Attribute bias is a characteristic of quantitative techniques or economic models whereby they tend to choose investment instruments that have similar fundamental characteristics. Some models used in finance will tend toward attribute bias, and investors should be aware of this as part of choosing a balanced portfolio.
Attribute bias should not be confused with attribution bias, a finding in behavioral economics whereby people blame others for their own mistakes or faults.
Attribute bias is simply a characteristic that is likely to happen unless models and techniques are specifically designed not to include it.
Because of this bias, a model or statistical technique may lead to concentrated positions.
Attribute bias describes the fact that securities that are chosen using one predictive model or technique tend to have similar fundamental characteristics. This makes sense, because a model that looks for specific sets of data points will only return investment instruments with those similar parameters.
Attribute bias is neither positive nor negative. It’s simply a characteristic that is likely to happen unless models and techniques are specifically designed not to include it. The danger in choosing a portfolio using a model with attribute bias is that the portfolio may contain similar securities, which can amplify market downturns. Attribute bias leads to an unbalanced portfolio. Most investors prefer a balanced portfolio to protect themselves from sudden or extreme movements of the market.
One way to correct for attribute bias and choose a balanced portfolio is simply to use several different models to choose securities, and use different parameters for each model. Each model can have attribute bias, but since the investor has balanced the parameters of the different models, the portfolio will be balanced even if each smaller subset of securities is not.
Let’s say you are an investor wanting to build a portfolio of stocks growing their revenues 20%+ per year and with growing earnings. You also add in technical factors that find stocks that also have strong recent performance. By setting these parameters, you may expose your portfolio to concentration in stocks that behave similarly.
Maybe your portfolio is heavy in growth areas like Discretionary and Technology. If those sectors face a rotation out of growth, you could be hit with steep losses due to over-concentration.
While attribute bias refers to a bias in the methodology of picking financial instruments for a portfolio, self-attribution bias refers to a bias a person can have that causes them to think that the success they have in business, choosing investments or other financial situations is because of their own personal characteristics. Self-attribution bias is a phenomenon in which a person disregards the role of luck or external forces in their own success and attributes success to their own strengths and work.
Attribute bias is a neutral concept, and is used as a descriptor to give information about how a group of securities was chosen. If attribute bias causes problems with a portfolio, understanding that it exists allows those problems to be corrected. In contrast, self-attribution bias is a negative phenomenon that can lead to skills deficits in the short term and failures over the long term for a person who has self-attribution bias. It is inherently a negative bias, and should be corrected if the person wants to maintain success in investing.