Quantitative Methods IV Flashcards
Binomial distribution
Success and failure, where prob is constant for each trial
Binomial expected value
E(x) = N * P
Number of trials * probability of success
Binomial variance
N * P * (1-P)
Continuous uniform distribution
Prob of outcome = % of range
I.e. range = 2 to 12, prob of 4 to 8 = (8-4) / (12-2)
Normal distribution
mean = mode = median Kurtosis = 3 Skewness = 0
Multivariate distributions
Mean
Variance
Correlation
N means
N variances
0.5N(n-1) pairs of correlation
Confidence intervals for normal distribution
90% = -1.65s to 1.65s 95% = -1.96s to 1.96s 99% = -2.58s to 2.58s
Z-score
(sample mean - hypothesized mean) / standard error
How many stdevs away from the mean you are
Standardized so mean = 0, stdev = 1
Lognormal distributions
e^x
Skewed to the right and bounded by zero, useful with non negative asset prices
Compounding with continuous periods
Effective annual rate
e^r - 1
R = return
Compounding with continuos periods
Convert effective rate to stated rate
Ln(effective rate + 1)
Stratified random sample
Classification system to separate into subgroups and then take samples based on size relative to population.
Guaranteed to sample from each stratum, not true of simple random sample.
Central limit theorem
Sample distribution approaches normal as sample grows
Variance of distribution of sample means = variance / n
Standard error of sample mean
standard deviation / sqrt(N)
Level of significance
5% significance = 95% confidence