Which Forex Pair Trends the Most - 2023 Data
The present guide provides an updated review of the most trending currency pairs in 2023. Additionally, it also provides a script that you can use to calculate trend statistics for any set of trading instruments and timeframes.
Measuring a trendedness of a currency pair (or any other trading instrument) is always a challenge. It is accentuated by the problem of this trendedness changing over time. A currency pair might be trending strongly one year and be completely trendless the next year. Still, it is possible (and important if you trade the trend) to compare the trendedness of currency pairs based on a set of metrics to get a better understanding of which currency pairs trend the most and also how exactly they trend.
The post below analyzes 10 currency pairs based on five metrics. It explains how these metrics work and why they can serve as a rough proxy of a pair's trendedness.
For the analysis, we chose 10 currency pairs that meet three conditions: they are very liquid (according to the 2022 Triennial Central Bank Survey), they have low spreads, and they are readily available at retail Forex brokers. For example, the rather liquid USD/CNY currency pair is omitted (which is the sixth most liquid in the world) because it is available only at a few brokers, its spreads are high, and trading is severely restricted by the People's Bank of China. Instead, we will look at the following currency pairs for this study (presented in alphabetical order):
We use the following methods to assess the trendedness of the currency pairs:
- Mean and median rate of change.
- Mean and median volatility.
- Average and median number of consecutive closes above/below 50-period simple and exponential moving averages.
- Mean number of consecutive Higher High + Higher Low or Lower Low + Low High occurrences.
- Mean number of consecutive bullish or consecutive bearish candles.
Rate of change is calculated as previous the Close minus current the Close and divided by the previous Close to get the percentage value. Obviously, this is a crude method of analysis. However, it can give us some hints on the pairs that trend often.
A currency pair’s volatility is calculated as a candle's High minus Low divided by its Open. It is calculated in percentage points too.
The above calculation would be only a starting point. To identify the best of the trending currency pairs, we need to calculate precisely the number of periods a pair had been in a trend during some span of time. We need a dependable indicator to identify trends in three different timeframes. We use the moving average for that purpose. We calculate the mean and median number of consecutive closes above/below the moving average. By ranking the average of the number of closes above/below a moving average, we can get additional insights regarding how trending the pairs are. Beginners are often advised to use an exponential moving average instead of a simple one as the former lags less (i.e., it follows a trend more quickly). We verify that as well by applying the calculations to both simple and exponential 50-period moving averages.
Consecutive Higher High + Higher Low or Lower Low + Lower High show exactly that — the streaks of bars that are formed according to one of the most popular definitions of trend.
Consecutive bullish and bearish candles show how likely a bearish candle is to be followed by other bearish candles, and the same for bullish ones.
All calculations are repeated on three timeframes: daily, weekly, and monthly. All currency pairs are analyzed using the data of 5 years back from March 5, 2023. The data is derived from MetaTrader 5 platform with a server in GMT+2 time zone, which means that the weekly session goes from Monday 00:00 to Friday 23:59.
Rate of change
We can calculate the absolute change in the exchange rate of a currency pair in a given period (day, week, month) using the following formula:
where N is the total number of periods.
The median rate of change is calculated by sorting the individual rates of change (Tn) and either picking the middle one (for an odd number) or calculating the mean of the two middle-most rates of change.
We have to use the percentage values because the direct (pips) rate of change would differ significantly among currency pairs as their exchange rates are not comparable.
The table provides the 5-year mean and median percentage (%) rate of change values for the studied currency pairs for three timeframes from March 6, 2018, until March 5, 2023.
The table above shows how mean and median changes (per day, per week, and per month) differ among currency pairs. The first noticeable thing is that they don't vary by a lot, at least on the daily and weekly timeframes. The situation is different on the monthly timeframe, in particular, if we'll look at the median values. The rate of change of the most active currency pairs (AUD/USD and NZD/USD) is about three times bigger than that of the slowest currency pair (EUR/GBP). Let's look at the charts below to better analyze the differences among the currency pairs' average change for the period.
Unlike in the previous year, there was a clear winner across almost all timeframes this year — NZD/USD. It lost only on the daily timeframe, with its mean change of 0.48% being just a percentage point below 0.49% demonstrated by AUD/USD. AUD/USD was also showing a significant rate of change on all other timeframes, usually trailing close to NZD/USD. GBP/JPY was the third on all timeframes, sharing its third place with GBP/USD on the weekly timeframe as both currency pairs demonstrate a 0.85% median rate of change. The rest of the pairs fell behind, though EUR/JPY showed a bigger rate of change on all timeframes compared to other losers. EUR/GBP showed the lowest rate of change, both mean and median, on all timeframes.
The volatility of a currency pair can be calculated using the formula:
where N is the total number of periods.
The median volatility is calculated by sorting the individual volatility values (Vn) and either picking the middle one (for odd N) or calculating the mean of the two middle-most values (for even N).
The table summarizes the mean and median percentage (%) volatility values of the studied currency pairs for daily, weekly, and monthly timeframes from March 6, 2018, till March 5, 2023.
As you can see, the mean and median volatility of the studied currency pairs varies less than the rate of change. However, it also demonstrated noticeably more significant variations than in the previous analysis. Unsurprisingly, the most volatile currency pairs were the same as the most trending pair as measured above using the mean rate of change. Below, you can find six charts that illustrate and help to compare the differences in volatility for the studied currency pairs.
AUD/USD and NZD/USD were fighting for the first place among the most volatile currency pairs. NZD/USD was the most volatile on the daily timeframe as well as by the mean measure on the weekly timeframe and the median measure on the monthly timeframe. AUD/USD beat its rival by the median estimate on the weekly timeframe and the mean measure on the monthly timeframe. GBP/USD was the third on all timeframes, except for the monthly median measure, which exceeded all other currency pairs. All other pairs were close to each other in terms of volatility. Even EUR/GBP did not fall as far behind as in the Rate of Change measurements, and even was not the last on most timeframes.
Consecutive closing above/below moving average
One of the most intuitive methods to detect Forex trends is to use a moving average. We calculate the mean and median number of consecutive closes above and below a 50-period (daily, weekly, and monthly) moving average (both simple and exponential).
The analysis made in 2021 showed that consecutive closes below or above moving averages seemed to result in more significant differences between the pairs compared with the previous measures of trendedness. But the situation was completely different this year as all measurements were rather close to each other. The comparison of other pairs is well illustrated by the charts below.
Unlike in the previous measures, there were no clear winners. According to the mean estimate, GBP/USD demonstrated the biggest number of consecutive days above/below SMA with EUR/USD being close second and NZD/USD the third. But if we look at the mean consecutive daily closes above/below EMA, EUR/USD show the highest number while GBP/USD and NZD/USD are sharing the second place with the result of 12.2. If we turn to median measures, EUR/JPY showed the highest result when looking at SMA, while GBP/JPY, GBP/USD, and USD/CHF were slightly below it, with the same result across all three currency pairs. But if we look at the median EMA measure, EUR/USD had the highest number of consecutive days while NZD/USD and USD/JPY were the second. USD/CAD demonstrated the lowest median number of consecutive days above/below EMA, even though this pair was never the last one on other measures.
The table for the consecutive weeks above and below the moving averages is presented below.
These charts show a clear winner, though it is different depending on whether we look at the SMA or the EMA measures. EUR/USD demonstrated the highest number of consecutive weeks above/below SMA, while AUD/USD had the second highest. But AUD/USD showed the highest number of consecutive weeks above/below EMA.
The table for the monthly data is presented below. Unfortunately, it doesn't offer reliable information because 5 years contain not so many monthly candles to work with.
GBP/JPY and NZD/USD had the highest number of consecutive months above/below SMA. While it was largely true for EMA as well, EUR/USD managed to beat them in that measurement.
Higher High + Higher Low and Lower Low + Lower High
A currency pair is generally believed to be trending if it forms consecutive Higher Highs with Higher Lows (HHHL) in an uptrend or Lower Low with Lower Highs (LLLH) in a downtrend. We calculate the mean number of HHHL and LLLH patterns for each currency pair on the daily, weekly, and monthly timeframes.
|Currency pair||Mean length of HHHL or LLLH streak|
In the charts below, you can see the illustration of the data presented in the table above. The results were all over the place in this measurement. EUR/USD demonstrated the biggest amount of consecutive HHHL/LLLH days, and EUR/JPY was the close second, trailing by just a few days. USD/CAD, meanwhile, was the least trending pair in this measurement. But things change dramatically if we look at the weekly timeframe. In this timeframe, USD/CAD came third, with NZD/USD taking the second place, and GBP/USD demonstrating the largest amount of consecutive HHHL/LLLH weeks. And on the monthly timeframe, all currency pairs were close to each other, with EUR/USD getting slightly ahead of the rest and GBP/JPY lagging behind a bit. The only consistent feature among all the timeframes was AUD/USD demonstrating a relatively large amount of consecutive higher highs and lower lows.
Consecutive bullish/bearish candles
A blunter way to measure trends is to record the average number of consecutive bullish or bearish candles. It ignores the Higher High + High Low and Lower Low + Lower High condition outlined in the previous analysis but still captures useful information about the currency pair's trendedness. The following table contains the mean values of consecutive bullish/bearish candles for each currency pair on the daily, weekly, and monthly timeframes.
|Currency pair||Mean consecutive bullish/bearish candles|
On the daily timeframe, GBP/JPY and NZD/USD went ahead of the rest, though EUR/JPY, GBP/USD, and USD/CHF were not that far behind. In general, the rest of the currency pairs were not slacking either, with the exception of USD/CAD, which had a peculiarly low amount of consecutive bullish or bearish candles. The weekly timeframe did not favor any one currency pair, as all contenders were very close.
Our research has revealed the following facts based on the studied period of 5 years:
- NZD/USD, GBP/JPY, and AUD/USD have the largest expected rate of change for any of the studied periods — day, week, and month. These should be your currency pairs of choice if your trading strategy involves opening a trade and holding it for some fixed period of time.
- NZD/USD, GBP/JPY, and AUD/USD are also the most volatile pairs. It means that an average candle on these pairs' charts is likely to be longer than on the charts of other currency pairs. This can be used to capture large movements (spikes) with well-placed take-profit orders. This conclusion (along with the one above) also seems to be very reliable as the currency pairs lead not only by mean but also the median values.
- EUR/USD, GBP/USD, and NZD/USD are the best-trending pairs when measuring trends with a moving average on a daily timeframe. However, the results were too different depending on the method involved in the measuring, making it hard to choose the most trending pairs.
- AUD/USD and EUR/USD enter longer trends on average on a weekly timeframe.
- Monthly timeframe comparison to moving averages is highly unreliable, so there is little point in trying to get any insights from EUR/USD, GBP/JPY, and NZD/USD dominance there.
- The data on consecutive Higher High + Higher Low or Lower Low + Lower High is very mixed with no clear leaders (especially on monthly timeframe).
- Consecutive bullish/bearish candles data is also very mixed but suggests that trading in GBP/USD, EUR/JPY, NZD/USD, EUR/GBP, and GBP/JPY on a daily timeframe can be more lucrative if your strategy relies on candles repeating their color.
- The higher volatility on GBP/JPY, NZD/USD, and AUD/USD also warrants a wider stop-loss for your trades.
- The low median number of days above/below a moving average for most currency pairs (just 3 or 4 days, with only EUR/USD reaching 5) suggests that the basic moving average crossover strategy is ineffective with most trading instruments. Whether expecting the high median values to hold for pairs that had them high during the previous 5 years is a good idea is another (unanswered) question.
- If you were to answer the question of what the most trending currency pair is based on all the data in this guide, it would make sense to say that it is either AUD/USD or NZD/USD. However, it should be noted that the latter usually involves lower spreads.
Important note: Past performance is not an indication of future performance. This means that it might be impractical to base actual trading on expectations of the trending behavior to remain the same as they were during the studied period.
And now to the most important stuff — a MetaTrader script that can be used to get the same data that is presented in this guide and even more. TrendStats script consists of two files that should be unarchived into the same subfolder inside your /MQL4/Scripts/ folder (/MQL5/Scripts/ for MetaTrader 5). You need to compile TrendStats.mq4 (for MetaTrader 4) or TrendStats.mq5 (for MetaTrader 5); TrendsStats.mqh is an included file used by TrendStats.mq4 and TrendStats.mq5.
The script, when run on any chart, will analyze a list of currency pairs (given via input parameters) on a range of timeframes (also given via input parameters) and on a given time period (also changeable via input parameters). It will produce .csv files with output results and will also output the results in the Experts tab of the terminal. Here is the list of input parameters for the script:
- Symbols — a list of currency pairs and other trading instruments you want to analyze. Enter them as they are listed in your Market Watch window. You can use space, comma, or semicolon to separate them.
- Timeframes — a list of timeframes to process. Enter them as
PERIOD_H4, and so on. Same as with Symbols, recognized separators are space, comma, and semicolon.
- PeriodToProcess — a period to process by the script. It is a choice of either
Last_5_Years(same as was used in this guide),
Time_Period(you then set the exact start and finish dates via StartDate and FinishDate inputs), and Last_N_Candles (you then set the exact number of candles to process via N input parameter).
- StartDate — ignored unless
Time_Periodis selected in PeriodToProcess.
- FinishDate — ignored unless
Time_Periodis selected in PeriodToProcess.
- N — ignored unless
Last_N_Candlesis selected in PeriodToProcess.
- Time_Shift — you can set the time shift in hours to move the start of the date. This is useful if your broker has an unconventional time zone. For example, if your server is UTC-7 and you want the day to start exactly at 00:00 UTC, you set this parameter to 7. Please note that setting non-zero Time_Shift will make the script calculate everything using H1 data only — it will be converted to other timeframes you request via the Timeframes parameter, but there might be not enough H1 candles to generate enough high-timeframe data.
- MA_Period — a moving average period for moving average comparison stats.
- FileNamePrefix — a prefix for .csv file names.
- SilentMode — if true, the silent mode will prevent the script from outputting any calculation data into the Experts tab of the terminal. Service and error messages will still be printed.
If you have some questions about this study of the major currency pairs' trendedness, if you want to suggest some other measures of trendedness to analyze, or if you find some bugs in the TrendStats script, please proceed to our Forex forum.