Anchoring Bias in Pricing Models

Anchoring bias in pricing models happens when an algorithm is overly reliant on a specific, initial piece of information, such as a historical price level or a specific indicator, when making new decisions. This "anchor" prevents the model from adjusting to new information as effectively as it should.

In options pricing, for example, if an algorithm is anchored to a specific implied volatility level, it may fail to adjust when the market moves to a new volatility regime. This leads to mispricing and potential losses.

To mitigate this, models should be designed to constantly re-evaluate their assumptions based on the latest data, rather than being tied to past benchmarks. This is essential for maintaining accuracy in fast-moving markets.

Being aware of where an algorithm is "anchored" is a key part of maintaining its predictive integrity and adaptability.