English
Open in Telegram
PPython Algo Trading discussion

Python Algo Trading discussion

✅ High trust
@pythonalgodiscussion · group · Crypto · indexed since 2026-09-07
1 770members
27writing in 30 days
149messages in 30 days
17 298messages in the index
Whole thread · 1 reply →
P
Interesting find of today Zerodha, Dhan, and Upstox all host their live trading engines and WebSockets directly inside AWS Mumbai (`ap-south-1`). If you run your trading bot on a standard AWS Mumbai EC2 instance, your orders and market ticks travel entirely over AWS's internal datacenter fiber instead of the public internet. This cuts your network latency down to 1 millisecond TO 5 millisecond levels—essentially giving you budget "co-location" speed without paying premium server fees.
👍1
  1. G
    that's not what colocation speed is bro. co location literally means you are hardwired with exchange's network itself. you are still few milliseconds behind it which you will never know because you are getting feed from broker and not directly from exchange. I have seen co location speed and had bought dump data too. it's insane and the difference it makes is ridiculous when you trade with large capital. you can use vultr server and get similar latency too.
  2. M
    Have you seen dhan their orders gets rejected cause of shit margin systems for exact same setup. Even normal order placement is bad in them
      1. M
        Had lost money most cause of their stupidity, had left them
        😱1
    1. G
      Dhan is one of the worst. even data quality is extremely bad. just my opinion. others may like it and never faced issues.
      1. M
        I had used their option chain endpoint and built system on it and it was fine while mcx data had issues
Whole thread · 6 replies →
G
just use any server in mumbai and you should get huge drop in latency. I have been using it since 2 years now.
  1. P
    yes i know that , but some Saas who are not authorized from nse are just hosting on AWS with the netwroking modifion and if you do this you can save good amount of time in trunaround , while achiving 50 to 60 percent of colo speed
Whole thread · 1 reply →
G
and no broker is allowed to host co location server in aws. it's NSE's own facility. their primary servers which serves clients maybe in aws but it's not legally allowed to host their own servers to connect to exchange.
👍1
P
i am not saying that brokers are hosting the co location server in aws , i am saying that with netwrok digging i found that all of those three brokers servers are in aws , so with right netwroking setting you could beat the normal vm within the AWS
M
Didn't needed websocket for that system
A
Has anyone here used Arrow Trade (https://www.arrow.trade/) for active trading? They claim to be one of India's FASTEST BROKER and say that 95% of their orders are executed under 2.9 microseconds. They also say that they are making available to the retail clients, the same Technology which is being used by Professional Proprietary Trading Firms for getting super fast trade execution, that too at the same price of any other retail broker. I’d appreciate your feedback, particularly on order execution speed, slippage, tick frequency (ticks per second), real-time data feed latency, and overall feed quality during high-volatility market conditions. Does it really offer any practical advantage over zerodha ? Any practical views would be really helpful. Thank you. PS: One interesting point about Arrow Trade is its connection with iRage, an established Indian proprietary and high-frequency algorithmic trading firm with a background in developing low-latency trading systems. iRage’s ecosystem also includes QuantInsti, the algorithmic and quantitative trading education company founded by iRage’s partners. Furthermore, Vivek Bajaj, known for his work in trading education and StockEdge, is listed as a partner at iRage Broking Services LLP, the entity operating Arrow Trade. Given this team’s background in quantitative trading and trading technology, Arrow Trade might be offering some advantages over the normal brokers. Therefor it would be great to hear from traders who have tested it in live markets. I am in no way linked to them, my query is simply from the point of view of retail trader.
A
There is someone who is selling L1 tick data, saying he bought it from the vendor. How can I validate it? Is there any way to verify the data and integrity? Thanks.
  1. G
    no way to verify it. it can be broker websocket collected data and can be dummy data too.
    1. A
      I see, isn't there any way like any ai or something to validate it
Whole thread · 2 replies →

An open public feed from the search index ChatCrawler — “Google for public Telegram”; refreshed as the venue is crawled. Times are UTC.

Public content only, official Telegram API. About · FAQ · Index limits · Remove a page · Catalog · Search · How we count