Markets have been extremely volatile lately and I’ve started wondering whether manual trading actually works better than algos in these conditions.
Algos obviously have advantages like speed and emotionless execution, but during sudden reversals and messy price action, sometimes it feels like experienced discretionary traders adapt much better.
So wanted to ask:
- Are you trading manually or using algos?
- Are rule-based algos still working properly for you?
- Is AI trading actually useful for retail traders or mostly hype?
Well for me its all manual order processing. Mostly because
-
Can’t get all these python, Static IP and the latest tech surrounding automated trading into my head.
-
On an Intraday perspective, I’m in the arena to spend some spare time and keep my mind busy. Manual Trading and Analysis gets the job done for people who want to be busy.
And about the AI, In India I think traders and investors are still caught in the loop of using century old indicators and outdated principles to find answers using AI and cant think outside the box. Unless that perspective changes and focus towards statistical and mathematical probabilities, based on multi dimensional granular data , AI is not much useful in the current form as accessible to the retail crowd. .
1 Like
Using an algo — specifically a rule-based Python bot on
Zerodha Kite Connect.
On volatile markets: rule-based algos actually perform
better for retail in volatility, not worse. Here is why —
discretionary traders “adapt” but most of that adaptation
is emotional. They widen stops, hold losers hoping for
reversal, skip valid entries because the last one failed.
The algo just follows the rules.
The real challenge in volatility is not the strategy —
it is position sizing. I keep lot size smaller when VIX
is elevated and only trade Iron Condor or Iron Butterfly
when VIX is in a stable range. Directional trades only
on strong trend confirmation.
On AI trading for retail: mostly hype at this stage.
The edge is not in prediction — it is in consistent
execution of a simple rule set with strict risk management.
No ML model needed for that.
Rule-based algos are still working. The ones that fail
in volatility are usually over-optimised on calm market
data with no volatility filter built in.
Key is having a kill switch — when daily loss limit hits,
bot stops. No override. That single rule saves more money
than any fancy prediction model.