intra day stop loss hit probability

tomerico · October 23rd, 2010, 11:57 am

Hello all . About a question i"ve recently asked but without a solution yet :suppose i want to estimate an asset (currency pair , stock ,etc ) hitting probability of some level in close time period ( e.g say EU/USD is 1.4 now and i wanna estimate it's hit probability in the 1.45 OR 1.36 values in the coming hour ) - how can i estimate that ? Thing is the assumption of Brownian Motion for the intraday return is not true so i can't use Monte Carlo or Barrier options procing models .Anyone ? ...

list · October 23rd, 2010, 12:10 pm

It's a subjective point. It might be make sense to find factors from wider area which effect on your currency pair , stock. It can be also your own developed index. If you interested in say 3 hour period look at the previous 3 hours factor that suggests your expectation. In more sophisticated model your specified favourable and unfavourable factors. 3-hour back is somewhat markovian approach.

tomerico · October 23rd, 2010, 12:17 pm

Thanks . I could look for "similiar " market conditions in the past (thousands of them ) and build some mean and STD to build a some probability distribution ? This looks naive statistics (?) ...Are there any microsturcture models which are "accessible" enough ? Tom

list · October 23rd, 2010, 3:34 pm

use distant past history can be useful as well as not. we need some evidence to use distant past.