Michael Mauboussin on Luck, Investment Process and Valuing Uncertain Businesses

Harry Stebbings welcomes Michael Mauboussin, head of consilient research at Counterpoint Global, part of Morgan Stanley Investment Management. Mauboussin describes his work as helping the investment team with process, researching and writing, and sharing ideas externally. He calls the pattern input and output: learn something useful, then communicate it.

Mauboussin defines luck as an event affecting a person or organization that can be good or bad and could reasonably have turned out differently. He distinguishes that from randomness at the level of a whole system. In a class coin-toss exercise, the random sequence is the system; a particular student guessing several correctly has experienced luck.

To have a chance of winning a lottery, one must buy a ticket and risk losing. The important question is expected value: an outcome can be possible without making the bet attractive. He describes a lottery as negative expected value in monetary terms and says people might still play for enjoyment, not as a guaranteed route to a return.

People tend to credit success to their own skill and blame failure on external forces, Mauboussin says. Different activities sit at different points on a skill–luck continuum, and separating the two is difficult. Chess is closer to the skill end, a lottery to luck; many investment decisions fall somewhere in between.

Stebbings asks whether luck misattributed to skill might actually improve an investor’s confidence. Mauboussin notes that venture has historically shown more persistence in fund performance than some other asset classes. One possible mechanism is preferential attachment: an early success, even partly lucky, may attract strong founders and opportunities to a well-known firm.

That feedback can coexist with skill, and Mauboussin says returns are widely dispersed across venture firms. He urges an investor to ask how much of a track record belongs to the individual and how much to the institution’s name and access. His comments on persistence and historical returns are broad research interpretations, not predictions about a particular fund.

A sound process begins with a clear idea of the advantage one is trying to create. Mauboussin contrasts highly systematic, frequent trading with Warren Buffett’s selective, research-intensive approach: both can be successful while operating very differently. There is no universal process; the method should fit the investor’s proposed edge.

Stebbings describes combining a media platform that introduces him to founders with a small investment committee of experienced operators. Mauboussin says the process is consistent with that stated advantage if the participants contribute relevant knowledge and it can be repeated. A process should be economically grounded, manage bias and evolve as the world changes.

Stebbings prefers three committee members to two because the conversation feels less like a personal argument. Mauboussin adds that an odd number can permit a decision while retaining a dissenting voice. This is a practical group-design suggestion, not a guarantee that three people will always decide well.

Two recurring process failures are departing from the stated method in an unsystematic way and letting biases enter the decision. Mauboussin highlights overprecision—believing one knows the future more narrowly than is justified—and confirmation bias, which favors evidence for a preferred view while discounting contradictions.

To counter confirmation bias in thesis-driven investing, he recommends writing specific signposts and probabilities before making a decision. When events diverge from the expected path, the team should stop and ask whether there is a good explanation or whether the thesis should change. A trusted decision partner can help enforce that precommitment.

Good team decisions require attention to size, cognitive diversity and the way discussion is managed. Mauboussin says different views often exist in the room but remain unheard. Leaders can invite the most junior participant to speak first, use a premortem and actively encourage disagreement instead of letting the highest-paid person’s opinion dominate.

Stebbings collects independent written views before a meeting so people do not simply follow a shared document. Mauboussin suggests a premortem: imagine that an investment has failed in one or two years, and have each person write a plausible explanation. The group can then identify risks and consider how to respond if they materialize.

A process should adapt when new information changes the problem. Mauboussin describes updating a prior view in the right direction and magnitude as difficult. For example, growth in intangible investment changes how accounting records spending, so a simple rule that all losses are bad can mislead; equally, losses alone do not prove a company is making valuable investments.

When he says everything is a discounted cash flow, Mauboussin does not mean every seed investor should build an elaborate model. The practical question is how a company could ultimately make money and generate cash. A strategic acquirer or public-market investor eventually has to assess the value of those future economics.

Young companies also have real options: uncertain future opportunities that capable management may develop if capital is available. He says a leading position may expand the alternatives, but even an exercised option eventually has to support cash flows. His framework is a way of thinking about economic value, not an exact forecast for an early startup.

Stebbings points out that consumer products can grow before their eventual monetization model is known, and premature models may cause investors to miss them. Mauboussin agrees that the initial model can be uncertain, but says an investor seeking financial returns must believe some way to monetize can eventually emerge. Google’s early search audience is his example.

A seemingly high early valuation cannot be assessed only by the frequency of losses. Mauboussin describes research on the skewed distribution of venture outcomes: many investments lose money while a small number can produce very large gains. An investment can lose in hindsight and still have had positive expected value at the time; that depends on probabilities and payoffs, which are uncertain.

He compares this mindset to betting on combinations with low frequency but potentially large payouts, rather than frequent small wins. The analogy illustrates payoff distributions, not an assertion that every speculative investment is attractive. Investors must consider the chance and size of outcomes before committing capital.

Stebbings worries that more capital entering venture increases prices and reduces returns. Mauboussin agrees that competition for assets can pressure expected returns. He describes venture’s historical public-market-equivalent results as episodic and notes that exit routes also matter; his comments about 2020–2022 are assessments made at the time, not current performance data.

On when to sell a winner, Mauboussin says there is no easy universal answer. As a company matures, an investor may be able to estimate a range of values and reduce a holding when the price meets or exceeds it. The competing case for holding is that a small minority of public companies has driven much of long-term market wealth, but identifying those winners in advance is hard.

Stebbings compares a very broad portfolio with a concentrated, hands-on model such as Benchmark’s. Mauboussin says both involve different activities and constraints. Hands-on early investing depends on the supply of exceptional founders and investors able to help them; he suspects that capacity is limited and explicitly says venture portfolio construction is not his area of expertise.

For younger investors questioning their ability during a downturn, he suggests revisiting their process instead of treating short-term price changes as a verdict on skill. He also cautions that macroeconomic variables are difficult to forecast. Investors should understand how rates and economic conditions affect a business, while concentrating on analysis they can actually perform.

Mauboussin recalls starting on Wall Street shortly before the 1987 crash and working on margin calls sent by telegram. The collapse felt catastrophic at the time. In hindsight, the economy continued to function and the stock market recovered, but he offers this as historical perspective rather than a promise that every crisis resolves the same way.

He observes that advances can be gradual while declines feel sudden, and that volatility tends to cluster rather than appear evenly through time. During the 2022 discussion he expected the volatile period eventually to give way to a quieter one, without specifying when. That was a view about market behavior, not a timetable for recovery.

Stebbings worries about the combination of pandemic, geopolitical conflict and supply-chain disruption. Mauboussin agrees the events are disruptive and notes that globalization changes their effects, but points to earlier pandemics and wars as historical parallels. He does not argue that the specific circumstances or consequences are identical.

On the 2022 repricing of public companies, he says pandemic demand pulled some sales forward and made planning difficult even for established retailers. Whether prices then represented a new normal was uncertain. He explains that when expected cash flows are far in the future, valuations are particularly sensitive to changes in real interest rates.

He cites a historical increase in US real yields of roughly 240 basis points during the period discussed. Higher real rates can lower the present value of long-duration assets. Companies with pricing power may better offset nominal inflation, he argues, especially when they deliver value above the price customers pay; none of this guarantees a stock’s return.

He considers whether rates were returning toward a more historically typical range, while avoiding a precise market prediction. He recalls a 10-year US Treasury yield below 1% during work completed in 2020. These figures and his interpretation belong to the historical discussion and should not be read as present-day rates.

For a favorite book, Mauboussin chooses E. O. Wilson’s Consilience, which argues for bringing disciplines together to address complex problems. He links that idea to his work and involvement with the Santa Fe Institute. The appeal for him is the unification of knowledge alongside the scientific practice of breaking problems into parts.

He credits investor Bill Gurley’s curiosity, communication and early grasp of cash flows, returns on capital and increasing returns. They worked in the same research department early in Gurley’s career. Mauboussin’s assessment is personal and reflects a longstanding professional relationship.

Asked about his own strengths and weaknesses, he chooses curiosity as a strength and a tendency toward abstraction rather than concrete detail as a weakness. He says revisiting a familiar research topic can still teach him a great deal, an approach consistent with his emphasis on continual learning.

His favorite investment advice is “this too shall pass”: both exuberance and fear can dominate judgment in the moment. Asked where he would put $100,000, he says he would probably buy equities, while explicitly saying he cannot predict the next month or several months. That was his personal view in 2022, not a recommendation for an individual reader.

He recalls buying Sears Holdings because he liked the CEO’s thinking and process; the investment nevertheless turned out badly. He distinguishes an unfavorable outcome from proof that every part of the decision was wrong. The example also reinforced for him how a difficult business can overwhelm even capable management.

He describes aiming for eight hours of sleep and exercising daily, routines that he says became easier after his children left home. He finds them helpful to his own cognitive performance. These are personal habits, not clinical advice or evidence that they ensure investment success.

He continues working because he enjoys the intellectual stimulation, autonomy and opportunity to be useful. Teaching at Columbia Business School is another way to turn learning into output. The best teachers, he says, remain students themselves: they keep learning their subject and improving how they communicate it.

Stebbings thanks Mauboussin for the conversation, and Mauboussin thanks him in return. Their discussion ends as a complete interview, spanning decision process, valuation, market history and the habits behind his research.