Is trading at night more profitable?

Is trading at night more profitable?
4 (80%) 5 votes

It has long been held that the most profitable traders trade at night – or during the Asian session.

The explanation for this is that markets are less volatile at night and so traders are less likely to have their stops run by sudden spikes in price.

Another reason is that the simple momentum strategies which most traders use tend to be more successful at night, because prices are less likely to trend strongly and more likely to respect support and resistance levels.

In an experiment I decided to put this theory to the test – was it indeed more profitable for a trader to operate at night?
A simple RSI strategy, using RSI on a 14 period setting with standard levels, on a 5 minute chart of Eurodollar was tested for effectiveness, to see if signals generated during the night were superior to those in the day.

Buy signals were generated when the indicator passed from oversold to neutral and a sell signals when it passed from overbought to neutral. A further bar making a new high or low was then required as confirmation before entering a trade order.

‘Night’ was defined as between 9pm and 6am GMT and ‘day’ between 6.01am and 8.59pm. Two different tests were run. The first was a sequential back-test from the first day of the experiment in May 2012 to March. The second was a random back-test using set-ups generated on days selected at random from a period over the last 10 years.

The results are shown in the tables in the index however a summary is outlined just below.

The data suggests that there is no advantage to trading a simple RSI strategy at night.

If, for example, we traded the strategy with a 6 pip profit target and a stop just above the extreme lows and highs of the oversold and overbought zone then the results show that there is no significant advantage to trading at night.

These results cover all the set-ups which were tested – both randomly and sequentially back-tested, which means there is obviously a greater weighting to more recent data:
Table 1a:

RSI with 6 pip target

Winners Losers Winning pips Losing pips Net Pips
Night 81 51 486 422 64
Day 94 29 564 434 130

The results show that the strategy would actually have resulted in a higher win rate if traded during the day – not the night as originally hypothesized!

However, to be more certain we could try using a different target – maybe increasing it to 10pips. This would result in the following results in
Table 1b:

RSI with 10 pip target

Winners Losers Winning pips Losing pips Net Pips
Night 65 67 585 586 -1
Day 75 48 675 513 162

The results gained from using a longer 10 pip target show even more bias to the day time as a better time to trade, with 162 winning pips vs -1 for the night.

Nevertheless, there is still the possibility that recent data is not reflective of the strategy’s longer-term success at night and it could be unduly biasing the results.

One way to deal with the bias would be to look at the random data from the over the last 10 years on its own. Tables 2a and b show the results for 6 and 10 pip strategies using only the randomly selected data sample:

6pips Winners Losers Winning pips Losing pips Net Pips
Night 50 27 300 261 39
Day 45 17 270 268 2
Night 41 36 410 365 45
Day 37 25 370 347 23

These results are interesting because they show that in both cases the night actually outperformed the day in the random samples. This would seem to back up the original assertion that trading at night is more profitable.

The biggest difference between the day and night results was with the 6 pip strategy where the night earned 37 more pips than the day, however is this a significant enough difference?

Using Binomial Statistical testing methods the result was tested for significance. It was found that in this case the difference was not great enough to prove statistical significance according to normal scientific research standards – although it was very close.

Overall the Binomial test showed that the chances of the difference being put simply down to ‘chance’ was only 6% although this was still 1% above the 5% or less required for establishing significance.

We can conclude, therefore, that the results were mixed overall and proved neither that an RSI strategy was more or less successful at night.

Whilst the data showed a definite bias towards greater success during daytime hours in a recent sample, the random data found no statistically significant bias either way.

Luckily you don’t have to stay up all night to make money from trading Forex!

Detailed Results

The data below lists all the trading set-ups tested.

The first column shows the date, the second shows the potential of the signal, which means the most pips that could have been earned if the trade had been allowed to run its maximum length and the final column is the ‘risk’ which means the amount which would be lost if the stop was triggered.

Triggers were placed one pip above the 5 minute bar immediately after the bar which broke out of oversold or overbought. This was as long as the second bar after the breakout rose higher or lower than the breakout bar and gave confirmation of a reversal.

Spread the love

Related Articles

Leave a Comment