David Morehead on Endowment Liquidity, Venture Returns and Portfolio Construction

Baylor University chief investment officer David Morehead says the endowment’s work matters amid pressure on higher-education finances. He points to demographic changes, competition for domestic students and difficulties facing some international students, and expects endowment distributions to become more important.

Morehead says his office has historically focused on downside protection, citing periods when Baylor held up better than the broader market. Over the past several years, the team has also tried to improve participation in rising markets; his examples are his account, not a complete performance comparison.

He describes a value-oriented, high-quality equity book that can lag in momentum-driven markets. Rather than relying solely on costly put protection, the office has sought more upside exposure through portfolio construction while trying to control the cost of carrying that exposure.

Many endowments invest through pooled funds, where a manager runs one portfolio for many limited partners. Morehead says the average risk-and-return profile of such a fund may not match Baylor’s needs at a particular time.

Baylor has therefore asked some managers to run separate accounts for the university, using a similar strategy but allowing greater visibility and input. This gives the endowment a way to consider exposures across managers instead of treating each fund in isolation.

His example is Nvidia: if another manager wants to add it when Baylor already has ample exposure, the office might decline. If the portfolio has none and the exposure fits its needs, it might ask for more. He says the approach has worked well recently, without offering a controlled comparison.

On a blank-sheet portfolio, Morehead says he spends extensive time on allocation, sometimes more than on choosing individual managers. At the interview, he puts Baylor’s portfolio at roughly 45%–47% private and 53%–55% public assets.

He starts with private investments because they consume liquidity and are difficult to rearrange. An institution should decide how much illiquidity it can tolerate, then construct the rest of the portfolio around that constraint; secondary sales can help at the margin but are not frictionless.

Baylor’s private allocation range is 35%–55%, with roughly 45% as a target, he says. The upper end is meant to leave room for a ‘denominator effect’: when public holdings fall, private investments can become a larger percentage even without new private purchases.

The aim is to avoid becoming a forced seller during a difficult market. Morehead says Baylor’s private share rose into the low 50s during previous stress without forcing sales or eliminating its ability to allocate. Future conditions can of course be different.

The team reassessed private investments as it anticipated the university’s future distribution needs. Morehead says Baylor has allowed some real-asset exposures to wind down and has focused on venture capital, expansion and growth equity, and buyouts, seeking returns sufficient to justify the lockup.

On established venture firms versus emerging managers, he says Baylor arrived later than some larger endowments and has not always obtained access to prominent names. He credits colleagues with strong results in expansion and growth equity, while saying the venture book includes newer managers.

Morehead likes venture but questions the length of some funds and whether general-partner incentives always match an endowment’s need for compounding and cash distributions. A large investment multiple is incomplete information without the time needed to earn it.

His illustration compares 15 times capital over 18 years with three successive investments each returning three times capital over six years. The latter compounds to 27 times only if the full proceeds can actually be reinvested on those exact terms; it is an arithmetic example, not a forecast.

Holding a winner longer can make a fund’s headline multiple look better, he says, but the endowment needs dollars to support students. He wants to evaluate the velocity and compounding of capital, not just the eventual multiple or a manager’s fundraising narrative.

Stebbings asks why Baylor invests in venture at all if faster growth-equity realizations can be attractive. Morehead says the office considers diversification across stages and the timing of returns, with some investments potentially paying out sooner and others much later.

The office has a rule against discussing returns without mentioning time, he says. Five times capital over five months and five times over 30 years describe very different experiences, even though the headline multiple is the same.

Morehead says venture also gives Baylor exposure to companies that may grow exceptionally. He reports that around 2.5% of the endowment is exposed to Anthropic through managers, while saying Baylor did not have exposure to SpaceX or OpenAI at that point. These are his reported positions at the time.

When an underlying private holding goes public, Baylor does not apply a single rule to sell immediately or hold indefinitely. Morehead says it considers the company, the resulting position size, the whole portfolio and the associated risk before deciding.

Stebbings asks how Baylor responded to a sharp decline in software stocks in early 2026. Morehead says his contribution is understanding human behavior: he spoke with business owners about whether they would replace essential software with unproven AI-generated alternatives.

The business owners he contacted were reluctant to remove systems central to their operations. Morehead infers that trusted software vendors may also become a channel for delivering AI features. This is his investment thesis, based partly on conversations, rather than proof that every incumbent is protected.

He says software had fallen substantially from its prior highs, creating an opportunity in the team’s view. Rather than pick every company himself, he allocated to a specialist manager and discussed which holdings seemed least vulnerable to AI disruption.

Morehead sees such decisions as part of an allocator’s job: decide where the next dollar belongs while relying on managers for their domain expertise. He describes frequent exchanges with the manager about individual companies, market information and relative opportunities.

Stebbings asks whether venture investments are a learning tool for the endowment. Morehead says he learns more from public-market managers and is skeptical of some technology forecasts that ignore the time required for regulation, adoption and execution.

Morehead argues that public markets incorporate decisions by many participants, whereas a private financing may be set by a much smaller group. Stebbings counters that public markets can also be irrational. Morehead agrees prices can be wrong, but considers the broader price-discovery process useful.

On private valuations, Morehead wants managers to avoid optimistic marks. An inflated internal price can distort decisions: an offer below the mark may appear to be a loss even if it exceeds a more realistic valuation. He sees disciplined marks as a way to reduce that psychological bias.

He cites Baylor’s reported gains between prior marks and exits as evidence that its managers have tended to be conservative relative to the comparisons he uses. Those figures are his account of the portfolio, not independent verification of any specific mark or valuation method.

Morehead says growth equity is Baylor’s largest private allocation and that the team is comfortable with its managers. He favors its return profile and believes it can have fewer complete losses than earlier-stage venture, although outcomes vary by fund and vintage.

He cites a roughly 30% annualized figure for the growth-equity book as he understands it and compares it with the endowment’s hurdle. This is a performance claim from the interview, not a promise or an independently audited result.

Bad vintages are part of investing, Morehead says. His team sets overall allocations and expected return hurdles for the private portfolio, while colleagues choose managers and construct the book within those parameters.

Baylor distributes about 5% annually, according to Morehead, supporting scholarships, professorships and other university needs. That spending requirement is a concrete part of the liquidity plan, separate from capital the team wants available for new opportunities.

He also assigns cash an opportunity value: if an investor expects to find an attractive investment during the next several years, liquidity allows that investor to act. His example adds an estimated opportunity to cash yield; it is a heuristic dependent on future opportunities, not a guaranteed cash return.

Morehead says Baylor entered the pandemic with around 15%–16% cash because the team did not see enough attractive uses for it at the time. In the more recent environment, he says the balance has been lower because it has identified more opportunities.

Asked about mistakes, he does not name a particular trade. Instead he says losses and extended drawdowns are part of investing, and that experience taught him not to go all in: an asset can fall further even after what looks like a steep decline.

When Baylor added to software after a decline, the team did not put every available dollar to work at once, he says. It used a planned process to allocate through a difficult market and reduce the influence of emotion; the process does not eliminate the possibility of loss.

Morehead describes thinking in increments of roughly ten percentage points of market decline. In his illustration, the team might invest a portion after a 20% decline and more at lower levels. The actual allocation depends on the situation and available liquidity.

The market may rebound before the team invests the entire intended amount, so this approach can miss part of a recovery. Morehead accepts that trade-off to avoid exhausting capital just before an asset falls further.

Decisions on individual stocks are generally in managers’ hands, but Morehead talks with them about downside scenarios and relative risk-adjusted opportunities. During the software decline, he says, he and a manager exchanged information frequently and discussed what they would do if prices fell further.

He says a manager may concentrate a portfolio during a drawdown when some positions appear more attractive. Morehead is comfortable with some additional concentration because the endowment owns a wide range of assets overall; that does not make any individual position risk-free.

A distinctive choice at Baylor’s size, he says, is recruiting many team members from undergraduate ranks. Waco’s location makes it harder to attract and retain some mid-career professionals for a long period, whereas graduates have already chosen the school and area.

The trade-off is years of training and a heavy early burden on the leader. Morehead says he initially emphasized the benefit of a stable team but had underestimated the time required to develop junior staff.

Hiring experienced professionals presents the opposite trade-off: they can contribute more quickly, but a team may face greater turnover. He prefers to describe these as alternative risks rather than say colleagues choosing a different model are wrong.

On endowment compensation, Morehead says the work requires a mission orientation. He describes his motivation as expanding opportunity for future Baylor students, while acknowledging that personal financial rewards matter to people too.

Asked about AI and education, Morehead worries that delegating too much thinking to a tool could weaken human reasoning. He mentions research but explicitly says he does not know its validity; the conversation does not establish a medical or educational finding.

He also believes education can find beneficial uses for AI and describes the issue partly as one of human discipline. Stebbings then asks about taxes on large endowments; Morehead says Baylor’s endowment per student is below the relevant threshold as he understands it.

Morehead jokes that he would welcome the tax if it came with a much larger endowment. The remark expresses his aspiration for the fund rather than an assessment of any particular institution’s tax liability.

On comparisons with peers, he says Baylor both tracks others and follows its own goals, including increasing the endowment per student. He acknowledges disappointment with relative performance in one year and describes changes to the portfolio that he believes will take time to show results.

He says the team had increased private commitments and experienced a ‘J-curve’ effect as investments were made before returns were realized. Morehead compares the restructuring to changing an engine while the car was moving; the explanation is his account of performance drivers.

Morehead estimates a substantially stronger return for the then-current year and expects structural tailwinds, while noting particular high-profile positions Baylor did not own. This is an estimate and outlook from the interview, not a verified final result or guarantee.

For private position sizing, the team works backward from the amount it wants exposed to each underlying company. A large fund multiple may barely affect the endowment if Baylor’s original share was tiny, so the office wants a meaningful dollar exposure.

He gives an illustrative target of roughly $2.5 million to $3 million per underlying company in some strategies. If a manager expects ten companies, Baylor might size its commitment accordingly, while leaving actual investment decisions to that manager.

Morehead wants a manager to carry out the role for which it was hired. He compares portfolio construction to fielding a baseball team: if a third baseman changes positions without discussion, the team may have an unintended gap even if that player performs well elsewhere.

Stebbings asks whether a manager who changes strategy and earns strong returns should still be retained. Morehead says he wants a conversation before a material change and wants evidence the manager has skill in the new activity; fit with the whole portfolio matters.

His public-market example distinguishes a fully invested stock picker from a manager who deliberately changes cash exposure. Both can be judged against a benchmark, he says, but he does not want a manager with no relevant track record to begin timing cash unexpectedly.

He distinguishes a minor move between adjacent financing stages from a fundamental change in approach. Moving from investing after product-market fit to backing untested teams with no product is, in his view, a different mandate and deserves explicit discussion.

Asked whether an LP needs to commit to three funds to judge a venture manager, Morehead says Baylor has often taken a long view because one or two funds may not provide enough evidence. He says his team is strongest when it has data to analyze and less comfortable with purely narrative cases.

If Baylor had a much larger endowment, Morehead is unsure that its principles would change, but says implementation would become harder. He talks with other investment offices about when asset size starts to constrain access to smaller opportunities.

At very large scale, even an exceptional return on a small commitment can have limited portfolio impact. The speakers use a hypothetical $20 million investment returning 50 times capital to illustrate that $1 billion is significant but only a small percentage of a $40 billion or $50 billion endowment.

That arithmetic helps explain the appeal of large investment platforms able to accept bigger commitments. Morehead nevertheless believes producing attractive returns over the required period becomes harder as a fund grows, because a larger amount of capital must be put to work.

The interview turns to data centers. Morehead says local opposition is a real issue in places where facilities are built, even though Baylor has investments connected to the sector. In his view, valuable sites increasingly need not just land and power but also permits.

He identifies electricity and water costs as concerns for residents, especially in dry regions. He expects better solutions for water to help acceptance, while believing power constraints may persist; actual resource use and price effects depend on local projects, infrastructure and tariffs.

Morehead says permitted and powered sites in Baylor’s portfolio have risen in value and describes utilities offering earlier power to some projects as others are delayed. These are his reported portfolio observations, not independently verified market-wide price changes.

Asked what share of projects will fail, he says he is not sufficiently expert to provide a reliable number. He identifies permitting and community opposition as important current bottlenecks, based on what he sees in the portfolio.

The speakers compare permitting in the US, UK and China. Morehead says Baylor has a permitted UK site that he considers valuable because of the difficulty of obtaining permission. Their broad geopolitical comparisons are opinions and do not establish how any specific foreign project was approved.

Morehead is bearish on Europe at the broad market level, citing policy, security and technology concerns. That does not prevent Baylor from using European long-short managers: he expects differences between companies and says some of the endowment’s macro hedges involve European indices.

In a quick-fire round, he says Baylor increased software exposure, reduced some energy exposure after a rise in crude prices and added to private-equity sponsors. These describe decisions he reports having made, not instructions for another investor.

He calls private credit overhyped in his view. His concern is that some credit investments may have significant downside exposure to a business without offering equity-like upside. He acknowledges he is not personally invested in the category, and individual structures and risks differ.

Morehead praises Brown’s endowment team and says its long-term record has earned his respect. He mentions Benchmark as a venture fund he would like access to, while treating both comments as personal assessments rather than recommendations.

Looking ahead, he is excited about biotechnology and says Baylor is spending more time studying the field. He points to scientific progress and possible diversification, while acknowledging the team is still considering whether to increase an existing exposure.

Morehead is also focused on building the investment office as the endowment grows. He wants systems that can track a larger portfolio and sustain the team without losing the creativity that helped it develop.

He describes the transition from a smaller to a larger endowment as requiring many decisions about staffing, processes and investment scale. His hope is that over the following few years the office will emerge better prepared for further growth.

Stebbings thanks Morehead for an unusually candid conversation about the allocator’s perspective. Morehead closes by describing the team’s aim to do a good job for Baylor from central Texas.