Resources · EN
How a portfolio gets validated.
Long-form guides with cited sources and an encyclopedia of systematic trading: the same method VEEMAN implements, explained so you can verify it.
Guides
validation · overfitting · backtesting
How to validate a trading strategy before risking capital
A good-looking backtest is cheap. This guide walks the statistical checklist that separates edge from sample luck: out-of-sample discipline, deflated Sharpe ratios, the probability of backtest overfitting, Monte Carlo stress and live monitoring.
Updated 2026-07-11 · 9 min read
portfolio · correlation · sizing
Correlation is not constant: sizing a multi-strategy book for the days it breaks
The average correlation between your strategies is the least useful number about them. What matters is what correlation does in drawdowns — and how to size a portfolio for those days instead of the average ones.
Updated 2026-07-11 · 8 min read
Encyclopedia
All entries →Short, connected entries: the concepts behind every number in the program.
Sharpe ratio
Excess return per unit of volatility — the most quoted and most abused performance metric.
Probabilistic Sharpe ratio (PSR)
The probability that a measured Sharpe ratio is genuinely above a benchmark, given sample length, skew and kurtosis.
Deflated Sharpe ratio (DSR)
A Sharpe ratio corrected for the number of trials it took to find it — the antidote to strategy mining.
Probability of backtest overfitting (PBO)
The probability that the configuration chosen in-sample underperforms the median out-of-sample.
Block bootstrap
Resampling contiguous blocks of returns to preserve autocorrelation and volatility clustering.
Monte Carlo simulation
Generating thousands of alternative histories from your returns to read the distribution, not the anecdote.
Value at Risk (VaR)
The loss threshold that a day should only breach with a given small probability.
Expected shortfall (CVaR)
The average loss on the days beyond VaR — the size of the tail, not just its doorway.
Maximum drawdown
The deepest peak-to-trough decline of the equity curve — the number that actually removes traders from the game.
Probability of ruin
The chance that an equity path touches a loss threshold you cannot or will not survive.
Kelly criterion
The bet size that maximizes long-run growth — and why practitioners trade a fraction of it.
CUSUM (cumulative sum control)
A sequential change-point detector: the earliest statistically disciplined alarm that a process has drifted.