E(X) is the expected, long-run average value of a random variable—found through foundational principles like linearity of expectation. As you move down the page, you’ll see multiple ways to compute it, including integral-based definitions and transform tools such as the moment generating function. The guide then connects E(X) to inference via the law of large numbers and the central limit theorem, and to modeling examples like risk-neutral pricing and portfolio expected returns.
Key Takeaways
- Law of large numbers implies sample mean converges to E(X), central to statistical inference
- Central Limit Theorem states sqrt(n)(bar X_n - E(X)) -> N(0, Var(X)) under mild conditions
- Moment generating function M_X(t) = E[exp(tX)], uniquely determines distribution if exists
- In Black-Scholes model, E(S_T) = S_0 exp((r - q)T) under risk-neutral measure for dividend yield q
- Portfolio expected return E(R_p) = sum w_i E(R_i) by linearity, regardless of correlations
- CAPM predicts E(R_i) = R_f + β_i (E(R_m) - R_f), linear security market line
- The expected value E(X) of a Bernoulli random variable with success probability p is exactly p, representing the long-run average proportion of successes in repeated independent trials
- Linearity of expectation states that E(aX + bY) = aE(X) + bE(Y) for any random variables X and Y and constants a, b, holding regardless of dependence between X and Y
- For any random variable X, E(X) equals the integral over the probability space of X(ω) dP(ω), providing the foundational measure-theoretic definition
- Exponential(λ) rate has E(X) = 1/λ, memoryless interarrival time mean
- Normal(μ,σ²) has E(X) = μ, the location parameter defining the mean
- Uniform[a,b] continuous has E(X) = (a+b)/2, identical to discrete case by symmetry
- For a Binomial(n,p) distribution, E(X) = np, representing the expected number of successes in n independent Bernoulli trials each with success probability p
- Poisson(λ) random variable has E(X) = λ, where λ is both mean and variance parameter, modeling rare events count
- Geometric distribution (trials until first success, p) has E(X) = 1/p, the average trials needed for first success
Expectations underpin inference and pricing, governing long run averages and enabling key limit theorems.
Related reading
01 · Category
Advanced Topics20 stats
01
Law of large numbers implies sample mean converges to E(X), central to statistical inference
02
Central Limit Theorem states sqrt(n)(bar X_n - E(X)) -> N(0, Var(X)) under mild conditions
03
Moment generating function M_X(t) = E[exp(tX)], uniquely determines distribution if exists
04
Characteristic function φ_X(t) = E[exp(i t X)], always exists, Fourier transform of density
05
Stein's lemma for normal X ~ N(μ,σ²), E[(X-μ) f(X)] = σ² E[f'(X)] for differentiable f
06
Efron-Stein inequality bounds Var(E[Xi | X_{-i}]) ≤ Var(X)/n for sum X= sum Xi
07
Optional stopping theorem: for martingale M_t, E[M_τ] = E[M_0] under stopping time conditions
08
Doob's martingale convergence: E[sup |M_n|]<∞ implies M_n -> M_∞ a.s. with E[|M_∞|]<∞
09
Burkholder-Davis-Gundy inequality relates E[sup |M_t|^p] to E[<M>_t^{p/2}] for martingales
10
Concentration inequalities like McDiarmid: P(|E(X|S)-E(X)| ≥ t) ≤ 2 exp(-2 t² / sum c_i²) for bounded differences
11
For sub-Gaussian X with variance proxy σ², P(|X - E(X)| ≥ t) ≤ 2 exp(-t²/(2σ²)), tail bound
12
Hoeffding's inequality for bounded [a_i,b_i] independent sum S_n: P(|S_n - E S_n| ≥ t) ≤ 2 exp(-2 t² / sum (b_i - a_i)²)
13
Wald's equation for sequential analysis: E[sum_{i=1}^N X_i] = E(N) E(X) under independence
14
Azuma-Hoeffding for martingale diff bounded c_i: P(|S_n|≥t)≤2exp(-t²/(2 sum c_i²))
15
Freedman's inequality for martingale with bounded diff and variance process, tighter than Azuma
16
Talagrand's inequality for convex lipschitz functions on product spaces, concentration
17
Transportation inequality: Wasserstein distance W_2(μ,ν) ≤ const sqrt( KL(μ||ν) ) relates means indirectly
18
Posterior mean E(θ | data) = integral θ π(θ|data) dθ in Bayesian
19
Empirical Bayes shrinks E(θ_i | data_i) towards grand mean, James-Stein
20
Reinforcement learning policy gradient ∇ E[reward] ≈ sum ∇log π(a|s) A(s,a)
Interpretation
Advanced Topics Interpretation
Across these advanced topics, the key trend is that increasingly large samples make E(X) both predictable and usable, since the sample mean converges by the law of large numbers and the central limit theorem then provides a normal approximation scaled by √n.
02 · Category
Applications In Finance24 stats
01
In Black-Scholes model, E(S_T) = S_0 exp((r - q)T) under risk-neutral measure for dividend yield q
02
Portfolio expected return E(R_p) = sum w_i E(R_i) by linearity, regardless of correlations
03
CAPM predicts E(R_i) = R_f + β_i (E(R_m) - R_f), linear security market line
04
For geometric Brownian motion dS = μ S dt + σ S dW, E(S_t) = S_0 exp(μ t), exponential growth mean
05
Value at Risk VaR_α ≈ -μ_p + z_α σ_p for normal returns, but E(loss | loss > VaR) involves tail expectation
06
Actuarial present value E[discounted payoff] underlies insurance premium calculation
07
Optimal stopping in American options uses E[continuation value] vs exercise
08
Kelly criterion maximizes E[log wealth] for bet sizing f* = (p b - q)/b in favorable games
09
Arbitrage-free pricing sets E^Q[discounted payoff] = price under risk-neutral Q
10
Bond duration approximates -dP/dr / P ≈ E[time-weighted cashflows], Macaulay duration
11
In martingale pricing, discounted asset price is martingale so E_t[S_T exp(-r(T-t))] = S_t
12
Fourier transform methods compute E[payoff(S_T)] via characteristic function for option pricing
13
In inventory theory, EOQ model has expected holding + setup cost minimized at Q* = sqrt(2 K D / h)
14
In S&P500 historical, average annual return E(R)≈10-12% nominal 1926-2023
15
Bitcoin daily log returns have E(R)≈0.003 or 0.3% but high vol, 2010-2023
16
US Treasury 10yr yield E(annual change)≈0% long-run stationary
17
Sharpe ratio = (E(R_p) - R_f)/σ_p, typical equity 0.4-0.6
18
Implied vol from options gives E^Q[log S_T/S_0] = (r-q)T - σ²T/2
19
Monte Carlo simulation estimates E[payoff] with std err σ/sqrt(N), convergence rate
20
Binomial tree for options converges to BS as n→∞, E[payoff] discounted
21
GARCH(1,1) forecasts conditional E(R_t | past)= μ + effects, volatility clustering
22
Factor models E(R_i)= α + β1 E(F1) + ... , Fama-French 3-factor avg premiums
23
In gambling, house edge = -E(player payoff per unit bet), roulette ≈5.26% American
24
Equity risk premium E(R_m - R_f) US historical 1926-2023 ≈6.5%
Interpretation
Applications In Finance Interpretation
Across major applications in finance, expected values often turn into clean exponential or linear forms such as E(S_t) = S_0 exp((r − q)t) in Black Scholes and CAPM’s E(R_i) = R_f + β_i(E(R_m) − R_f), showing how risk-neutral and model based assumptions make forecasting mean outcomes sharply more tractable.
03 · Category
Basic Properties10 stats
01
The expected value E(X) of a Bernoulli random variable with success probability p is exactly p, representing the long-run average proportion of successes in repeated independent trials
02
Linearity of expectation states that E(aX + bY) = aE(X) + bE(Y) for any random variables X and Y and constants a, b, holding regardless of dependence between X and Y
03
For any random variable X, E(X) equals the integral over the probability space of X(ω) dP(ω), providing the foundational measure-theoretic definition
04
The expected value E(X) is always between the minimum and maximum possible values of X, specifically min ≤ E(X) ≤ max for bounded X
05
Jensen's inequality asserts that for convex function φ, φ(E(X)) ≤ E(φ(X)), with equality if X is constant, quantifying the convexity effect on expectations
06
E(X) for a uniform distribution on [a,b] is precisely (a+b)/2, the midpoint of the interval, reflecting symmetry
07
Non-negativity preservation: if X ≥ 0 almost surely, then E(X) ≥ 0, a fundamental monotonicity property
08
For indicator random variable I_A, E(I_A) = P(A), linking expectation directly to probability of event A
09
Monotonicity: if X ≤ Y almost surely, then E(X) ≤ E(Y), provided expectations exist
10
E(c) = c for any constant c, the degenerate case where variance is zero
Interpretation
Basic Properties Interpretation
In the Basic Properties category, the key trend is that expected value behaves reliably across models, from the clean midpoint rule E(X) = (a+b)/2 for a uniform distribution to the universal bounds min ≤ E(X) ≤ max for bounded variables.
More related reading
04 · Category
Continuous Distributions24 stats
01
Exponential(λ) rate has E(X) = 1/λ, memoryless interarrival time mean
02
Normal(μ,σ²) has E(X) = μ, the location parameter defining the mean
03
Uniform[a,b] continuous has E(X) = (a+b)/2, identical to discrete case by symmetry
04
Gamma(α,β) shape-rate has E(X) = α/β, sum of exponentials mean
05
Beta(α,β) on [0,1] has E(X) = α/(α+β), mean proportion
06
Weibull(k,λ) shape-scale has E(X) = λ Γ(1 + 1/k), involving gamma function for lifetime modeling
07
Lognormal(μ,σ²) has E(X) = exp(μ + σ²/2), moment-generating derived mean
08
Pareto(xm, α) minimum xm, shape α>1 has E(X) = α xm / (α-1), power-law tail mean
09
Cauchy(μ,γ) has undefined E(X) due to heavy tails, no finite mean exists
10
Chi-squared(k) degrees freedom has E(X) = k, sum of squares of standard normals
11
Normal(0,1) E(X)=0, defining standard mean
12
Exponential(λ=2) E(X)=0.5, half-life like
13
Gamma(α=3,β=1) E(X)=3, Erlang special case
14
Beta(2,5) E(X)=2/7≈0.2857
15
Lognormal(μ=0,σ=1) E(X)=exp(0.5)≈1.6487
16
Pareto(xm=1,α=2.5) E(X)=2.5/1.5≈1.6667
17
Weibull(k=2,λ=1) E(X)=Γ(1.5)≈0.8862, Rayleigh special
18
Student-t(df=5) E(X)=0 for df>1
19
Logistic(μ=0,s=1) E(X)=0, sech² density symmetric
20
For Uniform[0,1] E(X)=0.5
21
Exponential(1) E(X)=1
22
Normal(5,2) E(X)=5
23
Beta(1,1)=Uniform[0,1] E=0.5
24
Gamma(1,1)=Exp(1) E=1
Interpretation
Continuous Distributions Interpretation
Across continuous distributions, the mean is determined by the key parameter that shapes the distribution, from the exponential giving E(X)=1/λ and gamma giving E(X)=α/β to the uniform averaging endpoints at (a+b)/2, showing that in this category the central tendency is directly controlled by the defining rates, locations, or interval bounds.
05 · Category
Discrete Distributions21 stats
01
For a Binomial(n,p) distribution, E(X) = np, representing the expected number of successes in n independent Bernoulli trials each with success probability p
02
Poisson(λ) random variable has E(X) = λ, where λ is both mean and variance parameter, modeling rare events count
03
Geometric distribution (trials until first success, p) has E(X) = 1/p, the average trials needed for first success
04
Negative Binomial(r,p) for r successes has E(X) = r/p, expected trials for r-th success
05
Hypergeometric(N,K,n) population N with K successes, draw n, has E(X) = n(K/N), unbiased estimator of proportion
06
For Discrete Uniform {1,2,...,k}, E(X) = (k+1)/2, average of first k naturals
07
Multinomial(n, p1,...,pm) marginal for i-th category has E(X_i) = n p_i, generalizing binomial
08
Zeta distribution with parameter s>1 has E(X) = ζ(s-1)/ζ(s), involving Riemann zeta function for tail-heavy counts
09
Log-series distribution (p) has E(X) = -p / ((1-p) log(1-p)), modeling species abundance
10
Discrete Pareto (xm, α) has E(X) = α xm / (α-1) for α>1, heavy-tailed discrete analog
11
For Binomial(n,p), E(X) = np exactly, with variance np(1-p)
12
Poisson(λ=5) has E(X)=5, P(X=k)= e^{-5} 5^k / k!
13
Geometric(p=0.3) E(X)=1/0.3 ≈3.333, variance (1-p)/p²≈7.111
14
Negative Binomial(r=2,p=0.4) E(X)=2/0.4=5
15
Hypergeometric(N=50,K=20,n=10) E(X)=10*(20/50)=4
16
Multinomial(n=100, p=(0.3,0.4,0.3)) E(X1)=30, E(X2)=40, E(X3)=30
17
Zeta(s=2) E(X)= ζ(1)/ζ(2) but ζ(1) diverges, actually for truncated finite mean ≈1.64493/1.64493 wait no, properly ζ(s-1)/ζ(s)≈ π²/6 / π²/6 *ζ(1) invalid, for s>2
18
For Binomial(n=100,p=0.5) E(X)=50
19
Poisson(λ=10) E(X)=10
20
Geometric(p=0.1) E(X)=10
21
Hypergeometric(N=100,K=30,n=20) E(X)=6
Interpretation
Discrete Distributions Interpretation
Across common discrete distributions, the expected value often scales in a simple, interpretable way with the key parameter, such as E(X)=np for Binomial and E(X)=λ for Poisson where both mean and variance match, so in the discrete setting the “average count” is directly tied to the underlying process rate or probability through formulas like 1/p for Geometric and (k+1)/2 for Discrete Uniform.
Reference
Cite This Report
This report is designed to be cited. We maintain stable URLs and versioned verification dates. Copy the format appropriate for your publication below.
APA
Lukas Bauer. (2026, February 13). E(X) Statistics. Gitnux. https://gitnux.org/e-x-statistics
MLA
Lukas Bauer. "E(X) Statistics." Gitnux, 13 Feb 2026, https://gitnux.org/e-x-statistics.
Chicago
Lukas Bauer. 2026. "E(X) Statistics." Gitnux. https://gitnux.org/e-x-statistics.
Sources & references
40 datasets cited across this report · attribution is report-level