On building a profitable trading strategy

On building a profitable trading strategy

After lots of thinking, backtesting and hearing from traders, I am convinced that not every trade can be profitable and market direction can not be predicted better than a coin toss probability. Any simple strategy should work in markets as long as some fundamental concepts are taken care of. All these talk about trading psychology feels good to hear, but it is of no use if it does not produce something which has real potential to earn money. Thus my quest on building a profitable trading strategy began.

Choosing the stock & direction

Before I started building a profitable trading strategy, I had following constraints:

  • Strategy should be able to be deployed on the cash market before I move to FnO, so that I do not lose enough capital in the testing phase.
  • Underlying used should be highly liquid

Keeping the above two constraints in mind, I started building a positional strategy. Since the cash market does not allow going Short, I have to build a Long only strategy. In a way, this simplified the process as I do not have to worry about finding short opportunities, and I can focus on one thing at a time. For highly liquid underlying, I decided to go with Reliance. No science here, just I like the stock. Traders can choose a different stock based on their convictions. Obviously backtesting helps solidify their convictions.

Determining the risk

Now I know that I have to bet on Reliance going up. The next step is determining the risk I can take on any trade. As I know that at least 50% of trades will be in loss, I need to know how much I can lose on each trade. This is the risk I am willing to take and I should be comfortable with that risk, else I would not be able to accept it when the time comes.

Risk is another word for stop loss. Basically if I bet that Reliance will go up and initiate a trade, Reliance may not immediately start going up. It happens many times that it goes down a little and then starts moving up. So I need to keep a stop loss on my assumption of how much low Reliance can go. If my stop loss is too wide, I will lose a lot on a single trade. If my stop loss is too narrow, I may mis slot of good rallies because of whipsaws. 

I observed that 2% movement in stocks like Reliance is pretty common. So any stop loss below 2% can be narrow. I decided to go with 4% stop loss , so that I am giving enough room for trade to unfold. I was not comfortable with any stop loss wider than 4% as the risk per trade would have increased. So for example, if I go long on Reliance at 2000, my stop loss will be 4% of 2000, which is 80 points. If Reliance falls to 1920, I will exit the trade.

There is no science to choosing a stop loss. Every trader may have a different feeling for it and they may use a different stop loss as per their risk appetite and strategy. Obviously, backtesting it on past data helps get conviction if the feelings are in the right direction or not.

So 3 things are finalised till now:

  • Reliance
  • Going Long
  • Stop loss of 4%

Determining the reward

I have the risk decided. Now I need to know the reward to complete the risk reward equation. There are two ways to exit a trade in profit:

  • Setting a fixed target
  • Keep trailing the stop loss and exit only when stop loss is hit

I believe that if I set a target on my profits, I tend to miss those rare opportunities where I can get insane returns. These few opportunities can increase the account value by a lot. These opportunities also compensate for any unfortunate black swan losses. So the decision was simple for me that I will keep trailing the stop loss as position moves in my favour.

For trailing stop loss, I decided to keep the technique extremely simple. If position moves 1 point in my favor, I will trail the stop loss by 1 point. Since my initial stop loss is at 4%, my trailing stop loss will always be at ~4% from the most recent high point. For example, if I go long in Reliance at 2000, my initial stop loss is at 1920, and after a few minutes if price moves to 2010, I will move my stop loss to 1930.

Using the above trailing stop loss technique, I will lose the last 4% of uptrend everytime, but I keep myself exposed to super big trends which may happen once in a while. If I have a target set for taking profits, I will miss these super big trends. 

So following things are finalised now:

  • Going long on Reliance
  • Initial stop loss at 4% from entry price
  • Trails stop loss by 1 point, if price moves 1 point in favour

Position sizing

Once I have determined the risk taken on a trade, I can easily determine the quantity to trade. The quantity is decided such that one does not lose say 1-2% on the capital. Position sizing in itself deserves a separate post, so I will take it in upcoming ones. 

For this strategy, I will keep it simple and take 1 lot position for backtesting. 1 lot of Reliance has 505 units.

Entering a trade

I need to decide an entry criteria so that the trade can be initiated. Since I am going long, I just need to ensure that Reliance has shown indication of going up before I take the trade. I strongly believe that any indicator whether simple or complex does not perform better than a coin toss, so I will go with a very simple one. 

If Reliance has broken it’s 20 day high, I will initiate the long trade. Experienced traders might also know it as Donchian Channel (DC20) . Breaking 20 day high shows that Reliance has started to move up. Again there is no science to it, I just needed some indicator to enter, so I chose a 20 day high criteria. A different indicator should work equally well. Backtesting different criterias helps get a strong conviction on what works and what does not.

Backtest results

I have a strategy ready using some basic rules of trading psychology. It is time to backtest it and verify if it actually worked on the past data or not. To summarize, the strategy rules are as follows:

  1. Long only trade on Reliance
  2. Go long when Reliance crosses last 20 day high
  3. Keep initial stop loss at 4% from entry price
  4. Trail stop loss by 1 point if price moves up by 1 point
  5. Play 1 lot -> 505 units

I backtested the above strategy on Reliance spot data for a 4 year period starting from 1 Jan 2016 till 31 Dec 2019. Below are the results I got:

PnL chart for the profitable trading strategy

Net returns: Rs. 2,31,664    (Includes slippage & charges)

Total trades: 32

Number of profits: 18

Number of losses: 14

Win rate : 55%

Max profit streak: 4

Max loss streak: 4

Max drawdown: Rs. -71,769

Top 3 profits: [119697, 53917, 42905]

Top 3 losses: [-31381, -24872, -23093]

Average return per trade: 7239

Sharpe ratio: 3.93

Trade logs

One of the most important aspects of understanding a strategy is studying its trade logs. This helps in understanding each trade and how the strategy has performed during different market cycles. By cross verifying each trade from the chart, one gets a strong conviction and identifies opportunities to make the strategy better.

The charges for trade logs and month wise PnL of the backtest are just 10 Nifty points (Rs. 750) Click here to get it.

Is it this simple ?

Yes.

A profitable trading strategy can be built using basic trading elements like risk, reward, position sizing and a simple entry criteria. A 50% win rate is super good for long term profitability. A long only strategy can also work, so one need not keep predicting direction each time. Traders need to keep their psychology in line with what trading demands. Losses and drawdowns will be there, but if fundamentals are right, and some luck favors, profits will come.

Don’t miss the trading logs & month wise PnL. Get your copy by clicking here.

This is the third post of the series “Reminiscence of a retail trader”. Read the introduction post by clicking here.

Follow me on twitter and subscribe to my blog to get all the latest updates on upcoming posts of this series.

Disclaimer : https://optionji.com/disclaimer/

Leave a Reply

Your email address will not be published. Required fields are marked *