Brownian barrier survival asymptotic Created 2026-10-07 Updated 2026-10-14
For standard one-dimensional Brownian motion and a fixed , the Brownian reflection principle gives the survival probability . The standard normal distribution function has derivative at zero, giving the displayed asymptotic. Thus the persistence exponent is one half, with leading constant .
Call price in an arithmetic stock model with interest Created 2026-10-06 Updated 2026-11-05
For the arithmetic stock model with constant volatility, let , , , and
Thus for , and for . With the standard normal distribution function and standard normal density, the European call option price is
Indeed, the discounted terminal price is conditionally Gaussian with mean and variance . Integrating its positive part above the discounted strike price gives this expression. At maturity it reduces continuously to .
Differentiation gives option delta . Holding stock units and bank-account units gives a self-financing portfolio replicating the financial payoff, under the usual stochastic-integrability conditions. Its nonnegative wealth is admissible, and the discounted-wealth supermartingale bound shows that this is the least replication capital among nonnegative-wealth strategies.
Central limit theorem Created 2026-09-24 Updated 2026-10-14
Let be real independent and identically distributed random variables with expected value and finite, strictly positive variance . Write and for their sum and sample mean. The central limit theorem states
Precisely, for every real ,
where is the standard normal distribution function. The conclusion is convergence in distribution to the standard normal distribution. If the variance is zero, every observation equals almost surely; the displayed normalization is then undefined and there is no fluctuating error to describe.
The striking feature is that the observations need not have a normal distribution. They may be binary, bounded, or strongly skewed. Their detailed probability distribution disappears from this limit except through the expected value and variance. The subtraction removes predictable growth: . The division implements variance normalization: variance additivity for independent random variables gives , so every has expected value zero and variance one. For nonzero , an uncentered sum drifts; without rescaling its spread grows.
For a concrete example, let each have a Bernoulli distribution with success probability . Then has a binomial distribution, with expected value and variance . Its exact probability mass function and its normalized count are
Here is a binomial coefficient. Individual trials are conspicuously skewed, yet their centered, rescaled counts approach the same standard normal distribution as sums of fair trials.
Figure 1.
Exact standardized binomial distributions approach the standard normal curve
. For independent Bernoulli trials with , each blue bar has width equal to the normalized lattice spacing and area equal to the exact binomial distribution mass at its center. Heights are mass divided by spacing, so they can be compared with the orange probability density function of the standard normal distribution. All four normalized counts have expected value zero and variance one. The rectangles visualize discrete masses; they do not make the discrete probability distributions continuous. Details.
At , has expected value and standard deviation . The exact binomial distribution gives . A normal approximation with the continuity correction uses the interval to :
The small discrepancy is an approximation error, not an equality supplied by the central limit theorem.
The strong law of large numbers and weak law of large numbers locate the average: , respectively by almost sure convergence and convergence in probability. The central limit theorem magnifies that shrinking error and describes its shape. In terms of the sample mean, the resulting normal approximation is , with standard error . This explains why four times as many independent observations halve the standard error. When is known, has approximately coverage for at large , giving a confidence interval. This coverage refers to repeated samples, not a guarantee for an individual interval.
A useful application is the Monte Carlo method: estimate a success probability by the proportion of successful simulated trials. Nicholas Metropolis and S. Ulam's original paper, "The Monte Carlo Method" (1949), illustrates this with games of solitaire and estimation of a region's volume by randomly selected points, and discusses why estimates require probabilistic error control. The strong law of large numbers justifies convergence of such proportions for independent repetitions; the central limit theorem supplies their large-sample error scale when .[ref]
There are three useful limits to this picture:
A historical original-paper reference is J. W. Lindeberg's "Eine neue Herleitung des Exponentialgesetzes in der Wahrscheinlichkeitsrechnung" (1922), published in Mathematische Zeitschrift 15, pp. 211–225. The Lindeberg condition and Lindeberg-Feller central limit theorem below describe a version for sums whose summands need not share a probability distribution.[ref]
A brief proof outline uses the characteristic function of . Its expected value is zero and its variance is one, so the finite second moment gives
The Taylor expansion of the complex exponential, with its remainder controlled by , and the dominated convergence theorem justify this step without a third-moment assumption. Independence then gives the characteristic function of :
The limit is the characteristic function of the standard normal distribution; the Lévy continuity theorem turns this into convergence in distribution.
A conservative variance estimator, in the finite-sample expectation sense used here, is an estimator of a sampling variance satisfying
for every statistical parameter value in the stated statistical model. Assume and are finite. In design-based inference, the same definition applies for every permitted fixed population, with the sampling design supplying the randomness. The bias of an estimator is then nonnegative for this target. Aronow and Samii use this expectation criterion for their correction to the Horvitz-Thompson variance estimator when some pairs cannot be sampled together.[ref]
A samplewise variance upper bound would require the stronger inequality almost surely. An expectation bound allows some estimates below . It does not even require that every realized value be nonnegative; in practice a variance estimate is commonly chosen to be a nonnegative random variable.
One construction starts with an integrable random variable that is an unbiased estimator of and might take negative values. Its positive part satisfies , hence is a conservative variance estimator. The amount of upward bias of an estimator is exactly
The identity follows from the positive part and negative part decomposition . The Horvitz-Thompson variance estimator supplies a concrete setting in which the untruncated unbiased estimator can be negative.
A deliberately constructed example proves that conservative variance estimation does not guarantee Wald interval coverage. Observe from a normal distribution and an independent from a Bernoulli distribution with success probability . Use and . This nonnegative random variable has
Thus it is even an unbiased estimator of the sampling variance. Yet the usual Wald confidence interval
has coverage probability
where is the standard normal distribution function. When , the interval is the singleton , which contains with probability zero because is a continuous random variable. When , its conditional coverage probability is the bracketed expression. At , overall coverage probability is at most 1 percent, despite the nominal 95 percent rule and the exactly normal centre.
Figure 1.
An unbiased variance estimate can give a Wald interval with coverage at most one percent
. Original randomized variance-estimation counterexample. The variance estimate is zero with probability 0.99 and 100 with probability 0.01, so its expected value is the true variance 1. The interval can cover its target only in the rare second case. Details.
This example concerns the logical strength of the expectation criterion. Controlling the distribution of an estimated standard error, for example through an appropriate consistent estimator in an asymptotic argument, is additional work needed to justify confidence interval coverage.
Delta bound in an arithmetic stock model Created 2026-10-06 Updated 2026-11-05
For positive volatility and positive remaining maturity, the call price in an arithmetic stock model with interest has option delta , where is the standardized discounted moneyness. Since a finite argument of the standard normal distribution function gives a value strictly between zero and one,
For a fixed spot , the maturity limit is , the derivative of the call financial payoff away from its kink. The payoff derivative is undefined at . Under the nondegenerate Gaussian terminal law, the event has probability zero; it is this strike event that is exceptional. At zero interest use the same formula with normal scale .
For a standard normal random variable , its standard normal distribution function is
The change of variables and the symmetric distribution of the standard normal random variable give the second expression. For with , standardization of a normal random variable then gives
In particular, the intervals within one, two, and three standard deviations have calculated probabilities about , , and . These values depend on a normal distribution; they are not universal coverage guarantees for an arbitrary random variable. This is the direct connection between a Gaussian function area and the familiar central intervals.
For a small upper-tail probability, use the complementary error function directly:
Computing this as instead can lose the tail to catastrophic cancellation. The Gaussian tail bound gives another useful estimate without requiring an exact error function value.
In the dividend-free Black-Scholes model, let , , , and
Then the displayed finite sum prices the European option with power put option payoff. Here is the standard normal distribution function. Indeed, conditionally has a normal distribution with mean and variance . Expand by the binomial theorem on and apply the lower truncated lognormal moment to each term. Equivalently,
This expectation representation proves monotonicity and convexity in . It also gives and as by the dominated convergence theorem. For positive the lognormal distribution smoothing makes the price smooth and strictly positive.
For with and , is equivalent to the displayed chance constraint, because the standard normal distribution function is continuous and strictly increasing. The coefficient is a standard normal quantile, which can be negative. If , equality is an atom and the exact condition becomes instead.
For a normal distribution with mean , positive variance , and , the target is equivalent to the displayed constraint, where is the standard normal distribution function. This follows from and strict monotonicity of . Unlike the sufficient Chernoff bound margin , this constraint is necessary as well as sufficient. The exact quantile can be negative when .
For with a standard normal distribution and , split the Gaussian integral at zero and complete the square to obtain , where is the standard normal distribution function. This converts a bound on an absolute Gaussian random variable into a two-sided Gaussian tail bound using Markov inequality.
The Gaussian error linear unit activation is , where is the standard normal distribution function.
Take the original synthetic model and , mutually independent random variables, with the second argument denoting variance. The pair has a multivariate normal distribution with mean zero, unit marginal variances, and Pearson correlation coefficient . Its conditional multivariate normal distribution is
For a first reading , the second reading has conditional expectation and conditional variance , hence standard deviation . There is still conditional probability of , a movement farther from zero, where is the standard normal distribution function. The line describes a conditional mean, not an individual trajectory.
Figure 1.
An extreme first reading predicts a less extreme conditional mean, with substantial scatter
. Original synthetic scatter plot of 1,800 generated pairs, explicitly not actual measurements. Each pair has unit marginal variances and Pearson correlation coefficient ; orange points satisfy . The blue line is the conditional expectation , the band is one conditional standard deviation on either side, and the dashed line denotes . The band is not a confidence interval for an estimated line. The axes show a finite display window. Details.
A short derivation isolates . Its covariance with is zero and its variance is . Because has a multivariate normal distribution, zero covariance implies independence: independently of . This also explains why reversing the two measurements gives . The statistical association does not identify a direction of causal effect.
Half-normal distribution Created 2026-09-24 Updated 2026-10-14
The half-normal distribution describes the absolute value of a centered Gaussian random variable. Its unit-scale version is for a standard normal random variable . It folds the negative half of the standard normal distribution onto the positive half, retaining the size of a deviation while discarding its sign. If a centered Gaussian random variable models measurement error, its magnitude therefore has a half-normal distribution.
Here is the standard normal density and is the standard normal distribution function. For , the event is , so the symmetric probability distribution of gives . The upper-tail probability is therefore . For example, . The doubled probability density function integrates to one on , and .
The mean and variance of the unit-scale half-normal distribution are
For a short derivation, because , while gives . More generally, for has a half-normal distribution with probability density function for . Its mean is and its variance is ; here is the scale of the underlying Gaussian random variable, not the standard deviation of its absolute value.
Fix a scale and cutoff . Define the population Huber location estimator by the root
Here is the Huber score. It has Lipschitz continuity, so dominated convergence justifies differentiation when puts no mass at . Then , and the nonzero derivative identifies the local population root. The implicit function theorem applies away from the contaminating score's kinks; its Lipschitz continuity also gives the first-order expansion at those kinks. Differentiating its contaminated equation yields
The normalization matters: the Huber score alone is not the influence function. The gross-error sensitivity is , while the rejection point of an influence function is infinite because the clipped tails stay nonzero. The original clipped estimating criterion and its derivative normalization also appear in the population analysis of M-estimators.[ref]
Scale sets what counts as a large residual. For a normal distribution with fixed , symmetry gives and , where is the standard normal distribution function. If the scale is instead a statistical functional , differentiation of the joint equation adds the scale contribution:
Under a symmetric baseline and an odd Huber score, , so scale estimation drops out at first order. This cancellation needs regularity of and does not protect finite contamination changes: an inflated standard deviation enlarges the physical cutoff . A robust choice, such as a suitable Scale M-estimator, needs its own influence function and breakdown point checked. Fixing a numeric cutoff in raw units also sacrifices scale equivariance when units change.[ref]
Lattice obstruction to normal approximation Created 2026-10-06 Updated 2026-11-05
If a cumulative distribution function is constant between consecutive lattice points spaced by near zero, then its Kolmogorov distance from the standard normal distribution function is at least half the increase of that function across one such interval. This gives a lower bound of order .
Lilliefors test 2026-10-14
The Lilliefors test calibrates the Kolmogorov-Smirnov statistic against a normal distribution whose mean and variance are estimated from the tested observations. In the original procedure, is the sample variance with denominator , and is the standard normal distribution function. Its reference law includes this fitting step. The original paper's opening procedure and Table 1 explain the changed calibration.[ref]
Linear-boundary Brownian first-passage distribution Created 2026-10-14 Updated 2026-11-05
For , real and standard Brownian motion, the first-passage time to a linear boundary has
Here is the standard normal distribution function. Differentiation gives , the drifted Brownian first-passage density. Its Laplace transform is for . There is mass at infinity when ; for the hitting time is finite almost surely.
For , define , where has the standard normal distribution. For , completing the square in the standard normal density gives
where is the standard normal distribution function. The boundary values are and . Multiplying by the current stock price produces a zero-interest Black-Scholes formula with total variance and relative strike .
For the same normal distribution model, a two-sided test compares the null hypothesis with the alternative hypothesis . Under the null hypothesis, the standardized test statistic has standard normal distribution. Write for the standard normal distribution function. With , , and , the observed test statistic is . Its p-value is
Thus the p-value is a tail probability computed under the null hypothesis, including departures in both directions. It is not , the posterior probability of the null hypothesis, nor the probability that the observations arose "by chance". The normal distribution test statistic and its calibration appear in NIST's teaching reference for testing a process mean; the tail interpretation is stated in Penn State's hypothesis-testing notes.
Choose the significance level before observing the sample. The rejection region is , whose probability under is exactly 0.05. Here , so reject . This agrees with zero lying outside . More generally, a proposed belongs to the unrounded 95% confidence interval precisely when its two-sided test does not reject under this strict cutoff. The equality boundary has probability zero in this continuous statistical model.
Figure 1.
The chosen 5% rejection tails and the observed two-sided p-value
. Details.
The significance level and statistical power describe different repeated-experiment probabilities. At a specified alternative , put and . Then
The first is the Type I error probability; the second is the Type II error probability at that alternative. For a planning alternative , keeping , and gives , and statistical power . At , and statistical power rises to . These are model calculations for specified alternatives, not posterior probabilities inferred from a nonsignificant result. They explain why failure to reject cannot establish equality and why choosing sample size before an experiment is useful.
Historically, Ronald Fisher's 1925 book discusses the normal cutoff and tail probability 0.05 as a convenient convention on page 47, rather than a mathematically compulsory threshold: [ref].
Under the risk-neutral measure, set and . Then with a standard normal random variable, and . The log-call option is the discounted Gaussian positive-part expectation:
since . Therefore
Here and are the standard normal density and standard normal distribution function. At expiry the continuous terminal limit is .