top of page

Understanding Bradford-Hill’s Criteria for Causation: A Guide for Investors

Updated: Feb 13



Investors are consistently challenged to identify the determinants of financial success. Is it the company's growth rate, management quality, or perhaps the current economic environment that significantly impacts an investment's performance? Knowing what causes something to happen is vital for decision-making and future predictions. In the field of medical epidemiology, Bradford-Hill's criteria serve as a set of principles to establish a causal link between a particular factor and an outcome. In a similar way, these criteria can be adapted to guide investors in their decisions by distinguishing between correlation and causation.



The Bradford-Hill Criteria: An Overview


Originally proposed by Sir Austin Bradford Hill in 1965, the Bradford-Hill criteria aim to provide a framework for understanding the relationship between an exposure (e.g., a drug or toxin) and an outcome (e.g., a disease or health condition). There are nine criteria:


  • Strength of Association: A strong relationship between cause and effect is more likely to be causal. Example: Investors looking into renewable energy stocks might observe a strong correlation between government subsidies and company profitability. A strong association might be an indication that subsidies significantly impact profitability.

  • Consistency: The cause-effect relationship should be consistently observed in different settings. Example: If every time the Federal Reserve cuts interest rates, a particular bond fund increases in value across multiple market conditions, that's a consistent cause-effect relationship.

  • Specificity: The cause should lead to a specific effect, not multiple, unconnected effects. Example: A particular software company’s stock price rises only when they release a new product. If the stock price isn’t similarly affected by other market factors, specificity is established.

  • Temporality: The cause must precede the effect in time. Example: In the case of an acquisition, if the acquirer's share price consistently rises after the acquisition announcements, temporality is established. The announcement precedes the stock's increase.

  • Biological Gradient (Dose-Response Relationship): Increasing the dose of the "cause" should increase the likelihood of the "effect." Example: An increase in marketing spending by a company consistently leading to a proportionate increase in revenue can be an example of a dose-response relationship.

  • Plausibility: The relationship must make biological or logical sense. Example: A retail company's increased revenue during holiday seasons is plausible due to increased consumer spending during holidays.

  • Coherence: The relationship should not seriously conflict with existing theories and laws. Example: If your hypothesis is that lower interest rates boost real estate investment trusts (REITs), this should align with the known economic theory that lower borrowing costs can lead to increased property purchases and rentals.

  • Experiment: An experiment that removes the "cause" should result in a reduction or elimination of the "effect." Example: If a tech company changes its monetization strategy from a one-time payment to a subscription model and sees a consistent increase in long-term revenue, then eliminating the "one-time payment model" has resulted in an "increase in long-term revenue."

  • Analogy: Similar causes should have similar effects. Example: If Company A in the tech sector performs well when it invests in R&D, and Company B is very similar to Company A, it's reasonable to expect that Company B will also benefit from investing in R&D.


Limitations and Caveats


It's essential to remember that not all criteria have to be met for a causal relationship to exist. The criteria serve as a guide rather than a definitive checklist. Financial markets are complex and influenced by multiple variables, making it challenging to establish causation definitively.


The Bradford-Hill criteria, although developed for medical research, provide a useful framework for investors seeking to establish causative relationships between variables. Understanding these criteria can sharpen investment decisions and improve the quality of financial analysis. By identifying these causal relationships, investors can make more informed decisions, minimize risks, and optimize returns. Therefore, applying Bradford-Hill’s criteria for causation can add another layer of rigor to the investment decision-making process.

9 views0 comments

Comments


bottom of page