Average True Range

A big change in my trading consists of using the average true range (ATR) to set take profit targets. This is based on the assumption that ATR, a measure of volatility, will show the likely movement potential. In other words, a take profit of 1,000 pips will unlikely be hit in the near future if the 10-day ATR is only 30 pips.

However, I think using ATR-based take profit is also somewhat flawed. This draws from recent events and past experiences back in my banking days.

Currently, I do my scans on the weekly, daily, and then hourly timeframes. Once I establish a trend-continuation signal, I would enter a trade with a take profit set at 2x ATR, meaning I would set a 60 pips take profit if the five-day ATR is 30 pips. 

This is based on the interpretation that this particular pair is capable of a daily movement of just 30 pips. If we observe a 60-pips move, that's a positive deviation from the mean and it's better to lock in that profit rather than let price revert back and continue holding in hopes that price would break new highs or lows.

There's a limitation to this approach, and it's that it doesn't account for the fact that volatility can change. If the five-day ATR surges to 150 pips, it doesn't necessarily mean that this pair will observe an 150 pips movement the next day.

On the other hand, a pair stuck in a range will have a much lower ATR. Once it breaks out of that range, you shouldn't really be setting a profit target based on the volatility in that range. It's very possible price trades in a 20-pips range and then breaks out 150 pips in the course of the next three days, especially with the Darvas box trend trading approach.

The topic of volatility itself draws on my experience performing Black Scholes warrant valuations back in my banking days. I was instructed to perform these valuations using Bloomberg's HVG function - historical volatility.

My speculation is because historical volatility can be easily observed and calculated. You can use the ATR indicator, Bloomberg's HVG function, or some other proprietary tool out there that's beyond my knowledge. While it's possible to use implied volatility, a forward-looking metric, it's not as easily observed and requires making possibly subjective assumptions.

With this in mind, I'm making yet another small but important tweak in my market scanning process. What ended up inadvertently happening with using 2x ATR profit targets is that I started tightening up my stop loss in order to maintain the reward to risk ratio. This was problematic because it provides less breathing room for the trades to play out and market spikes could stop me out.

Instead, profits should be trailed, especially when strong trends are formed. The use of ATR is more efficient with stop loss placement (risk management). In other words, setting a 20 pips stop loss when the daily ATR for the past five days is 100 pip means a high probability of getting pushed out of this trade.

Since we're not dealing with implied volatility, it's not as important to track volatility changes. If the volatility picks up, the daily ATR stat will slowly increase, which won't pose any serious risks to a stop loss that provides breathing room for the trade. On the other hand, a decreasing ATR doesn't pose any risk concerns either.

In essence, this approach helps optimize risks without limiting profit potential of trending trades.