Golden RatioResearch Community
Join free community
Educational content only. Nothing on this page or in the community is investment advice, a trading signal, or a recommendation to buy or sell any security. We are not a SEBI-registered investment adviser.
Free · WhatsApp · weekly research session

A community for people who'd rather test an idea than trust one.

We study how algorithmic trading research is actually built — the programming, the historical data, the backtesting methodology, the statistics that separate a real result from a lucky one. No calls, no tips, no signals — ever.

Free forever · Education only · Leave anytime
session_04_crossover_study.py
# illustrative example — historical data, for study only
def signal(fast, slow):
  return fast.rolling(20).mean() > slow.rolling(50).mean()
# evaluated on 3 yrs of sample index data
report = backtest(signal, data, costs=True)
Sharpe (sample)
0.71
Max drawdown
−18.4%
Trades
212
Illustrative backtest output on historical data. Hypothetical, for teaching evaluation methodology — not a projection, promise, or record of real results.
What's inside

Four things, kept free.

Everything below already lives in the group — pinned, organised, and added to every week.

Every week

Live research sessions

A walkthrough each week — methodology, code on screen, open Q&A after. Recordings go into the group for anyone who misses it.

Step by step

The research guide

From a blank environment to a fully evaluated strategy: handling data, coding rules, and the checks that separate a real result from a lucky one.

Free tool

A research backtester

Run your own rules against historical data and read the standard evaluation stats — before you take any conclusion further.

Daily

A room that answers

Stuck on a bug, a data quirk, or a methodology question? Post it. This is peer and practitioner discussion, not a broadcast channel — and not licensed advisory.

Why "Golden Ratio"

Structure over impulse.

A rule, written down, coded, and checked against real history, holds up better than a hunch on a Tuesday afternoon. That's the whole idea behind this community: understand the method before you trust the output.

φ
1.618 — the ratio we're named for
The path

Five steps, in order.

A real sequence — each step depends on the one before it. Most members work through it over a few weeks, at their own pace.

1

Set up a research environment

Python, an editor, and a source of historical market data. We walk through it end to end on session one.

2

Learn to read the data like a researcher

Cleaning, structuring, and the basic statistics you need before any rule is worth testing.

3

Turn an idea into a testable rule

Take a plain-English hunch — "a fast average crossing a slow one" — and write it as logic a program can evaluate.

4

Backtest it, and learn to distrust it

Use the free tool, then learn to spot lookahead bias, overfitting, and survivorship bias — the three ways a backtest lies to you.

5

Keep a research log

Write down what the evidence actually showed, what didn't hold up, and why. The log — not the first good-looking result — is the point.

Questions

Before you join.

Is this investment advice?−

No. Nothing shared in this community — live sessions, the guide, the backtester, or discussion in the group — is a recommendation to buy, sell, or hold any security. It's a space to learn the programming and research methods behind algorithmic trading. We are not a SEBI-registered investment adviser, and nothing here should be treated as personalised advice.

Will I get trading calls, tips, or signals?+

No, and we won't post any — not in the group, not in a DM, not as a "bonus." If someone claims to represent this community while sending calls or signals, they don't.

Do I need to know how to code?+

No. The guide starts from a blank environment. Comfort with basic Python helps but isn't required going in — most members pick it up alongside the material.

What data does the material use?+

Historical, publicly available market data used purely for study and backtesting exercises. Examples reference index or instrument data the way a textbook would — never as a recommendation to trade that instrument.

Who runs the sessions?+

Practitioners and enthusiasts sharing what they've learned building and testing their own research. This is peer education, not a licensed advisory service.

How much time does it take each week?+

The live session runs about an hour. Members who put in two or three focused hours a week tend to move through the guide comfortably.

One tap

Join the Golden Ratio research community.

Free education, a weekly live session, the full research guide, and the backtester — waiting in the group chat.

Join on WhatsApp
Your number stays visible only to group members, as with any WhatsApp group.