Thread by @theBuoyantMan

The scenario in trading today is not as it was 20 years back. 20 years back systematic or algorithmic trading was almost non-existent. Today it's growing in numbers.

Throughout the last 20-30 years, traders have had to evolve to adapt to the changing technology and market.

It's clear what the next 20 years look like. By next 20 years, almost 90% of global markets will become algorithmic and systematically driven. Already we see discretionary prop funds withering away. Very few are surviving and barely scraping to get by.

How do you prepare yourselves for trading as a lifelong career?

No other go. You have to learn programming, learn math, statistics, and adapt to the moving technology goalposts. Otherwise you will face competition from those who do learn those things.

If you don't want to learn any of these, you can simply resign yourself from trading, invest the money in good businesses and earn through dividends and share price growth. That method of making a living in the market won't go out of fashion. But even there you have competition.

The edge in value investing or any form of long-term investing is psychology + quantitative analysis + qualitative analysis.

Quantitative analysis wise you can't beat those with the information advantage - the big funds, investors, etc., with state of the art tech.

These people/funds with cutting edge tech, analytics, etc., will make sure all the quantitative analysis based information is priced in, coz it will be.

Qualitative analysis is then doing all the field work, talking to management, customers, suppliers, vendors, etc.

Doing the ground work is where the edge already is, and in having strong psychology, being aware of your biases, and holding through the difficult periods, being steadfast in your conviction, etc.

All these will come into play if you go into the long-term investing way.

But these things are outside of the quantitative analysis. Even for long-term investing, to do quantitative analysis, you need to understand at least basic statistics, and to analyse many companies and screen out unfavorables, you should use programming. Otherwise it's tedious.

Slowly in the next two decades, you will see intraday game getting crowded by algos, and Indian market will get way more efficient than it is now, with way more participants, liquidity, etc.

You'll either compete with systematic traders, or you'll be pushed out of lower tf.

That's what's been happening in the last few many years. Algos have come to dominate scalping, so scalpers are slowly going out of business. Slowly, algos will come to work with the 5-15m timeframes as companies like Tower Research have started India focused mid-frequency ops.

You'll eventually have to come out of those timeframes crowded by algos and work with timeframes where there will be lesser competition - like 30m, 1h, etc. Slowly your returns will decline too. Otherwise you'll have to take leverage which SEBI won't allow you to take.

If you have to take leverage, you can only go to bank and apply for a personal loan or something else, which is dangerous. So, you'll eventually end up playing swing trading in cash or in futures. Even in those timeframes, systematic trend followers will dominate.

Discretionary trading has its place. But slowly discretionary traders will get pushed out of lower timeframes into higher timeframes. This is my guess based on the pattern I am observing in the last 20 years of evolution of stock markets.

Indian markets may not have dark pools or orderflow selling as a predominant thing right now. But already there's a lot of spoofing by institutions going on. It will eventually be algos competing with each other rather than people in the lower timeframes.

So, either adapt or perish. That's the message for today. Starting now, slowly build your math and programming skills. It's not difficult. 1-2 hours a day should help you get there. Worst case excel + vba, best case - C++ and Python with a bit of R programming should help.

The reason I don't suggest amibroker-afl above others is because it's a proprietary tool and language. No one knows when Amibroker will go out of existence. It's always better to have command over one programming language and build your own stack for backtesting & live trading.