VaR (Value at Risk) = the loss you should NOT exceed with a given confidence level. At 95% confidence, 1-day VaR of 3% means: on 19 out of 20 days, you won't lose more than 3% — but on 1 day you might.
CVaR (Expected Shortfall) = the average loss in the worst scenarios that break through the VaR threshold. CVaR is a more complete picture of tail risk because it asks "when VaR is breached, how bad does it actually get?"
Method: Historical simulation using actual daily returns — no normality assumption. Multi-day figures use overlapping return windows.

Skewness: Negative skew = left tail is fatter (more chance of large losses than large gains — typical for stocks). Positive skew = right tail fatter (lottery-like payoff). Kurtosis (excess): > 0 means fatter tails than a normal distribution — extreme moves happen more often than models predict. Most stocks have positive kurtosis ("leptokurtosis"). This is why VaR based on normality is dangerously optimistic. Fat tail %: Days with |return| > 2 standard deviations. Under normality, this should be ~4.55%. Higher = fatter tails.

Drawdown = how far below its previous peak the stock is at any point. A drawdown of −30% means the stock has fallen 30% from its recent high. The chart shows the full underwater period — the deeper and longer it stays down, the higher the psychological and capital cost of holding.

Beta = sensitivity to the overall market (SPY). Beta 1.5 means when the market falls 10%, this stock tends to fall ~15%. Beta matters because market risk is diversifiable at the portfolio level — you may already have market exposure; idiosyncratic risk is the stock-specific risk you're adding. Idiosyncratic vol = the volatility left over after removing market beta. High idio vol means stock-specific factors dominate the risk — earnings, news, sector rotation.