top of page

Understanding Binary Outcomes with Uneven Distributions



In investing and probability, you often encounter binary outcomes - two possible scenarios such as a market going up or down. However, the chances of each outcome are not always equal or 50/50. There can be an uneven distribution of probabilities between the two outcomes. As an investor, understanding these uneven distributions is key for assessing risks and returns.



What is an Uneven Distribution?


An uneven distribution refers to when the probabilities of two outcomes are skewed heavily in favor of one outcome over the other. For example, there could be a 20% probability of outcome A happening and an 80% probability for outcome B. This is very different from a 50/50 coin flip scenario.


Examples of Binary Outcomes with Uneven Distributions


There are many examples of uneven binary probabilities in investing and business:


  • Merger or acquisition deals - Most potential deals end up falling apart rather than being completed. One study found only 20% of negotiated M&A deals end up closing while 80% fail to go through.

  • Clinical drug trials - Drug compounds often fail in clinical trials. Only about 10-15% of drugs make it through all three trial phases and end up getting approved.

  • Venture capital investments - About 75-80% of VC investments do not return investor capital while 20-25% generate large returns. This uneven distribution allows VCs to offset losses through home run investments.

  • Options contracts - Roughly 80% of options end up expiring worthless while around 20% of options finish in-the-money at expiration. Traders sell more options contracts than they buy to benefit from this uneven distribution.


Implications for Investors


The asymmetric probabilities of these binary outcomes have important implications for investors and risk management:


  • The lower probability event often carries higher risk and higher potential reward. Concentrated bets increase chances of large gains or losses.

  • Uneven distributions allow portfolio approaches like venture investing where a few large returns offset multiple losses.

  • Underdogs beat odds sometimes. Low probability does not mean no chance, especially over longer time periods.


Quantifying Uneven Distributions


As an investor, how can you quantify and evaluate uneven binary probabilities? A few useful approaches include:


  • Use Historical Frequencies: Analyze historical frequencies of events to estimate future probabilities. For example, if a merger closed successfully in 15 out of 100 prior attempts, you could estimate a 15% probability of success. Past frequencies give reasonable base rates to judge future outcomes.

  • Employ Conditional Probability: Conditional probability assesses the odds of an outcome by placing certain conditions. For venture investments, what is the probability of a positive return conditioned on the company reaching an IPO? By fixing certain conditions, conditional probability better quantifies subsets rather than overall binary outcomes.

  • Model with Decision Trees: Decision trees graphically represent uneven probability distributions. Each branch endpoints in a binary outcome with branches weighted by probability. Decision trees allow modeling complex combinations of conditional, sequential, and independent probabilities.

  • Set Confidence Intervals: Statistics like confidence intervals quantify the variance around probability estimates. Confidence intervals grow larger with less data or random outcomes. Investors use confidence ranges to size positions based on distribution extremes.

  • Adapt Over Time: Do not anchor uneven distribution assumptions as conditions change. Monitor outcomes to detect shifts in probabilities. Some venture firms have increased late stage investing as more startups reach mature milestones. Uneven distributions remain uneven but with evolving probabilities.


Key Takeaways for Investors


Here are a few key takeaways about uneven binary outcomes:


  • Assume binary outcomes have uneven probabilities absent data suggesting otherwise.

  • Gather frequency data to quantify probabilities and confidence ranges.

  • Use conditional probability and decision trees to model complex sequential outcomes.

  • Adapt probability estimates as new condition information arises.


While most investments have binary win/lose outcomes, smart investors use history and analytics rather than hopes and intuition to quantify the uneven distributions between binary results. An analytical approach to uneven distributions allows better risk management.


By understanding uneven probabilities in binary outcomes, investors can better size positions, hedge risks, and evaluate strategic payoffs. An analytical approach helps when outcomes may appear binary but have very different distributions between the two results.

17 views0 comments

Comments


bottom of page