Scott Wilson on WashU’s Endowment, Concentration and Co-Investing

Harry Stebbings welcomes WashU chief investment officer Scott Wilson and asks what he brought from Grinnell. Wilson credits David Clay and describes a concentrated approach combining external managers with direct and co-investments: find strong partners, add capital to their most compelling opportunities and let successful positions compound.

Wilson clarifies a story about his arrival at WashU. His team submitted redemption requests for more than half the portfolio in the first week, after he had reviewed it for months; the larger turnover of more than 75% happened over roughly 18 months, including private secondary sales.

The board had asked for meaningful performance improvement within three years, he says. The public portfolio needed substantial changes, while the private side was in better shape. The team exited all 36 or 37 hedge funds, a transition he compares to turning an aircraft carrier rather than a speedboat.

Wilson says the managed pool was just under $8 billion when he arrived and is now just under $20 billion. He views public and private holdings through a common long-term investment lens, though he acknowledges the very different liquidity available to the institution.

The broad allocation target is roughly half public and half private, with liquidity requirements shaping that mix. WashU pays out around 5% on a smoothed basis. Without that obligation Wilson says he might target 70–80% private assets, but he emphasizes that private distributions cannot be counted on exactly when needed.

He illustrates the constraint: a 5% payout from a portfolio that is half liquid consumes the equivalent of 10% of the liquid portion annually. Wilson reports stronger private than public compounding over the prior decade for his portfolio, while arguing that some private markets offer less crowded opportunities.

More than half of WashU's private portfolio is in direct investments and co-investments, according to Wilson. The team finds external partners, studies companies they identify and adds capital to selected individual names. Team members are generalists rather than a separate direct-investing group.

Building this capability requires internal expertise and time with companies. Wilson estimates that 80–90% of his team's time is spent meeting individual businesses, even though it also meets managers. He credits direct and co-investments for much of the portfolio's outperformance relative to peers.

Wilson cautions against treating the results as proof that his team sources every opportunity independently. A trusted manager has already identified a company; WashU starts there and conducts its own work to test the thesis. The team sometimes disagrees and distinguishes partners by the quality of the signals they provide.

Asked about concentration, Wilson distinguishes original cost from current portfolio value. He says SpaceX grew to more than 20% of the portfolio at one point. The team advocated retaining exposure through opportunities to sell and used continuation vehicles, with board governance supporting that decision.

For a new single-name investment, Wilson says the team generally notifies the board above a certain cost threshold and needs approval at a higher one. He also argues that a concentrated position inside a fund can represent a much smaller look-through exposure for the endowment as a whole.

Concentration has also produced painful outcomes. Wilson mentions substantial investments in automated and additive manufacturing companies that had not worked out as hoped, while saying their final outcomes were not yet known. He says the direct share of the private book could rise, subject to the natural need for fund partners who supply opportunities.

Before committing directly, the team needs to understand and independently assess a thesis, monitor the company and prefer it to other uses of capital. Wilson says it passes when the team lacks the expertise to distinguish a company from competitors, and invests in only a tiny fraction of the opportunities it sees.

One of Wilson's largest stated mistakes was not adding to winners once they became large positions or rose in price. He believes a model combining concentrated manager relationships with direct exposure can scale without simply adding managers until each contributes very little.

Wilson says WashU looks through funds to the underlying companies rather than counting funds as if they were independent exposures. He reports that roughly half of the portfolio is in its top 50 look-through positions. His view of diversification is a portfolio-management judgment, not a universal rule.

The team wants enough conviction for a meaningful eventual position, but may start with a smaller co-investment and add over time. A small fund commitment can make sense when it creates access to an interesting company, although Wilson says many tiny commitments are difficult to scale.

WashU adds managers infrequently, sometimes one or two in a year. Wilson says its co-investment model does not fit mega-funds as well: they may not need its capital, and a larger fund must either make more investments or write larger checks, often at higher valuations.

Stebbings challenges that view, pointing to exceptionally large AI and space outcomes that might support venture returns at greater scale. Wilson agrees such outcomes can make it possible, but questions prospective multiples and the long-term return on capital-intensive AI infrastructure.

Wilson says WashU has many fund relationships above $1 billion but none above $10 billion. Its preference is to partner with smaller managers early, learn about the underlying companies and build meaningful direct positions over time. He says the team is less equipped to underwrite the very earliest stages without customers or visible economics.

Rather than wait for a manager's investor-relations team to offer a co-investment, Wilson says his team proactively asks which portfolio companies excite the manager and meets them. If conviction develops, it may seek participation in a later round, an employee tender or another direct purchase.

He believes relying only on inbound syndicated deals can create adverse selection. Looking back at WashU's manager book, he says its proactive approach to finding companies has served it better than waiting defensively for opportunities to arrive.

Stebbings asks about the cycle and private valuations. Wilson sees similarities with previous periods of excessive pricing and believes late-stage private companies can sometimes command a premium to what public investors will pay. These are his current market judgments and predictions, not established future outcomes.

Stebbings asks why successful companies would list if they can raise privately and provide employee liquidity. Wilson answers that public markets are deeper pools of capital, even though private markets have grown. They discuss, without resolving, whether a future listing could change demand for other public assets.

Wilson worries that some AI-related companies will not survive a decade and says their long-term competitive advantages and profit pools are uncertain. He argues that for many tasks customers may eventually switch from expensive frontier models to cheaper alternatives. The claim is his forecast, not a conclusion about every AI business.

WashU has exposure to Anthropic through fund partners. Asked why it has not sold, Wilson says those holdings sit within broader funds and relationships rather than easily sold standalone vehicles. He says the endowment can consider hedging if a company becomes public and the valuation warrants it.

Stock distributions are handled case by case. Wilson says WashU continued holding Nubank years after receiving public shares, while other positions may be hedged or reduced if he believes valuations are excessive. For a very large position such as SpaceX, he describes taking some risk off over multiple years while retaining exposure to potential upside.

The conversation turns to Sequoia's evergreen structure. Wilson says the arrangement did not fit WashU's co-investment model and that the decision was discussed with the board. He acknowledges the firm's record and says passing may have been a mistake viewed solely through subsequent returns, while maintaining that fit mattered.

Wilson rejects a rigid rule that an institution must back a manager for three funds regardless of what it learns. Because the team follows underlying companies, it can reassess a relationship if its view of the portfolio diverges substantially from the manager's; he says such breaks are uncommon.

The most frequent reason he gives for dropping a manager is insufficient valuation discipline. Stebbings points out that exceptional companies can look expensive in real time. Wilson agrees the judgment is difficult but says his concern usually comes from a repeated pattern rather than one expensive deal.

He is not inherently opposed to rapid deployment if the expected returns justify it. He prefers a smaller fund that can write meaningful early checks and return to investors sooner over a much larger vehicle, because a single successful investment can matter more to the smaller fund.

WashU aggregates look-through exposure across managers and direct positions, tracking how much capital ultimately sits in each underlying company. Wilson says that view matters more to the team than discussing fund-level performance alone.

Discussing Cerebras, Wilson says WashU reduced and hedged part of a large position around its IPO while retaining exposure. He expresses confidence in the management team but says the valuation requires substantial future performance. His comments are an investor's assessment, not a recommendation.

Asked what he would change, Wilson says he might smooth university payouts over a longer period because a 5% distribution can be difficult in some markets. He recognizes the competing obligation to support the university's present mission and says the ability to see that impact matters.

Wilson expects a shakeout in parts of AI but cannot time it. He points to cyclical industries where earnings can surge when prices rise, then normalize as the cycle changes. Stebbings presses him on whether high private valuations are consistent with a preference for private holdings.

Wilson distinguishes the long-term hypothetical without liquidity constraints from the current opportunity set. He says private holdings had reached around 70% at fiscal year-end, but some became liquid and the team was redeploying proceeds into public companies, including sectors it sees as less exposed to AI infrastructure.

Over his nine years at WashU, Wilson says the original roughly $8 billion generated about $17 billion in profits while roughly $5 billion was paid to the university. He credits the board, administration and team, and says he remains because he values the people and the work.

Returning to mistakes, Wilson says anchoring on an earlier purchase price or an already-large position stopped him from adding to some high-conviction winners. He estimates that missed additions cost billions, while also acknowledging mistakes in the other direction: adding to companies that did not recover.

He regrets delaying a large secondary sale after markets rose in 2020–21. The team waited for higher reported values and missed an opportunity to sell at prices he now considers attractive. He describes the choice as greed and notes that subsequent write-downs reversed part of the earlier gains.

Wilson says WashU finds selected opportunities in Europe, including fintech and non-technology companies. He argues that comparable businesses can trade at lower multiples there than in the US, while Stebbings asks whether a persistent regional discount should simply be treated as the market price.

Wilson distinguishes price from his estimate of value, acknowledging that value is subjective. His argument is that a lower entry price for a business with comparable long-term growth can improve returns. This is an investment thesis rather than a guarantee that a valuation gap will close.

WashU has a Shanghai office and has invested in China, Wilson says, while acknowledging geopolitical and regulatory risk. He favors understanding local conditions and working with local partners; Stebbings challenges him with the possibility of abrupt policy changes and limits on foreign capital.

In a rapid-fire round, Wilson names Benchmark as a fund he wishes WashU had backed and praises Lee Fixel of Addition for disciplined underwriting. He also names several peer endowment teams he respects and says there is no regular formal dinner for all chief investment officers.

His closing advice is to excel at habits he considers broadly available: show up prepared, remain curious, work hard, ask good questions and treat people kindly. Stebbings adds that modern research tools make preparation easier; Wilson agrees that these habits should be easier to practice.