Probably someone with a little mathematical knowledge and a lot of experience could answer this easily

I would like to know the probability density function of the change in forex rates from say next monday 1:00pm to monday 1:01pm

The approach I am using (I'm not claiming this to be the best way) is to simply look back through past history. Look what happened between 1:00pm - 1:01pm on every Monday so far

So if we do this looking at the past year, we end up with 52 data points. Where do we go from here? I could easily assume a log-normal distribution or something like that, but I hear there tends to be a larger kurtosis in reality

Perhaps this is well studied by some and there's a popular form of probability distribution people approximate the markets with?

Are there forex options? If so I could maybe for some time intervals derive the probability distribution based on the various prices of options?

And for those who don't know the answer, for forex or stocks, how is it you calculate the variance of a play to sufficiently protect your bankroll? all feel?

HP, Here's a standard methodology for market risk:

http://www.riskmetrics.com/publications/techdoc.html

Edit: Also, I really hope you're not playing around in FX.

ty, this looks real good

not yet!

curious though, are you saying that simply because of my huge lack of knowledge or something else? Criticism of any kind is appreciated

You're using time as a set variable when there really should be no relevance or use for its past history. Time change doesn't affect prices.

as a trader in futures, stocks, and FOREX myself, I'll tell you FOREX is probably the toughest to trade. There are more fake technical signals than any other market that I engage in. It's like playing in teh highstakes games instead of grinding it out in mid stakes (stocks/futures would go there)

well, I would disagree

I'd lay you 1.1 to 1 odds I could pick out two one-minute intervals next week for the EUR/USD, and tell you which interval will have the larger change

The main effect I am trying to measure here is the variance of the EUR/USD at certain times of the week, like when the NYSE opens, as opposed to Tues. 3:25am

So when people are putting through forex orders, you think that the time of day plays an unconscious (or conscious I guess) factor on what move they make?

I don't really know what how most traders trade, but I would guess yes

like if you are taking breakouts into consideration, and you notice a breakout that occurred at 930am eastern time, you might treat it differently

Why? People don't look at the clock when placing orders. You're trying to find a correlation between price movements and raw time that doesn't exist, and it doesn't matter how many numbers you pull out that might show a trend, if the theory sucks your algorithm won't predict future behavior because you were essentially, fooled by randomness.

I was going to make a million dollars playing the Forex a few years ago (I used www.oanda.com)

I lost my $1200.00 bankroll in a pretty impressive amount of time. I seemed to have an incredible knack for placing my stop-losses at the PRECISE bottom of whatever time-frame I was going for.

I always wondered how the intreast that is added/subtracted from traders accounts effects trades in the last 60,30,15,10,and 5 minutes of trading. Wouldn't this be a good application where time plays a role in the trades being made?

Do you mean margin interest? if so I can't imagine anyone that trades that thinly will last long.

Jimbo

Change in net interest across all accounts would always be zero. If you could somehow figure out who's accumulating/unloading and who's day-trading, it'd be easy to make money.

Yes. I don't mean they are buying just to earn the intreast but that it would make more sense for orders to come in before 5 rather then after.

Isn't he more saying that there are time intervals during which expected volatility is higher than other time intervals? This is probably true, as certain times during the day (especially given a 24-hour trading day) have a higher rate of information flow than others. It's not like this violates any sort of arbitrage principle.

It looks as if there are two different conversations going on in this thread, so the best thing to do is to start clarifying things. It looks as if the OP is curious about the PDF of prices during specific time intervals, and not concerned about the directionality. Art is assuming that the OP is concerned about directionality.

The short answer: If you look on the internet, google: forex time of day trading, you can find many sites that can give you studies showing the pip range of most pairs, certainly the major pairs, for any time of day, usually hourly. If you are a member at a forex broker, Oanda, etc, they should have that information available to its members. To answer the OP: most pairs have it's highest volatility near its domestic open and around the open of its cross currency. In english: look at a pair like GBP/USD, you will find the highest volatility at 3 am/4 am EST, which is the open of the FTSE, and at around 8 am/9am which is around the time of the open of the NYSE. You will also see good volatility near the close.

This does affect trading, based on your style. If you are a long term trader, then this will have no effect on the decisions you make for trading. But for an intraday trader, those are the times where you are looking to make the bulk of you money, as the rest of the day ranges, and for some, ranges flatly and will not give you the volatility that you are looking for.

THE HUN.

While I am searching for correlations, here I am just trying to figure out the volatility and more precisely what type of volatility. And yes I still believe myself and others will trade differently based on this knowledge, of course it's not the entirety of the my system or theirs

I think thehun has summed up everything well