The cap table looked clean. Forty-two million dollars raised, reasonable liquidation preferences, founder equity properly structured. I reviewed it for the better part of an afternoon before I found the problem buried in a footnote: unassigned intellectual property. A portfolio of patents developed before the company's formation, never formally transferred, sitting in limbo between the founder's personal estate and the corporate entity.
The deal didn't fall apart that day. It fell apart eight months later, in the middle of a $190 million exit, when the acquirer's diligence team found the same footnote. The science was sound. The team was excellent. The market thesis was correct. None of that mattered once a structural defect became visible under the light of a serious transaction.
I've watched versions of that story play out enough times to have internalized the lesson: in capital allocation, the quality of your case determines the outcome. Not the strength of your conviction. Not the credibility of your narrative. The case.
Walk into any investment office and you'll hear the same language: pattern matching, gut instinct, seeing around corners. These are descriptions of intuition dressed as method. The investors who generate consistent returns across multiple cycles do something different.
They build evidence stacks. They set burden-of-proof thresholds before the debate begins. They size positions the way a careful analyst sizes risk — not according to enthusiasm but according to what the evidence actually supports.
This is not a metaphor borrowed from a courtroom. It is a method developed over 27 years of evaluating companies, structuring transactions, managing professional teams across complex situations, and learning the hard way what separates a defensible decision from a wishful one.
The investors who generate consistent returns across multiple cycles do something different. They build evidence stacks. They set thresholds before the debate begins.
The Three Thresholds
Legal reasoning distinguishes between levels of evidentiary confidence: probable cause, the preponderance of the evidence, and beyond a reasonable doubt. Each standard requires a different quality and quantity of proof. Each justifies a different level of commitment.
Investment decisions follow identical logic, and the failure to apply it is responsible for more capital destruction than any market cycle or macro event I have witnessed. The problem is not that investors make bad bets. It is that they make large bets on thin evidence, because conviction feels like certainty and narrative feels like proof.
The framework is straightforward. Each threshold supports a different level of capital commitment.
Probable Cause: Opening the Investigation
The lowest threshold requires basic evidence that an opportunity exists. Market problem identified. Technological solution theoretically viable. Regulatory or competitive barriers not immediately fatal. This level of evidence justifies one thing: dedicating resources to a deeper investigation. It does not justify writing a check.
Consider a clinical-stage oncology company with a novel mechanism of action targeting a validated pathway. Probable cause means the biology makes sense, the unmet need is real, and the team has the scientific credentials to run the program. That is a case worth opening. It is not a case worth concentrating capital behind. Too many investors stop the analysis here and call what they have found a thesis.
Preponderance: Building the Business Case
The intermediate threshold requires evidence that this specific company can capture this specific opportunity, profitably, within a capital structure that protects investors. This standard supports a meaningful allocation but acknowledges that significant execution risk remains.
For the oncology company, preponderance means Phase 1 or 2 data with a legible efficacy signal, a capital runway sufficient to reach the next value inflection, a clean IP assignment chain, governance that can manage a founding CEO's natural tendency toward mission over arithmetic, and a realistic path to liquidity that doesn't depend on a single acquirer.
Each element demands different evidence. The clinical data comes from the scientists. The capital structure analysis comes from the cap table and the term sheets. The governance assessment requires sitting in a board meeting and watching how the room handles bad news. You cannot outsource that last one to a data room.
Beyond Reasonable Doubt: Concentrated Conviction
The highest threshold requires evidence that success appears probable given current trajectories, with identified and bounded downside. This standard supports concentrated positions — the bets that generate returns meaningful enough to matter.
Very few opportunities reach this threshold. That is the point. An evidence-based approach to investing is, by design, selective. The investors who generate the best long-term returns are not those who see the most opportunities. They are those who wait for the evidence to accumulate before committing.
The investors who generate the best long-term returns are not those who see the most opportunities. They are those who wait for the evidence to accumulate before committing.
The Elements Every Investment Must Prove
Every legal claim requires the plaintiff to prove specific elements. Investment opportunities work the same way. Four elements appear in every investment case I have built or evaluated.
Market: Does genuine willingness to pay exist at the scale required? Usage metrics and expressed interest are not evidence. Signed contracts, pilot program results, retention data, and price sensitivity studies are evidence. In life sciences, this is the reimbursement question. Weak evidence on this element should stop a case before it advances.
Technology: Can the proposed solution work at the scale and cost structure required for the business to succeed? This requires expert assessment, not market research. In biotech it means sitting with the scientists and understanding what the data actually shows, not what the slide deck says it shows. Theranos failed on this element with devastating consequences. The narrative was compelling. The evidence, when eventually produced, was not there.
Execution: Can this specific management team navigate the specific challenges this company will face? Track record is relevant but insufficient. The better question is whether the team's decision-making holds under pressure. I evaluate this by examining how leadership has handled prior setbacks, how they respond to challenge in the room, and whether the board has the independence and authority to intervene when necessary.
A company with a good market, a real technology, and a team that cannot execute is still a failed investment.
What makes this element genuinely difficult is that execution risk looks different depending on the domain. In enterprise software, the core execution question is commercial: can this team sell into complex organizations and build recurring revenue? Having co-founded and operated technology companies before transitioning to life sciences, I find those failure modes are almost entirely about go-to-market discipline.
In regulated life sciences, the execution question is different in kind. The team must navigate clinical protocols, regulatory correspondence, and investor expectations simultaneously, often while managing a founding scientist whose identity is fused with the science in ways that create governance fragility.
Having evaluated both, I find that the investors who assess execution risk most accurately are those who can name specifically which execution challenges apply to this company in this industry at this stage. Execution is not a single element. It is a set of domain-specific competencies.
Economics: Can the business model generate attractive returns at scale given realistic capital requirements and competitive dynamics? Many investments that pass the first three elements fail here. The business works. The returns, given the capital required to get there, do not.
Weak evidence on any single element should be treated as a case-killer, regardless of how strong the other elements appear. You close those elements or you don't invest.
Setting the Burden Before the Debate
The most common failure mode in investment decision-making is not insufficient evidence. It is moving the goalposts. The evidence looks thin, so the threshold gets quietly lowered. The team is impressive, so the technology questions get softer. The market is exciting, so the economics get hand-waved.
The solution is to set the burden of proof before the investment conversation begins, when the room is not yet emotionally committed to an outcome. This requires naming, explicitly and in writing, what evidence would be sufficient to justify each level of capital commitment.
Once those thresholds are defined, the conversation that follows has a different quality. Questions become clarifying rather than adversarial. Data requests become targeted rather than exhaustive.
The people in the room know what they are trying to prove, which means they also know when they have not proved it.
This also creates a discipline around position sizing that is otherwise difficult to maintain. Small positions when evidence reaches probable cause. Meaningful allocations at preponderance, with acknowledged execution risk. Concentrated bets only at the highest threshold, and only when the downside is bounded.
Set the burden of proof before the conversation begins, when the room is not yet emotionally committed to an outcome.
What the Evidence Doesn't Solve
An evidentiary framework is not a formula. It does not eliminate judgment. It creates the conditions under which judgment can be exercised well.
The four elements I described require expert assessment. Understanding whether a Phase 2 efficacy signal is meaningful requires scientific fluency that most generalist investors do not have. Evaluating a management team's decision-making under pressure requires having watched them make decisions under pressure. Assessing governance requires having sat in governance situations yourself.
The cases that get built well are the ones where the investor can challenge every element at the level of depth the evidence requires.
Not just asking whether the IP is assigned, but understanding what it would take to contest the assignment if it later became disputed. Not just reviewing the clinical data, but understanding the statistical methodology well enough to identify what the p-value is actually telling you.
The goal is not certainty. Certainty is not available in early-stage investing or in most consequential decisions. The goal is a defensible decision — one that you could explain to a skeptical audience using the evidence you assembled, the standard you applied, and the reasoning you used to reach the conclusion.
The Verdict Reflects the Case
Returning to that cap table with the footnote: the lesson I took from that transaction was not about IP assignment protocols, though I have been meticulous about them ever since. The lesson was about what happens when evidence-gathering stops at the point where the narrative becomes compelling.
The deal looked good at the level of the story. Strong science, experienced team, clear market need. It looked good at that level because no one had asked hard questions about the footnote, and the footnote was not in the story. It was in the document. Those are not the same thing.
Evidence-based investing is, at its core, a commitment to the documents over the story. Signed contracts over verbal commitments. Cap table snapshots over founder assurances. Clinical data over scientific enthusiasm. Governance structure over cultural narrative.
It is slower than conviction-based investing. It is less exciting. It produces fewer heroic moments in investment committee meetings. It also produces fewer disasters of the kind that footnotes generate.
The investors who consistently generate exceptional returns are not the ones with the best pattern recognition or the highest risk tolerance. They are the ones who know how to build a case, how to test it against hostile questions, and how to size their commitment to match the evidence they actually have rather than the outcome they hope for.
In the court of capital markets, the verdict always reflects the quality of the case you built before you walked in the door.