/sci/ - Science & Math

>>10196612
In the 1970s, Robert Lucas made the Lucas critique. It basically goes like this:
Imagine you're a producer. Obviously, you want to maximize your profits, and to do that, you pay some general attention to how the government's acting on the economy, and the general theoretic basis of that, even if you don't read too much economics.
So imagine that the government notices that, if they bring down the interest rate they can trick you into investing, and then they let it go up again. So the government keeps doing this, but because you aren't an idiot, you notice. So, when the interest rate goes down, you let a couple of months pass before even thinking about investing.
In other words, even if there existed a correlation between the interest rate and investment, or more famously, between inflation and product, the moment the government's action on these things changes, the people start acting in different ways, because the game rules changed.
What this means is that using statistics in economics is something really important, but also strangely sensitive. If the government finds a correlation in the data, and tries to exploit it, it ends up shooting itself in the foot after people learn how to exploit that.
If you're using, say, the stock market, then you have a similar problem: the stock market efficiency hypothesis. Even if you want to try to use data to maximize profits, everyone else is already doing that.
You need to remember that people think. They might not be super rational geniuses, but they aren't retarded.

>>10180822
Check'em.

>>10174794
Economics isn't a science or math? Heavens no. You must be mistaking me for a finance major. I study the highly mathematical economical sciences. We do important work such as analysing firms and market structures and use advanced maximization techniques to theoretically derive free market's behavior from basic game theory. Without us, the economy would go broke. Our theoretical subject uses far more maths than any engineering discipline or science, including computer science, and we're practically a superset of maths.

>>10123477
>a
>stochastic component
>between 0 and 1
>up
>to
>a
>high
>potence