Conditional Probability

Conditional probability answers the question: given that something has happened, how likely is something else? The chance that a stock rises tomorrow might be about 50% on its own, but the chance that it rises given that it just reported a large earnings beat may be different. Most useful trading questions are conditional: what is the win rate of this setup when the trend is up, or when volatility is high? Understanding conditional probability helps traders build better filters and avoid common reasoning errors.

The formula#

P(A | B) = P(A and B) / P(B)

Read P(A | B) as "the probability of A given B".

A trading example#

Why conditional thinking matters#

• Filters: trend filters, volatility filters and time of day filters are all conditional probabilities. See Combining Signals.
• Regimes: a strategy's win rate can differ greatly between calm and volatile markets. See Structural Breaks and Regime Changes.
• Risk: the probability of a large loss given a gap or news event is higher than unconditionally.
• Correlation: if one position loses, the probability that related positions also lose rises. See Correlation Management.

Common mistakes#

Confusing P(A | B) with P(B | A)#

The probability of a crash given that the VIX spiked is not the same as the probability of a VIX spike given a crash. Nearly every crash comes with a VIX spike, but many VIX spikes are not followed by crashes. Mixing these up is called the inverse fallacy. Bayes' theorem connects the two. See Bayes' Theorem.

Ignoring base rates#

A pattern that appears before 80% of big rallies sounds powerful. But if the pattern also appears constantly at other times, and big rallies are rare, the chance of a big rally given the pattern may be small.

Overfitting conditions#

Adding many conditions can produce high win rates on historical data by chance. Each extra filter reduces sample size. Conditions should make economic sense and be tested out of sample. See Overfitting and Curve Fitting.

Independence#

Events A and B are independent if P(A | B) = P(A). Coin flips are independent; trading outcomes often are not. Assuming independence when outcomes are related underestimates risk.

Checking a filter honestly#

When you add a condition, record how many trades remain and whether the improvement holds in a later period you did not use to choose the filter. A filter that lifts the win rate from 50% to 60% on 600 trades in one period, and from 50% to 58% in the next, is far more convincing than one that only works in the data it was built on. See In-Sample vs Out-of-Sample Testing.

Frequently asked questions#

What is conditional probability?#

The probability of an event given that another event has occurred, written P(A | B).

How is conditional probability used in trading?#

To measure how a setup's win rate changes under certain conditions, such as trends, volatility levels or news, which helps build filters.

What is the base rate fallacy?#

Ignoring how common an outcome is overall when judging a signal, which can make rare events seem more predictable than they are.

Next, learn to update beliefs with evidence in Bayes' Theorem.