MONEY TALKS by Andrew Waitman
Due diligence of future events creates the potential danger of simplistic, trite and misleading conclusions. The farther out the events of interest, the substantially higher the cost of investigation and the noise or probability of error.
Cost and complexity rise rapidly with farsight
From SCAN's Print Edition
Due diligence is a critical part of the investment process for investors. Due diligence refers to investigating and verifying the facts and claims related to a business. This may include financial data, the backgrounds of individuals, contracts with customers, technical claims and a myriad of other matters material to the current state and future potential of a company. What this investigation encompasses and the efforts expended is germane to this discussion.
Gathering facts and validating claims on the current state of a business is a relatively straightforward review process. From a financial due diligence perspective, auditors perform this task every year on behalf of the board of directors for all public companies and most private companies. Financial and legal due diligence is well covered ground in audit, banking, investment and legal circles. I will not retread that ground here.
My observations and conclusions are related to future-oriented due diligence, particularly concerning customer and channel commitments leading to market share, size, growth and competitive dynamics.
Historical company information often can be validated with 100% certainty if you are willing to spend the time, effort and money. Customer invoices, contracts, cash transactions and so on can all be researched and verified by independent investigators. However, the economic cost of achieving 100% certainty often far exceeds the value of that information. Auditors, investors and investigators typically use a random or pseudo-random sampling methodology to ensure the cost of investigating and validating does not exceed some reasonable economic threshold.
Facts and future events are different animals. In due diligence they must be treated so. How does this information/cost trade-off affect investigations of claims that are about events in the future? The economic cost of “certainty” increases rapidly with time. This should be intuitive. In order to increase certainty of information for future events the efforts (and associated costs) of improving certainty for farther and farther out events increases at some unknown but accelerating and time-related rate. Knowing something with absolute certainty one month from now will be expensive but not impossible to achieve. However, as time increases between now and the future unknown outcome, the number of variables with a causal relationship to the outcome begins to expand at some rate. In order to investigate how each of these variables will contribute to the final outcome there is a cost of variable investigation outcome. The complexity is exacerbated as these variables inter-relate affecting each other in positive and negative ways.
For example, the variable that a person will want to carry around music on small battery powered device depends on the variable of “how much they are willing to spend on this device,” which is loosely impacted by the variable of “current macro economic conditions.”
Furthermore, the variable “how easy it is to transfer their current music collection onto this device” is connected with the variable “how many of their friends are doing this at any point in time” and so on and so on. Numerous and often increasing numbers of variables over time affect the final outcome.
In addition to the difficulty of identifying all the variables necessary to investigate a future outcome and the associated costs of investigating each variable’s outcome, there is of course an inherent error or noise function that is introduced with each causal variable. This error increases with time and reduces certainty while increasing the cost of acquiring sufficient information to improve predictive ability. The noise arises as a result of the fact you are not investigating the actual event but merely some future variable event that may not occur as “claimed” during the investigation. For example, the answer the VP of Engineering gave you changed between the time of questioning and the time of some future event. Context evolves.
Imagine you need to investigate the probability of a company reaching its annual revenue target one year from now. That revenue will be derived from a finite set of customers making decisions during the course of the next 12 months. A priori, the exact number of customers and specifically which non-customers will become customers are unknown factors. Even greater complexity surrounds the information challenge if you are investigating a new product or new product category. How many potential customers must you sample to get an indication of actual purchase interest? What if you fail to explain the product sufficiently? Does asking the question without “money-on-the-line” commitment bias the answer positively/negatively? How large does your sample size need to be to improve accuracy of the outcome?
Due diligence of future events creates the potential danger of simplistic, trite and misleading conclusions. The farther out the events of interest, the substantially higher the cost of investigation and the noise or probability of error. An illusion of knowledge or conviction based on thin-slicing the set of all possible variables that impact future events is the scourge of the venture business. The illusion delusion leads to adverse selection. Adverse selection is selecting companies that turn out to be poor investments, or avoiding investing in start-ups that turn out to provide substantial venture returns. If investment firms are competent at due diligence, a debatable conclusion, why is portfolio performance so poor in the majority of venture funds?
How to avoid the illusion delusion will be the topic of my next column. Ensuring investors focus on what they can know with certainty and what they cannot know a priori is the art of venture capital.