While the UK is focused on the latest Brexit plan, the GBP/USD could react sharply to the FOMC state...
Which Currency Trading Strategies Work Best in Times of High and Low Volatility?
02/18/2009 10:35 am EST
Extreme volatility across forex markets has forced substantial losses for many currency traders. Unexpectedly sharp moves have wiped out hard-earned gains for many common trading strategies that outperformed during calmer market conditions. Knowing when to switch between viable strategies has never been more important, and this article will go into the criteria used to determine which strategies are the most attractive.
In previous forex education articles, we discussed the relative merits of popular Relative Strength Index (RSI) range trading strategies as well as Moving Average Crossover momentum strategies. Our analysis showed that the Moving Average Crossover system was more profitable over the longer term, but RSI trading signals were nonetheless profitable for longer stretches of time. Given what we know about the strengths and weaknesses of these respective systems, these facts give us important details in market dynamics.
We need to find a filter that will tell us when to look to trade one type of strategy or look for another. At DailyFX we often make reference to our volatility indices, which use implied volatility levels from forex options markets to gauge overall volatility expectations across key currency pairs. Implied volatility is a key component of forex options prices, and it gives an estimate on how much a given currency will move within a specific amount of time.
It is subsequently no surprise that our volatility indices correlate strongly to profits and losses from highly sensitive range trading and momentum strategies. In fact, we see a strong correlation between our DailyFX 1-Month Volatility Index and the performance of both our range trading and momentum strategies.
The strong relationship between performance and our predictive volatility index tells us that we can likely use this volatility measure as a tool to filter our trading strategies. With the benefit of hindsight, we see that the RSI strategy severely underperformed in months where the 1-month Volatility Index reading (taken on the first trading day of the month) was high. According to a simple statistical regression, we can expect that the RSI strategy will lose money in months where our index reaches above 9%. In practice, we observe that the RSI strategy lost substantially in all months where the volatility index reached above 11%.
The link between our Moving Average Strategy and volatility is weaker, but we nonetheless see that the two tend to move in the same direction. Using the same basic statistical analysis, our model suggests that the Moving Average Strategy loses every time that the volatility index drops below 9%. Yet the strategy actually had a substantial loss in a month where the volatility index hit an elevated 13%-a clear warning that the filter is by no means infallible.
What's the Moral of the Story?
We have taken two very common trading strategies and identified the proper market conditions for both. In doing so, we have identified when these trading strategies fail. Though it is unlikely that traders follow these strategies exactly, we can use the same concept on a wide range of similar trading styles. If a strategy depends on strong price breakouts or sustained momentum to turn profits, it would likely underperform during times of exceptionally low price volatility. Range trading strategies, by comparison, would outperform in such environments.
Can We Do Better?|pagebreak|
It is clearly difficult to set an absolute number at which we will turn off our RSI strategy and use our Moving Average strategy and vice versa. The main difficulty is always that what worked in the past may not necessarily work in the future. There is likewise a particular issue with implied volatility: Looking at the absolute number tells us relatively little on what to expect through near-term trade. One thing we do with our own DailyFX+ Currency Trading Signals is compare implied volatility to its recent range. In doing so, we can gauge relative volatility expectations and trade accordingly.
After researching the topic, we decided that it was best to use rolling percentiles of implied volatility numbers. Our weekly trading outlook for DailyFX+ trading signals uses these percentiles on a per-currency basis to gauge the likelihood of range trading and breakouts.
Why Do We Like Using Rolling Implied Volatility Percentiles?
If we avoid range trading during the worst of financial market conditions, we can save ourselves from a great deal of losses through extraordinarily volatile price action. The chart below shows us the results of using the 90-day percentile on the 1-month Implied Volatility index on EUR/USD RSI trading. This is the same strategy as highlighted in our Relative Strength Index (RSI) range trading strategies article and uses an elementary RSI strategy on a 60-minute EUR/USD chart. The important thing to note in this case is that the strategy saves itself from substantial losses through the worst of breakout market conditions.
Where Can We Find This Data?
We have known of the effectiveness of filters on range trading systems for quite some time now, and we have accordingly published the above information on DailyFX.com since May 2008. Every week, we review the performance of our DailyFX+ Currency Trading Systems and identify which are likely to outperform through the near future. Given that, we transparently identify which trading signals are range, breakout, and momentum-based, and it is relatively straightforward to filter strategies based on market conditions.
By David Rodriguez, quantitative analyst, DailyFX.com
Related Articles on FOREX
Fluctuations in the value of the dollar directly impact the returns that U.S. investors see on their...
A higher currency price is a distinct disadvantage for a country wishing to sell products on the wor...
The consensus expectations for Friday’s headline non-farm payrolls data point to around 180,00...