Data Science in Finance

Lousy models are great!

The answer might surprise you!

Hungry for coin flips?

What's your $R^2$?

Our model:

$$ \begin{align*} Y =& \begin{cases} 1 & \text{ if "heads"} \\ 0 & \text{ if "tails"} \end{cases} \\ P(Y=1) =& w \\ \hat Y \equiv& 1 \end{align*} $$

$$ \begin{align*} \text{SSE} =& \sum\limits_{i=0}^{n-1}\left(y_i - \hat y_i\right)^2 & \text{SST} =& \sum\limits_{i=0}^{n-1}\left(y_i - \bar y\right)^2 \\ =& \sum\limits_{i=0}^{n-1}\left(y_i - 1\right)^2 & =& \sum\limits_{i=0}^{n-1}\left(y_i - 0.5\right)^2 \\ =& \sum\limits_{i=0}^{n-1}\left(y_i^2 -2y_i + 1^2\right) & =& \sum\limits_{i=0}^{n-1}\left(y_i^2 - 2(0.5)y_i + 0.5^2\right) \\ =& \sum\limits_{i=0}^{n-1}\left(-y_i + 1\right) & =& \sum\limits_{i=0}^{n-1}\left(0.25\right) \\ =& -nw + n & =& 0.5n \\ =& n(1-w) & =& 0.5n \\ \end{align*} $$

So, our $R^2$ is:

$$ \begin{align*} R^2 =& 1 - \frac{\text{SSE}}{\text{SST}} \\ =& 1 - \frac{n(1-w)}{0.5n} \\ =& 2w - 1 = 1.1 - 1 = 0.1 \end{align*} $$

Not necessarily bad Data Scientists

So what are the bets we're making?

Not just making money

Looking good?

Better yet!

So what's the game?

Active Portfolio Management

Richard C. Grinold, Ronald N. Kahn

The Hedge Fund Mission

Make money

Like Roulette

A paraphrasing

The CAPM

Capital Asset Pricing Model

Risk

What is it?

Variance?

Exceptional Returns