Have you noticed a pattern in xx50 or xx00?

Please Note: I am based in the UK so my midnight is currently GMT +/- 0.

Draw lines at the xx00 an xx50 on EUR/USD.
Note the last xx50 or xx00 that the price touches prior to midnight.

On the vast majority of occasions during the next 24 hours price will come back to that level (I realise that 'vast majority of occasions' is not very exact!)

It also seems to work on EUR/CHF and GBP/USD.....and probably others!

I've highlighted some examples on the attached GBP/USD chart.

The problem is price can shift enormously prior to returning and figuring out how to effectively trade it.

Comments welcome.

EUR/CHF crossed 13,000 at 16:00GMT yesterday and hasn't been back there since - i.e.this was the last xx00 or xx50 touched prior to midnight.

Will it get back there sometime today?

Price is currently 12964.

Just looking back over data since 25/10/10.

For the moment, ignoring any appropriate stop, price came back in the next 24 hours to the hit previous xx00 or xx50 at the following rates;

Eur/usd - 65 out of 70 - 92% success rate
GBP/USD - 60 out of 70 - 85% success rate
Eur/Chf - 62 out of 70 - 88% success rate.

These figures are based on price coming back within the following 24 hours and with no stop loss in place.

Obviously no SL is suicide and figuring that one is the deal breaker!

Eur/Aud - 5 out of 70 - 92% success rate
Eur/Gbp - 10 out of 70 - 85% success rate

This information alone isn't going to provide you with any edge. Same goes for divergence, price will ALWAYS come back to its divergence-target yet this information doesn't provide you with any edge alone.

Btw, this will work with random lines too.

True that. The chances of any given price in hour 1 being seen again in the subsequent next 23 hours are around 90% on EU if I remember right (10y test) so your sample of 70 is pretty accurate. Explains why people 'learn' to hold on to losers.

Unfortunately, the 1 in 10 chance gives you a belting that'll make you wish you never did hold on. But not enough of a belting to make it an edge to do the reverse either.

Thank you guys - saves me wasting my time on it. :>

As you rightly point out the 1 loss in 10 wipes out the measly return on the other nine, no matter how tempting it looks!