Win rate isn't everything: choosing the right trading system
Three systems with 90%, 10% and 50% win rates show why expectancy, losing streaks and time drawdown matter more than accuracy alone.
Education only, not investment advice. Figures here come from past data and backtests: illustrative, not live market data and not a performance claim.
Earlier, many traders traded on discretion, without any rules-based approach. Now a lot of traders are ready to move to algo trading but are confused by the many options. When you are bombarded with tons of different trading systems, how do you choose the best one?
Win rate versus risk reward
Many traders still think winning accuracy is the most important parameter in a trading system. But is it really true? Let's check this example. Consider three different trading systems: System A, System B and System C.
System A has 90% winning accuracy, System B has 10% winning accuracy and System C has 50% accuracy. When given a choice, most traders will opt for System A, as it has the highest winning accuracy of all at 90%.
But System A is not the right choice. Its average profit in a winning trade is Rs.1000, so 90% of the time System A is capable of making a Rs.1000 profit. But the remaining 10% of the time, when the system makes a loss, it loses a lot. So after 100 trades, the net result is -10,000.
In System B, the average profit in a winning trade is Rs.10,000, so 10% of the time System B is capable of making a Rs.10K profit, which is ten times its losing trade. The average loss in a losing trade is Rs.1,000, but 90% of the time the system will continue to lose.
It's like buying an OTM option trading at Rs.2 on expiry day. Option buyers tend to buy such strikes thinking that if there is an extreme move in the index on expiry day, this Rs.2 option can move to Rs.200, giving exponential returns. But that doesn't happen very frequently.
So after 100 trades, we will end up with a profit of Rs.10,000 with System B. In spite of losing 90% of the time, this system ended up in profit after 100 trades because of its positive expectancy. But practically, System B is hard to trade in real time.
System C has a winning accuracy of only 50%, but its average profit is 2x its average loss. When a trade goes right it makes Rs.2000, and when it goes wrong it loses only Rs.1000. So after 100 trades, System C is able to generate a profit of Rs.50,000, which is the highest of all.
If we simply focus on winning percentage alone, we might not end up with the correct trading system. Along with win %, the risk reward ratio should be given equal importance.
Plan your risk for losing streaks
Just because your backtest says that you can have 5 continuous losses, it doesn't mean that you are going to face at most 5 back to back losing trades. Statistically, any trend system which has only 40% winning accuracy can have 21 losses in a row.
Imagine facing 21 losses back to back. Will you be able to handle it? Yes, if you plan your risk management well in advance. If you are risking 1% of your capital in a trade, then even a losing streak of 21 will not do much damage to your trading capital.
So your risk per trade should be decided based on the probability of such losing streaks instead of relying solely on your backtest parameters.
Time drawdown
Many traders tend to ignore another important factor called time drawdown. That is, how many days can a system underperform? How many months can the equity curve stay under water? Imagine making zero returns for 4 months with your intraday trading system.
You will be placing trades every single day, only to regret it after 4 months when the system is still in the drawdown phase. During such periods many traders lose faith in their system and tend to abandon it, thinking it has stopped working.
The longer this time period, the more psychologically hard you might find it to trade that system. Trend following systems fall into this category: when markets are stuck in a sideways phase without a clear direction, trend systems are affected badly and may continue to underperform.
The key to surviving such a time drawdown is to trade a mix of uncorrelated strategies. When choosing a trading system, if you give importance to these three parameters, you can end up finding the right trading system for you. I cover this in more detail in this blog post.
Along with win %, the risk reward ratio should be given equal importance.
I first wrote this as a thread on X. Follow me there for new ones.