What algo trading is and is not
The misconceptions to clear first
- It is not a strategy. Automation executes a strategy; it does not supply one.
- It is not passive income. There is a daily routine, and strategies stop working.
- It is not lower risk. The same positions carry the same risk whoever places them.
- It does not require coding on a no-code platform, though it does require precision.
The order to learn things in
- Understand what an option is. Calls, puts, strikes, expiry, premium. Without this nothing else makes sense.
- Understand moneyness and the Greeks, roughly. Not to compute them — to know why a position behaved as it did.
- Build one simple strategy. Two legs. Specify it precisely enough that software could execute it.
- Backtest it and read the report from the risk end. Drawdown before profit.
- Paper trade it for two expiry cycles. Watch behaviour, not P&L.
- Set up a broker, an IP and the daily routine. Only now does this become necessary.
- Go live at one lot. Measure your real execution cost before scaling.
What to learn first, specifically
| Concept | Why it comes early | Where |
|---|---|---|
| Moneyness | Decides which strike you trade and why | <a href="/blog/itm-atm-otm-strikes-explained">ITM, ATM and OTM</a> |
| The Greeks, roughly | Explains why a position moved | <a href="/blog/option-greeks-explained">Option Greeks</a> |
| Implied volatility | Explains losing on an unchanged index | <a href="/blog/implied-volatility-explained">Implied volatility</a> |
| Margin | Decides what you can actually trade | <a href="/blog/margin-for-option-selling">Option selling margin</a> |
| Reading a backtest | Stops you funding a fitted strategy | <a href="/blog/how-to-read-a-backtest-report">Backtest reports</a> |
Choosing a first strategy
What to look for
- Two legs, not four — fewer things to go wrong in execution
- Defined risk, so a bad day cannot be catastrophic
- Rules you can state in one or two sentences
- A margin requirement comfortably inside your capital
- Liquid strikes, so execution is not the variable being tested
The mistakes that cost beginners money
| Mistake | What happens | The fix |
|---|---|---|
| Going live straight from a backtest | Logic errors found with real money | Paper trade for two cycles first |
| Sizing to available margin | A volatility rise creates a shortfall | Size to your risk rule, with headroom |
| Selling far out-of-the-money options | High win rate, then one large loss | Understand what the premium is paying for |
| No written stop condition | Abandoning at the worst moment | Decide it in rupees before deploying |
| Changing the strategy mid-test | Several strategies tested briefly | Leave it alone for a full cycle |
| Treating a backtest as a forecast | Expectations that cannot be met | Read it as evidence about the past |
The daily routine, so it is not a surprise
- Complete the broker login. Sessions expire daily on the self-serve brokers.
- Start your bots. Deploying is not starting, though the scheduler can automate this part.
- Check the logs once after your entry time. Thirty seconds, and it catches most problems on the day rather than at the end of the week.
Realistic expectations
- Most strategies do not work. That is what testing is for, and rejecting ideas cheaply is the skill.
- Live results trail backtests. Structurally, on every platform.
- Drawdowns are normal. Your backtest already showed you one; the live version will feel different.
- Nothing compounds without surviving. The first objective is not losing enough to stop.
The short version
- Automation executes a strategy; it does not supply one
- The first five steps cost nothing beyond a plan
- Learn moneyness, the Greeks roughly, IV and margin — enough to name causes
- Start with two legs and defined risk, not naked short options
- Write your stop condition in rupees before you deploy
- Expect most strategies not to work — that is the process functioning
Frequently asked questions
Not on a no-code platform. You define the rules through the interface. What you do need is precision — a rule vague enough to need your judgement cannot be automated.
Nothing to backtest or paper trade beyond a plan. For live trading you need the margin your chosen strategy requires with headroom, which varies hugely between a defined-risk spread and a naked short option.
Something with two legs, defined risk, and rules you can state in a sentence. A vertical spread fits all of those and has arithmetic you can verify by hand.
No. The same positions carry the same risk. Automation removes execution errors and the need to be present; it does not change what the position can do.
After backtesting over a varied period and paper trading for at least two expiry cycles. Then live at one lot for a few weeks before scaling.
Not writing a stop condition before deploying. Strategies get abandoned during normal drawdowns, which locks in the loss and misses the recovery.
Research Analyst strategies are included in the plan and are a reasonable starting point. Review the drawdown and margin yourself, and paper trade before committing — the responsibility for running it is still yours.
There is no honest answer to that, and any source giving you one is worth avoiding. The first month's realistic objective is learning your execution cost without losing enough to stop.
Start with a free 3-day trial
Build a strategy, backtest it and run it on paper — no broker, no IP and no money needed to try it.
