top of page
Search

From Manual Charts to Machine Intelligence: How Programming Transformed Technical Analysis Since the 1990s

  • Writer: ADARSH KUMAR MALPOTRA
    ADARSH KUMAR MALPOTRA
  • May 18
  • 3 min read

Updated: May 19


When I look back at my original 1996–98 project — “Technical Analysis of Select 35 Scrips with EPS > 10 and P/E < 10” — I am reminded of a very different world of computing. A world where:

  • Stock data was downloaded from NSE/BSE in text files

  • Charts were drawn manually or using Lotus 1-2-3 / early Excel

  • Technical indicators were coded using BASIC, FoxPro, or early C

  • Internet speeds were measured in kbps, not Mbps

  • And “automation” meant writing a macro that didn’t crash

Yet, despite these limitations, programming was already deeply embedded in financial analysis. The seeds of today’s AI‑driven analytics were being planted quietly in the background.

This blog is a reflection on that journey — from the early days of technical analysis programming to the radically simplified, AI‑powered world we live in today.

 

The 1990s: When Programming Was a Craft, Not a Commodity

In the mid‑90s, technical analysis required:

1. Manual Data Collection

Daily price and volume data were sourced from:

  • Economic Times

  • Capital Market magazine

  • NSE/BSE bulletins

  • Teletext feeds

For my project, I manually compiled data for 35 scrips such as Blue Star, Escorts, Jindal Strips, Reliance, Nahar Spinning, etc.A typical line from the project reads:

“The technical analysis shall concentrate on trend determining techniques primarily price patterns, trendlines, moving averages, momentum. The volume of the scrips traded shall be concurrently analysed.”

This meant writing formulas by hand, calculating moving averages manually, and plotting trendlines on printed charts.

2. Early Programming Tools

The most common tools were:

  • GW‑BASIC / QBasic

  • FoxPro for database‑style analysis

  • Lotus 1‑2‑3 macros

  • Excel 5.0 with limited VBA

  • C/C++ for those who wanted speed

Even a simple 14‑day RSI required 20–30 lines of code.

3. No APIs, No Cloud, No Automation

Everything was local.Everything was manual.Everything was slow.

But it worked — because the logic of technical analysis is timeless.

 

The 2000s: The Rise of Retail Trading Software

With the arrival of:

  • MetaStock

  • Amibroker

  • TradeStation

  • Bloomberg terminals becoming more accessible

Programming shifted from “writing code” to “writing formulas.”

A trader could now:

  • Backtest strategies

  • Run screeners

  • Automate charting

  • Use built‑in indicators

This decade democratized technical analysis.

 

The 2010s: Python Eats the World

Python changed everything.

Suddenly, anyone could:

  • Pull data from Yahoo Finance

  • Run TA‑Lib indicators

  • Build backtesting engines

  • Use Pandas for time‑series analysis

  • Deploy machine learning models

Technical analysis moved from: Charts → Algorithms → Predictive Models

The 2010s made programming powerful.

 

The 2020s: AI Makes Programming Invisible

Today, the landscape is unrecognizable.

You can:

  • Ask an AI to generate a trading strategy

  • Pull 20 years of data with one line of code

  • Run Monte Carlo simulations in seconds

  • Build dashboards without writing a line of code

  • Use no‑code platforms like Wix, Lovable, Emergent, Framer AI to publish insights instantly

Programming has shifted from:

1990s → “Write everything manually”

2000s → “Use formulas and software”

2010s → “Write Python for automation”

2020s → “Describe what you want; AI writes the code”

The complexity has disappeared.The capability has exploded.

Why This Evolution Matters for Finance Professionals

1. Technical analysis is no longer a specialist skill

Anyone can generate:

  • Trendlines

  • RSI

  • MACD

  • Volume studies

  • Pattern recognition

…with a single click.

2. The edge has shifted from “coding indicators” to “interpreting signals”

Human judgment is now the differentiator.

3. AI allows deeper, faster, broader analysis

What took me months in 1996 can be done in seconds today.

A Personal Reflection

When I wrote:

“The study of historical price and volume patterns provides clues as to which outcome follows which pattern.”

…I had no idea that one day AI would:

  • Detect patterns automatically

  • Backtest them instantly

  • Optimize them continuously

  • And explain them in natural language

The essence of technical analysis remains the same.But the tools have evolved beyond imagination.


 
 
 

Comments

bottom of page