Maybe - but your own algo would have missed the fire alarm, right?
Which is why mathematical/technical traders are always adding in more complexity/decision points to make up for all the fire alarms they appear to have missed in hindsight.
A bit like MSNBC always call the reason for moves in retrospect. Hindsight analysis - looking for effect in cause.
The fact is - it's NOT just about processing the available info - a good report can move a market up one day and down the next. The markets are fickle & emotional as much as they are mathematical.
It can be modeled:
You posted a hypothetical question, so let me give you several hypothetical answers:
1. You input into your model, all the informations on each individuals, who, what and where in a set of initial conditions, you also input into your model, the store informations, then run a time series calculation, you can also run a Monte Carlo, ending up with a set of most likely outcome over time series.
2. You do something easier, assume a Brownian motion but input the boundary conditions that include the layout, size and store information. You can calculate the probability density function of where the people are. You can even follow an individual "particle" (people) and calculate its individual path as a Weiner Process....
3. And by the way, you model it not by predicting the fire alarm but by including in your model an input perturbation function, like the fire alarm or other non random events. So, you model what happen AFTER the fire alarm.
In trading the Quants are trying to do #1, we amateurs are trying to do #2,
@destriero @TheBigShort et al tried to do #3 in this thread:
https://www.elitetrader.com/et/threads/earnings-journal.328822/ The fire alarm is Earning.
I am just babbling, not really making any sense, so you can ignore this post.
