Module 03 · AAPL stochastic paths

Apple as a process, not a point.

Geometric Brownian motion, calibrated on eight Apple annual log-returns, with σ blended against live realized vol. S0 is the last print.

Drift μ

28.3%

Annual log-return mean

Volatility σ

27.1%

blend · annual μ + live realized σ

S0

$313.45

Live tape

Paths

10,000

Stochastic paths

AAPL GBM fan — five years of Wiener noise

NowY1Y2Y3Y4Y5$3,003.39$238.63
Median $1,061.68 IQR 83.3% 5–95

GBM statistics

MetricYear 1Year 3Year 5
Mean$415.90$731.43$1,276.19
Median$400.44$658.81$1,061.68
P5$259.39$304.50$402.98
P95$626.62$1,423.24$2,860.37
P(> $300)86.1%95.3%98.4%
P(> $400)50.2%85.7%95.1%
P(< $150)0.0%0.1%0.1%

GBM finish table

AAPL finish-position probabilities

Each cell is the share of simulated paths that finish a given year inside that price band.

Horizon$50–150$150–250$250–350$350–450$450–650$650–∞
Year 1
Year 2
Year 3
Year 4
Year 5

Jump-diffusion

AAPL Merton jumps · λ = 0.3 / yr, mean jump −5%

NowY1Y2Y3Y4Y5$3,125.13$231.04
Median $1,057.81 IQR 87.3% 5–95

Median comparison

Jump-diffusion versus GBM (median path)

NowY1Y2Y3Y4Y5$1125

How to read this

GBM assumes log-returns are well-behaved. Jump-diffusion adds Poisson shocks — the model’s way of admitting a regulation case, a demand air-pocket, a China weekend. The medians stay close. The fifth percentile does not. That gap is the point. Drift is the eight-year 10-K mean; volatility is blended with whatever the live session has already printed.