Thread by @theBuoyantMan

I had a quick observation about intraday price movement in NIFTY (and probably even BankNifty indices). Days where price starts at or near the highest or lowest of the day and closes at or near the lowest or highest of day respectively are wonderful for trading.

So, I set out checking whether it's an anomaly or if it happens frequently. And, it does happen frequently. So frequently that with some effort, you can make a strategy out of it. The following are the statistics that i obtained with manual observation and calculation.

There are two types of days with respect to price of NIFTY that I am interested in.
1. Starts at or near the highest of the day, and closes at or near the lowest of the day.
2. Starts at or near the lowest of the day, and closes at or near the highest of the day.

I chose to ignore candles that had long wicks even if they'd covered good range in the direction and closed towards the highest or lowest point of the day from the opposite direction. I chose to ignore very small range candles too that fulfilled the 2 conditions mentioned above.

A year usually has around 240-255 trading days, depending upon the country and holidays. We'll just keep 250 as the count by default. The following are the statistics that came up from my observation calculating each candle that fulfilled the conditions manually.

(1) will represent price starting at or near High of Day (HOD) and closing at or near Low of Day (LOD)
(2) will represent price starting at or near LOD and closing at or near HOD.

2020: 37 days fulfilled (1) ; 25 days fulfilled (2) (till date)

2019: 57 days fulfilled (1) ; 39 days fulfilled (2)
2018: 80 days fulfilled (1) ; 62 days fulfilled (2)
2017: 65 days fulfilled (1) ; 47 days fulfilled (2)
2016: 74 days fulfilled (1) ; 58 days fulfilled (2)
2015: 56 days fulfilled (1) ; 45 days fulfilled (2)

If you consider 250 trading days a year
2020: 62 days out of 125 days.
2019: 96 days out of 250 days.
2018: 142 days out of 250 days.
2017: 112 days out of 250 days.
2016: 132 days out of 250 days.
2015: 101 days out of 250 days.

Calculating the % number of days fulfilling these conditions in each year, we arrive somewhere around 55% on average for years since 2015. You calculate the % for each year and average based on number of years if you do the calculations yourself for other prior years too.

This means, if you are able to identify which direction the market would trend and latch on to Nifty futures and hold till end of day, you would win 55% of times. You would essentially need a stop loss system in place that loses atleast 1/3rd < what you make in those 55% trades.

If you can come up with such a strategy and put on the trade early in the morning identifying the trend for the day, you'll either get stopped out or trail till end of day. This also means there should be only 1 trade a day. If the trade is off, it's done for the day.

This is something to think about. There are several strategies that could help you accomplish this, a lot of them freely available for you to find on the web. Now it's your turn to put in the work, manually observe on charts, and go through strategies and see which could work.

Once you arrive at a strategy that helps you accomplish this, test it from 2008 with the data available. Thoroughly test it, get the win rate, R:R, maxdd and other parameters of the system and then go at it with discipline in trading live. And once it's done, get out of your way.

For clarity, here's a visual.
The rightmost candle started at the highest of the day and closed near the lowest of the day.
The 8th candle from left started at the lowest of the day and closed near the highest of the day.
These are all daily candles.