The single number that tells you how bad a bad day gets
You have $250,000 invested and you feel comfortable. Then you run a Value at Risk calculation and the screen says: at 95% confidence, you could lose $8,200 in a single day. Suddenly comfortable feels like a guess. That is exactly what VaR is built to replace.
Value at Risk answers one question banks and hedge funds ask every morning: what is the most I could realistically lose over a given period, and how confident am I in that number? A 95% one-day VaR of $8,200 means that on 95 days out of 100, your loss should stay under $8,200. On the other 5 days, it could be worse.
The calculation depends on three inputs. First, your portfolio value. Second, its volatility, usually the standard deviation of daily returns — a stock-heavy portfolio might run 1.2% daily, a bond-heavy one closer to 0.4%. Third, your confidence level, typically 95% or 99%. Higher confidence means a larger, more conservative loss estimate.
Here is how the numbers move. At 95% confidence, the model uses a z-score of about 1.65. At 99% confidence, it jumps to 2.33. So bumping your confidence from 95% to 99% on that same 250,000 portfolio pushes the estimated worst-day loss from roughly8,200 to about $11,600. Same portfolio, stricter question, bigger number.
The point of VaR is not to predict the future. It is to put a concrete dollar figure on your downside so you can decide whether you can stomach it. A number you can name is a risk you can manage. A vague feeling of safety is not.