Central Tendency

How Central Tendency Works: Mean, Median, and Mode

There are three primary tools used to find the center of a dataset, each with its own specific logic and utility. The "Mean" is calculated by summing all the values and dividing by the count. It is the "center of weight"—if you put all the numbers on a balance beam, the mean is where the beam would balance. In finance, the mean is used for calculating "Expected Value" and is the basis for most portfolio optimization models. However, its greatest weakness is that it is pulled aggressively toward extreme values (outliers). One "fat finger" trade can drastically change the mean price, even if the rest of the market didn't move. The "Median" is the true middle value. To find it, you simply sort the data from smallest to largest and pick the one in the center. The median is far more "robust" than the mean because it doesn't care about the size of the extremes. If a stock has returns of -5%, -2%, 1%, 3%, and 1,000%, the median return is a modest 1%. The mean, however, would be a misleadingly high 199%. This is why the median is the preferred measure for describing data sets with heavy "tails," such as income distributions or home prices. The "Mode" is the most frequently occurring value in the set. While less common in general finance, it is a powerful tool for day traders. In the "Volume Profile" indicator, the "Point of Control" (POC) is the mode—the price level where the most trading volume occurred during the day. Traders view this as the "fairest value" of the session, as it represents the price that the highest number of market participants accepted. When the price moves away from the mode, it is considered to be "searching for a new value."

Important Considerations: Skewness and Distribution

The relationship between the mean and the median reveals the "skew" of the data, which is critical for risk management. In a "Normal Distribution" (the famous Bell Curve), the mean, median, and mode are all the same. Most financial models assume a normal distribution for stock returns, implying that positive and negative surprises are equally likely and symmetric. However, real-world financial data is often "Skewed." In a "Positively Skewed" distribution (Right Skew), a few massive positive outliers pull the mean above the median. This is common in options trading, where most trades expire worthless (low median), but a few provide 100x returns (high mean). In a "Negatively Skewed" distribution (Left Skew), the mean is lower than the median. This is the "picking up pennies in front of a steamroller" scenario common in insurance or short-volatility strategies: most days you win small (high median), but one catastrophic event wipes out all the gains (low mean). For an investor, simply knowing the "average" (mean) return is not enough. You must understand if that average is being propped up by a few lucky events (Positive Skew) or if it hides the risk of a total wipeout (Negative Skew). Analyzing the gap between the mean and the median allows you to see the true "shape" of your risk before you deploy capital.

Comparison: The Three Measures of Center

Each measure of central tendency provides a different perspective on the "typical" value.

Common Misconceptions in Statistics

Avoid these common errors when interpreting measures of central tendency:

• "Average" always means Mean: Many reports use the word average to mean Median, especially when discussing housing or income. Always verify which one is being used.
• The Mean represents the "most likely" outcome: Only if the data is perfectly bell-shaped. In skewed markets, the mode or median is often more likely.
• Mean reversion is guaranteed: Just because a price is far from its mean doesn't mean it must return. Sometimes the "mean" itself is moving because of a fundamental change.
• A single measure is enough: You can never fully understand a dataset with just one number. You must also understand the "spread" (standard deviation).