# Herd behavior

Herd behavior is the tendency to follow other investors’ actions instead of independently evaluating a trade, potentially reinforcing buying or selling pressure.

## What it means

When investors see many others buying or selling, they may treat that activity as a reason to do the same. Sometimes the crowd has relevant information; sometimes it is reacting to attention, uncertainty or other traders’ actions. Similar trades do not, by themselves, prove that anyone copied anyone else.

Herd behavior describes a decision-making pattern, not a price forecast. It can appear during a widely discussed news event or in a rapidly moving stock. It differs from *momentum*, which describes a price trend: investors can follow a trend independently, and herding can occur without a sustained trend.

## How it is calculated

There is no single formula that diagnoses herd behavior from an individual trade. Researchers may compare how often investors take the same side in a stock with how often they take that side across the broader market. One simple, illustrative measure is:

`H = |B / (B + S) − P|`

- `H` is the absolute difference between two buy proportions, measured in percentage points when multiplied by 100.
- `B` is the number of observed investors buying the stock during a specified period.
- `S` is the number of observed investors selling it during the same period.
- `P` is the proportion of observed investors buying across a broader comparison group during that period, expressed as a decimal from 0 to 1.

The calculation requires investor-level observations and a clearly defined comparison group. It is an illustration, not a standard diagnostic test. A large difference shows unusually one-sided activity relative to that group; it does **not** establish why investors acted alike. Public trading volume alone does not supply the investor counts needed for this example.

## Worked example

Suppose a hypothetical data set records decisions by 100 investors in one stock over a day. Of those investors, 80 buy and 20 sell. Across a broader set of observed stock decisions that day, 55% are buys.

1. Find the stock’s buy proportion: `80 / (80 + 20) = 0.80`.
2. Subtract the broader buy proportion: `0.80 − 0.55 = 0.25`.
3. Take the absolute value: `H = 0.25`, or **25 percentage points**.

The stock’s observed decisions lean more heavily toward buying than the comparison group. This does not show that the 80 buyers followed one another. They might have responded independently to the same company announcement. A different observation period or comparison group could produce a different result.

## Why it matters before you trade

Following a crowd can substitute visible activity for a stated reason to trade. Before placing an order, separate the information you can verify from signals such as social-media attention or a fast price move. Consider whether the proposed position fits your intended basket and risk limits, even if many other investors appear to agree.

Crowded activity may also coincide with changing execution conditions. A wide bid-ask spread raises the cost of crossing the spread; an order large relative to daily volume can face greater estimated market impact. Neither condition proves herding, but both matter to a pre-trade check when interest in a stock is intense.

## How PreTrAIde uses it

PreTrAIde does **not** claim to calculate a herd-behavior score or determine whether other investors are copying one another. Its pre-trade analysis checks a user-built stock basket before an order is placed. The quantitative layer estimates factors including spread, participation in daily volume, market impact, volatility and portfolio concentration. Its AI layer reasons over quantitative scores, recent news and analyst ratings, returning a GO, NO-GO or MODIFY verdict with plain-language reasoning. These verdicts are decision support, not recommendations.

That check can help a user examine execution and basket risk when a trade idea is attracting attention. It cannot establish that the crowd is right or wrong, or remove behavioral bias from the user’s decision.

## Common mistakes

- **Equating a price rise with herding.** A price can move because investors independently respond to new information.
- **Treating heavy volume as proof.** Volume shows trading activity, not investors’ motives.
- **Assuming the crowd must be wrong.** Shared decisions may reflect useful information; herding alone gives no reliable direction for a future price.
- **Ignoring the comparison group.** A buy-heavy stock may be less unusual on a day when buying is widespread across the market.

---
Source: https://pretraide.com/glossary/herd-behavior