Japanese yen (JPY) intraday seasonality overview, 2003-2010

User Rating: / 2
PoorBest 
Written by Forex Automaton   
Thursday, 17 February 2011 10:11
Article Index
Japanese yen (JPY) intraday seasonality overview, 2003-2010
Time stability
All Pages

Intra-day seasonality in JPY is researched by averaging hourly logarithmic returns, grouped systematically into temporal "bins" according to the time of the day. Two types of average, the average across years of observation for each instrument, and an average across instruments for each year, are presented. The instruments are USD/JPY and four popular crosses involving AUD, EUR, GBP and CHF. The effects look practically insignificant given the typical levels of trading costs available to a retail trader, and given the lack of time stability.

 

You can tell how unorthodoxal the idea of trading on the basis of wall clock reading rather than the (real or imagined) "direction" of the market is simply from the fact that popular trading platforms simply do not offer an option of submitting a trade order with a pre-set execution time, while submitting an order with a pre-set level of price (limit order) is a common feature of all platforms.

 

The report uses hourly data from January 1st, 2003 through January 1st, 2011.

As usual, the quantities I am going to look at are not the actual close prices. Since prices are always positive, they are trivially correlated; this feature is absent in the correlations of the so-called logarithmic returns (or logarithmic increments) which are the ratios of price levels at close to the values they had at the close of the previous hour-long time interval.

For intraday seasonality studies, I choose Central European time because it allows me to split the forex week into five full non-interupted trading days.

Seasonality effects in first-order statistics (means) are straightforward to utilize in trading, therefore one could argue that such effects are not likely to be found. Nevertheless, sometimes such effects are visible and significant. To look for them, I average logarithmic returns for each hour of the day separately using profile histograms. The resulting histograms are plotted in Fig.1.

Average log return as a function of the hour of the day, averaged for 2003-2010, for AUD/JPY, EUR/JPY, GBP/JPY, CHF/JPY, USD/JPY. 1.1 Average log return as a function of the hour of the day, averaged for 2003-2010, for AUD/USD, EUR/JPY, GBP/JPY, CHF/JPY, USD/JPY, in three-hour long time bins 1.2

Fig.1. Averaged logarithmic return vs hour of the day in CE time for the six leading exchange rates of the US dollar. 1.1: Hour-long time bin. 1.2: Three-hour long time bin.

Fig.1 presents hourly "seasonal" averages of the hourly logarithmic return in AUD/JPY, GBP/JPY, EUR/JPY, CHF/JPY, USD/JPY.

The errorbars assigned to the points represent the estimated precision of the mean for each time bin. One should bear in mind that due to the non-Gaussian nature of the distribution of (even logarithmic) returns, the confidence levels associated with these intervals are generally lower than would be expected for a Gaussian distribution.

Fig.1 shows a certain hints of a distinctive intra-day pattern, to some extent common to all six currencies. The pattern looks like a "hump" whose maximum lies around 14:00-16:00 CET (see the time zone conversion table) which corresponds to 8am-10am US Eastern time.

In other words it's best to be short yen for roughly 9 hours covering New York morning, and to be long yen for the rest of the day. However, the amplitude of the daily swing in the exchange rates involving yen is somewhat smaller than what was seen with USD.

The remarkable nose-dive the crosses take during the hour ending at 00:00 CET (6pm New York) is no doubt related to the end-of-business day on Wall St and in Chicago.

It is important to note that USD/JPY, the exchange rate involving USD, has somewhat a pattern of its own, obviously due to the influence of USD seasonality studied in a separate report.

When analyzing potential profitability of this effect, take into account trading costs. Out of the instruments involving yen, USD/JPY has by far the tightest spreads, usually being second only to EUR/USD. A spread as tight as 10-4 of the price level in USD/JPY, and about 1.5×10-4 in the case of EUR/JPY can be expected from a good broker. The rest of crosses have spreads much worse. This means that the amplitudes of the effect make it very difficult to profitably exploit the pattern in Fig.1, with the exception of USD/JPY which, as noted above, is influenced by the USD seasonality and in that capacity had already been discussed elsewhere.

In the next section, I look at the temporal stability of the effect year after year.



Last Updated ( Thursday, 17 May 2012 16:10 )