute. By the time a notification reaches your phone, the position may already be larger than you would ever have taken deliberately.
• Absence of hesitation. A human trader pauses at an absurd price or an impossible quantity. Code does not. It will send an order at a price no person would have typed.
• Correlated failure. Several strategies keyed off the same signal, the same data feed or the same underlying will lose together, on the same morning, and their combined margin call will arrive at the same time.
• Silent operation. A strategy that stops trading does not necessarily announce it. Neither does one that has been quietly losing since a change you made three weeks ago. Alerts tell you about events; they do not tell you about the absence of events.
• Over-trust. The most expensive assumption in automated trading is that because the system ran correctly yesterday, it is running correctly today. Verify against your broker’s own order book, which is the authoritative record.

6. Execution risk

An order leaving your instance is the beginning of the process, not the end of it. Between the decision and the fill sit your broker, the exchange and the order book, none of which we control.

• Slippage. Time passes between your strategy’s decision and the order reaching the exchange, and the order then queues and fills on its own terms. In fast, gapping, illiquid or circuit-limited markets the difference between the price you modelled and the price you got can be material.
• Rejections. An order can be refused by your broker or the exchange for insufficient margin, a product or segment restriction, a scrip in the F&O ban period, a freeze quantity, a price outside the band, an expired session, or an unregistered source address. A rejected entry with a successful exit, or the reverse, leaves you somewhere you did not intend to be.
• Partial fills. A part-filled order leaves a position smaller than your strategy believes it holds, and any exit sized off the intended quantity will then be wrong in the other direction.
• Modify and cancel are not instant. A cancellation races the fill. An order you believe you cancelled may already have executed.
• Nothing can be undone. An order that reached the exchange is a real order. The record on your instance will tell you exactly what happened and when. It cannot reverse it.

7. Technology, hosting and connectivity

Your instance depends on systems that neither you nor we control end to end. Any of the following can prevent an order, delay it, or interrupt a strategy midway:

• Your instance being unreachable. If your instance is unreachable, orders it would have placed are not placed, and nothing is replayed when it comes back: a strategy that missed an entry has missed it. Trading through a broker terminal is always available to you as a fallback, because the broker account is yours and we never stand between you and it.
• No uptime percentage is promised. We monitor your instance and respond to incidents, and we have deliberately not published an availability figure. Do not build a strategy on an assumption of continuous availability, and do not assume any figure you have not been given in writing on your order form.
• Broker API downtime and rate limiting. Broker APIs go down, degrade, queue or throttle, particularly at the open, at expiry and during high volatility, which is exactly when your strategy matters most.
• Session and token expiry. Indian broker access tokens expire on the broker’s schedule, in most cases daily around 3:00 AM IST, and some require a fresh login or a two-factor step. Until the session is renewed, that account cannot be read from or traded through, whether or not your strategies are running.
• The dedicated address not being registered. Your broker accepts API orders only from an address you have whitelisted against your own credentials. If that registration lapses, or your broker changes what it checks, your orders stop going through and the strategy will not know why.
• WebSocket disconnection and market-data gaps. Feeds drop and reconnect. A strategy that assumes an unbroken tick stream can act on a stale price or miss the tick that would have triggered its exit.
• Maintenance and upgrades. Patching the software or the host can require a brief restart. We schedule planned work outside NSE and BSE market hours wherever we can and tell you before it happens, but an upgrade can also change behaviour your code depended on.
• Defects in the software itself. The platform is software and contains defects. We fix the ones we find and the ones you report. Neither of us will find all of them first.
• Connectivity and power. Internet failure, routing problems, data-centre incidents or power loss at any point in the chain.

Nothing is replayed. A strategy that missed an entry has missed it, and there is no queue that catches up when service returns. Know how to square off or cancel directly with your broker, on their terminal or app, without relying on your instance. That is the fallback, it always exists because the account is yours, and the time to learn it is not during an incident.

8. Market data and backtest risk

A strategy is only as good as the data it was built on and the data it runs on, and both can be wrong in ways that are invisible until money is involved.

• Live data can be late, missing or wrong. Ticks are dropped, feeds lag, snapshots are stale after a reconnect, and an indicator computed on a partial candle is not the indicator you tested.
• Historical data is not the market. Adjustments, survivorship, contract rollovers and missing sessions all change a backtest’s answer.
• Look-ahead bias. A backtest that uses a candle’s close to decide a trade inside that candle produces excellent results and is not reproducible in life.
• Overfitting. Parameters tuned until the equity curve looks good describe the past precisely and predict nothing.
• Sandbox results are simulated. The paper-trading sandbox has its own database and its own notional capital. It has no queue position, no impact cost and none of the rejection behaviour of a real exchange. A strategy that is profitable in the sandbox has demonstrated that it runs, not that it makes money.

9. Liquidity and impact cost

An automated strategy can put more size into a book, faster, than the same person would by hand. In a liquid large cap or a near-month index contract that is usually immaterial. In a thinly traded stock, an illiquid strike, a far-month contract or a small cap under a price band, it is not.

• Order size that is large relative to the resting depth walks the book, so later fills come at worse prices than the first.
• Wide bid-ask spreads mean impact cost is incurred immediately on entry, before the position has moved at all. A strategy with a small edge per trade can be entirely consumed by it.
• Exiting is harder than entering. The same size has to find the other side, and it may have to do so when liquidity has thinned or the book has moved away.
• A strategy that trades frequently pays the spread and the brokerage every time. Costs are certain; the edge is not.

Treat instrument liquidity as a constraint on how you size and where you trade, not as an afterthought.

10. Regulatory and tax risk

The rules that apply to trading, to automation and to operating accounts on behalf of others are set by SEBI, the exchanges, the clearing corporations, your broker and the tax authorities. They change, sometimes at short notice.

• The framework for retail algorithmic trading is evolving. Requirements around API-based automated order placement, the registration or approval of algorithms, and the obligations of brokers and of the people using their APIs continue to develop. A change may require you to alter how you trade, may require us to alter, restrict or withdraw a feature, or may impose an obligation directly on you. What is settled and what is not is set out on the compliance page, including the parts we have deliberately not concluded on. Take your own advice.
• Your broker’s terms govern your account. Brokers place conditions on API use, on automated order placement, on static IP registration and on third-party tools. Confirming that your intended use is permitted under your own broker agreement is your responsibility.
• Operating other people’s money is regulated. If you manage, solicit, pool or trade funds belonging to anyone other than yourself, or charge for doing so, or share in outcomes, you may require registration or authorisation in your own name. Obtaining and maintaining it is entirely your responsibility. triStack AlgoTrader does not provide it, does not extend any registration to you, and does not verify whether you hold one.
• Tax. Classification of trading income, set-off, audit thresholds, advance tax and reporting depend on your circumstances and can change. We produce records, not tax advice. Consult a qualified professional.

11. No guarantee of outcome

triStack AlgoTrader does not offer, promise or imply any profit, assured return or performance outcome, and does not share in your profits or losses.

What we sell is infrastructure: isolation, a dedicated address, a machine that stays patched and monitored, and a record you can read. Those are engineering properties. They say nothing at all about whether your strategy makes money, and we will never present them as if they did. Consistent, reliable execution of a losing strategy produces a reliable loss.

triStack AlgoTrader is managed hosting for trading software you operate yourself. Tristack Technologies LLP is not registered with SEBI in any capacity, publishes no signals or recommendations, never handles your funds, and guarantees no return. Trading in securities carries a substantial risk of loss, and you are solely responsible for your trades.

12. Suitability: an honest self-assessment

Work through the following before you connect a live account. If you cannot answer yes to all of them, automated trading is probably not appropriate for you right now.

• I understand the instruments my strategies will trade, including how futures and options behave, and I would be willing to place these trades by hand.
• I wrote, or fully understand, every strategy I am going to run, and I can read what it does on every path including the error paths.
• My strategies have an exit for every entry, a bounded retry, and a limit that stops them after a loss I have decided in advance.
• I have run them in the sandbox, and I understand that a sandbox result is a simulation and not evidence of profitability.
• I know how to square off and cancel directly with my broker, without using my instance, and I could do it today.
• I can monitor my accounts during market hours, or I accept the consequence of not being able to, including being left in a position after a failed exit.
• My margin is sufficient for the size I have configured, with headroom for an intraday margin increase.
• I could absorb a total loss of the capital in these accounts without affecting my obligations, my household or my ability to meet a margin call.
• I am not trading with borrowed money, with money I need soon, or with money whose loss I would have to explain to someone else.

13. Acknowledgement

By accepting an order form, using an instance, or connecting a broker account to one, you confirm that you have read this Risk Disclosure Statement, that you understand it, and that you accept the risks described in it. You further confirm that:

• you are trading on your own account, or on an account you are lawfully entitled to operate;
• you wrote or chose every strategy running on your instance, and you set its limits, its schedule and its size;
• you are solely responsible for every order placed in your accounts and for every consequence of it, including losses; and
• you have not relied on triStack AlgoTrader for advice of any kind, for a view on any security, or for any assessment of whether a trade or a strategy is suitable for you.

This statement should be read with the Terms of Service, the Disclaimer and the compliance page. We may update it as the product, the market structure or the applicable rules change, and the date at the top of this page reflects the current version.