Inside the Mind of a University Endowment Manager: Strategic Investment Philosophy and LP Perspectives

University of Michigan endowment manager Dan Feder reveals how institutional investors navigate uncertainty, select venture partners, and build resilient portfolios across decades.

Key Takeaways

• Endowments are pools of 13,000+ separate funds requiring 8% nominal returns forever with limited volatility
• Risk-based investing handles "known unknowns" while venture capital's power lies in "unknown unknowns" - true uncertainty
• "Adventure capital" for ambitious founders differs from "capital for ventures" - traditional financing for established companies
• Career path creation in endowment management shifted focus from long-term inputs to short-term demonstrable outcomes
• Concentrated portfolios of 36 relationships across venture and private equity enable better manager selection than diversification
• Fund size matters less than strategy alignment - both too big and too small can hurt returns depending on approach
• LP herd dynamics create vulnerability when markets shift, requiring independent thinking at the periphery for resilience
• Institutional advantages include access to information/people, long time horizons, and occasional ability to influence outcomes

Timeline Overview

• Career Origin (2000) — Accidental entry into endowment management through private equity introduction, moving from law to Princeton investment office
• Learning Phase — Working with Andy Golden at Princeton on venture and private equity, developing foundational investment philosophy
• Career Evolution — Moving through Washington University and eventually to University of Michigan, observing industry professionalization
• Current Strategy — Managing concentrated portfolio of 27 venture relationships within broader endowment allocation framework
• Future Outlook — Anticipating fundamental changes in endowment management best practices over next 5-10 years

Endowment Construction: Building from Zero

• University endowments are not single entities but pools comprising thousands of separate endowments - Michigan has over 13,000 individual funds, each with specific mandates like scholarships for first-generation students, operating like shares in a mutual fund structure.
• The fundamental requirements are threefold: support spending of 4-5% annually, keep pace with higher education inflation (above CPI), and preserve intergenerational equity without excessive volatility. This translates to needing roughly 8% nominal returns "forever" - an ambitious target requiring equity orientation.
• Portfolio construction starts with deepest, most efficient markets - public equities - then adds asset classes using modern portfolio theory to reduce volatility per unit of return. The modest 4-5% annual liquidity requirement enables allocation to illiquid assets like venture capital and private equity.
• The allocation process differs from investing: allocation uses risk-based frameworks analyzing historical correlations and volatilities to determine asset class weights, while investing within each bucket requires different approaches, particularly for uncertain assets like venture capital.
• Mean variance optimization provides mathematical foundation for asset allocation, but behavioral frameworks prove equally important by setting parameters that prevent overconfidence in any single area and guide portfolios toward model states that would make sense to outside observers.
• The theoretical progression moves from fully liquid public markets toward increasingly illiquid alternatives, with each step justified by improved risk-adjusted returns or correlation benefits rather than pursuing illiquidity for its own sake.

Risk Versus Uncertainty: The Frank Knight Framework

• Frank Knight's distinction between risk and uncertainty fundamentally changes how venture capital should be viewed within portfolios. Risk involves "known unknowns" - calculable probabilities from historical data - while uncertainty encompasses "unknown unknowns" that are not knowable in advance.
• Risk-based investing works well for traditional asset allocation but misses venture capital's true value proposition. "If you set up a system that's risk-based, around these known or knowable things, you're not going to know how to deal with or incorporate those unknown unknowns into your portfolio in a sensible way."
• Unknown unknowns manifest as entrepreneurs working on problems only they understand, creating opportunities that are not known to others and not knowable through traditional analysis. Successful venture investing requires accessing information others don't have about these uncertain opportunities.
• Risk-based alpha gets absorbed by markets over time as advantages become known and arbitraged away. Uncertainty-based investing enables "durable economic profit" because the information advantages remain sustainable when properly cultivated through unique access and relationships.
• This framework explains why traditional diversification within venture capital misses the point. Rather than selecting multiple similar firms hoping one excels, endowments should concentrate on accessing the few managers capable of trafficking in true uncertainty effectively.
• The cascading herd effect occurs when clear, risk-based opportunities attract capital easily while uncertain, uncomfortable opportunities get overlooked despite potentially superior long-term returns. LPs face similar dynamics with certain investments appearing better for longer periods regardless of underlying fundamentals.

Adventure Capital Versus Capital for Ventures

• The term "venture capital" originally shortened "adventure capital" when coined by Benno Schmidt Sr. at J.H. Whitney in 1946, suggesting a fundamental distinction that has been lost over time between adventurous investing and traditional financing.
• "Adventure capital" targets founders working on hard, ambitious problems where the future is not known or knowable, requiring genuine risk-taking and exploration. This aligns with endowment advantages around long time horizons, unique information networks, and occasional ability to influence outcomes.
• "Capital for ventures" represents traditional financing for companies past the truly adventurous stage - still valuable and necessary work, but different from betting on unknown unknowns. Many venture firms perform this function well but it requires different evaluation criteria.
• The adventure/capital distinction eliminates artificial stage-based categorizations (seed, early, late) because even later-stage companies can pursue genuinely adventurous, uncertain paths that merit different investment approaches and risk tolerance.
• This framework helps LPs focus their "spotlight" on the adventure end of the spectrum where their institutional advantages create sustainable competitive moats rather than competing in efficient markets for traditional financing opportunities.
• The distinction also explains why many VCs resist admitting they do traditional financing - both functions are valuable, but adventure capital requires different skills, time horizons, and risk tolerance that align better with endowment characteristics.

The Professionalization Problem in Endowment Management

• Historical endowment management attracted people from "oddball places" who found themselves in the field accidentally, creating strong mission alignment and intellectual curiosity without predetermined career advancement expectations.
• The emergence of defined career paths in endowment management created positive and negative effects: more talent attraction and professional development opportunities, but also pressure to demonstrate track records for advancement between institutions.
• Career progression now follows traditional financial services patterns: analyst to associate to principal to director to managing director to CIO, often requiring moves between institutions to create promotional opportunities.
• Short-term performance demonstration requirements shift focus from long-term inputs (relationship building, knowledge accumulation, network development) to near-term outputs (visible metrics, demonstrable relevance, quick wins).
• This creates misalignment with venture capital's inherently long-term nature where the best opportunities may take years to develop and mature. "If you're much more long horizoned in your thinking, you're going to tend to be more input oriented."
• The solution involves cultural and structural changes to reward input-focused work: "spend a lot of time listening to people who I think are smart and who are going to tell me what's coming next" and letting ideas "percolate to the surface" over years.

Concentrated Portfolio Construction and Manager Selection

• Michigan's concentrated approach involves 36 active relationships across venture and private equity (27 in venture, 9 in private equity), deliberately constraining relationship numbers to force better selection decisions.
• Instead of diversifying within categories hoping to catch exceptional performers, concentration requires picking "the one" in each area, forcing complementary selection across different approaches and return sources.
• This creates "cognitive dissonance" by backing managers with fundamentally different worldviews: "we're constantly investing with people who have very different views of the world. We have specialists and generalists. We have solo GPs, we have firms, we have larger firms."
• The advantage of intentional constraints: "it forces you to pick among groups that do something specific" rather than taking diversified allocations among similar approaches, leading to more thoughtful portfolio construction.
• Manager selection focuses on five core functions every asset class must execute: source opportunities, transact them, own investments well, exit effectively, and run firms without detracting from the first four capabilities.
• Recent thinking evolution favors backing more smaller groups over fewer larger ones: "instead of backing that firm, maybe we should be backing two or three firms of a couple of people or one or single practitioners" to maximize exposure to truly productive individuals.

Fund Size and Strategy Alignment

• The question "is fund size the enemy of returns" requires clarification: "do you mean too big or too small?" because fund size represents a symptom of strategy rather than an independent variable.
• Undercapitalization creates problems: seed and early-stage funds sometimes lack sufficient capital to execute properly, missing opportunities for meaningful exposure in companies that become important.
• Overcapitalization can force managers to make suboptimal decisions, watering down what they should be doing or decreasing quality of underlying company selection as fund size increases beyond strategy requirements.
• The ability to lead rounds at appropriate scale benefits both investors and founders, translating back to better investor outcomes. Being undersized relative to strategy hurts execution regardless of absolute fund size.
• Rather than having predetermined views on optimal fund size, the focus should be understanding strategy requirements and ensuring fund size enables rather than constrains effective execution of that strategy.
• Engagement around fund sizing decisions proves valuable: being "in the conversation about how are you thinking about this? What's the strategy?" helps managers make uncomfortable but correct decisions rather than taking easy paths.

Institutional Advantages and Strategic Positioning

• Michigan's three core advantages guide investment selection: access to information, people, and opportunities; genuine long time horizons; and occasional ability to influence outcomes through introductions, connections, or high-marginal-impact interventions.
• The university's scale creates unique positioning: "very large university with a deep and broad research function with alumni network that numbers in the many hundreds of thousands with a lot of loyalty to the institution."
• Investment evaluation requires alignment with one or more institutional advantages rather than pursuing opportunities that others could execute equally well. "Does that investment, whether it's a fund investment or something we're doing in a more direct sense, align with one or more of our three advantages?"
• Different institutions require different approaches: "a small liberal arts college is not going to have that same portfolio of advantages, but it'll have other things." No single recipe works across all endowments.
• The shift from allocation-focused to advantage-focused thinking represents a fundamental change: "allocating among asset classes is not the same thing as investing" and each institution must develop approaches that play to specific strengths.
• Future endowment management requires customization: "each of those institutions to really do this thing, which is very hard, that 8% nominal return with limited volatility forever, is not going to be able to look at a recipe book and say, 'This is how it's done.'"

Manager Selection and Relationship Building

• Sourcing operates exclusively through "introductions and qualified referrals" rather than trying to see the entire market, leveraging network advantages rather than attempting comprehensive coverage with limited team resources.
• The referencing and networking process "holds a lot of weight" in manager selection, emphasizing listening carefully to trusted sources and understanding their perspectives on potential partners.
• Some people are "just special" in ways that are "pretty glaringly obvious" when encountered, though this quality remains difficult to articulate beyond acknowledging its rarity and importance.
• The framework avoids overusing terms like "taste" which can become "a copout" or "a way for people to fall back to when they can't explain why they're making a decision" rather than doing necessary analytical work.
• Selection requires distinguishing where "the seat ends and the person begins" because great firms with great people include individuals who wouldn't succeed independently and vice versa.
• Portfolio construction increasingly favors backing fewer, more productive individuals rather than larger firms where only some partners drive returns: "what we're trying to do is get the most exposure to that productivity."

Herd Dynamics and Independent Thinking

• LP herd behavior creates vulnerability during market disruptions: "when stuff happens and things start moving around then that structure gets broken and the kind of the weaker investors or the herd mentality investor becomes exposed and highly highly vulnerable."
• The aspiration involves being "independent thinkers willing to be on the outside of that or on the perimeter of the herd" while maintaining self-sufficiency when markets shift or structures break down.
• Independent thinking doesn't require disagreeing with consensus: "just being an independent thinker doesn't mean that you disagree with the herd. It just means you should think independently. Sometimes that means the herd is correct."
• Strong independent thinkers around the periphery help guide broader institutional behavior during stable periods, but this structure fails when underlying conditions change and peripheral guidance becomes insufficient.
• The water buffalo and lions analogy illustrates how herd protection works until external pressures force movement, creating chaos and exposing the most vulnerable members who lack independent navigation capabilities.
• Maintaining independence requires developing internal capabilities for decision-making rather than relying on external guidance that may disappear during the most critical moments when independent judgment becomes essential.

Dan Feder's perspective reveals how sophisticated institutional investors think about portfolio construction, manager selection, and long-term value creation in ways that differ fundamentally from the narratives commonly shared in venture capital. His emphasis on uncertainty over risk, inputs over outputs, and institutional advantage over generic diversification provides a framework for understanding how the most thoughtful LPs approach their craft.