Quant model demos
These eight models run in your browser. The data is randomly generated and does not show any strategy’s past or future results.
01 / 08 · How prices move
Geometric Brownian motionGBM
Where could the price go?
It treats log price changes as a random walk with drift. The higher the volatility, the wider future paths fan out.
Sample data · randomly generated · Log scale
- Sample paths
- 90% range
- Median
- Expected value
Parameters
Results
- One-year return, median path
- −7,7%
- Expected one-year return
- +10,5%
- Chance of a loss after a year
- 55%
In trading
The baseline for option pricing, risk budgets, and scenario generation. The chart shows volatility drag: even with a positive expected return, the median path can fall, because the median grows at μ − σ²/2.
Limits
It assumes constant volatility and normally distributed returns. Real crypto markets have gaps, fat tails, and volatility clustering.