click below
click below
Normal Size Small Size show me how
Quant: Statistic
Question | Answer |
---|---|
Geometric mean | [(1+R )(1+R )........(1+R )]^(1/n) - 1 |
Population Mean | m = Sum (Xi) / N |
Sample Mean | X = Sum(Xi) / n |
Weighted Mean | X = Sum(wiRi) = w1R1 + w2R2 +...+wnRn Where R , R , ..., R are the returns for assets 1,2,...,n and w , w , ...,w are the portfolio weights, so that w + w + ...+w = 1 |
Harmonic Mean | N / (Sum (1/Xi)) where : N = number of purchases (equal $ amounts) X = share price for each purchase |
Variance | SUM( Xi - mean)^2 / N |
Standard deviation | (Variance^2) ^ (1/2) |
Sample Variance | SUM( Xi - sample mean)^2 / n-1 |
Sample Standard deviation | Sample Variance ^ (1/2) |
Coefficient of Variation | stdev / mean where: the lower the better |
Sharpe Ratio | (Rp - Rf) / stdev p where: the higher the better |
Chebyshev’s Inequality | 1- (1/k^2) |
symmetrical distribution | skew = 0 |
Leptokurtic | kurtosis > 3 and more peaked with fatter tails |
Kurtosis for a normal distribution | 3.0 |
Excess kurtosis | kurtosis - 3 |
Excess kurtosis for a normal distribution | 0 |