The Twenty Minute VCRyan Akkina: How MIT Builds Their Venture Fund Portfolio & How MIT Approach Direct Investing | E1109
CHAPTERS
- 0:00 – 0:47
Why venture fund investing is harder now: complexity, competition, and capacity
Harry and Ryan open on whether it’s harder than ever to invest in venture funds. Ryan explains why the old playbook—back a handful of proven Series A firms—no longer works as the ecosystem has exploded by stage, geography, and business model. They also discuss venture’s secular shift toward more capital, more attention, and lower margins.
- •Venture has become more complex across stage, geography, and specialization
- •Brand alone no longer guarantees meaningful access or check sizes
- •Industry shift from boutique to more commoditized, capital-heavy environment
- •Competition is a secular trend with only cyclical relief
- 0:47 – 6:44
Ryan’s path to MIT: engineering roots, Stanford entrepreneurship, and discovering the LP world
Ryan shares how early experiences in chip design and Stanford’s startup ecosystem shaped his interest in technology and investing. After consulting briefly, he sought investing exposure with mission-aligned work and found the endowment/foundation LP model. A job posting led him to MIT in 2009, where he’s stayed for 15 years.
- •Early technical background at HP/Intel and Stanford MS&E
- •Exposure to VCs/entrepreneurs via Stanford’s BASES and early YC days
- •Chose not to start in VC due to “right to win” considerations
- •Consulting detour, then pivot to endowment investing for breadth and purpose
- 6:44 – 8:34
Access dynamics today: why even MIT can miss top funds (and what emerging managers offer)
They dig into how even elite institutions can miss important managers and later face small allocations and capacity limits. Ryan notes great firms prioritize early believers and existing LPs. In a tougher fundraising environment, emerging managers can be particularly attractive because dedication and quality become clearer while competition and valuations improve.
- •Late entry into top funds often means small, potentially uneconomic allocations
- •Existing LPs get priority as firms fill capacity earlier than before
- •Down cycles surface who is truly committed to building a franchise
- •Better entry valuations and less crowded markets can favor new managers
- 8:34 – 13:15
How MIT evaluates managers: “See, Pick, Win… and Service”
Ryan outlines MIT’s core framework for underwriting venture managers: see deals, pick the best, win allocations, and then service founders to sustain future access. They debate which dimensions matter most, with Ryan emphasizing that in today’s competitive market, winning and founder service can be the main differentiators. Track record is contextual—more useful when mature, but less decisive for new funds.
- •Evaluation framework: see (deal flow), pick (selection), win (allocation), service (founder support)
- •In competitive stages, “win” and “service” often decide outcomes
- •Early-career managers require more qualitative judgment than track record
- •Founder-manager fit and likability meaningfully affect term-sheet selection
- 13:15 – 16:45
What goes wrong: early signals, luck, and why strong funds go sideways
Ryan reflects on mistakes and the challenge of detecting outcomes early—often you can’t tell in the first three years, with clearer signals emerging around years 3–5. They discuss why previously great funds deteriorate, including over-scaling, arrogance, loss of humility, and fading motivation. MIT also evaluates whether fund size remains aligned with team capacity and historical deployment ability.
- •True fund quality often only shows after 3–5 years via fundamental company traction
- •Emerging manager success can hinge on a luck-driven credibility feedback loop
- •Failure modes: growing too big, leaving the sweet spot, arrogance, loss of motivation
- •Capacity diligence: number of great GPs, capital per GP, historical pace of deployment
- 16:45 – 18:35
Managing relationships: when to scale down and how to tell a GP you’re not re-upping
They address the delicate balance between LP loyalty and recognizing that every firm has a “half-life.” Ryan explains how MIT tries to anticipate decisions early and communicate non-reinvestment before it becomes a surprise. The reality is that even great firms can outgrow their strategy or face generational-transition risks, forcing LPs to adjust exposure.
- •Default posture is loyalty, but no firm stays great forever
- •Decision triggers: fund size vs. strategy fit, team capacity, and long-term return potential
- •Best practice: communicate non-re-ups early to avoid surprise
- •Scaling down is part of disciplined portfolio management
- 18:35 – 21:21
Case study: backing Greenoaks early—references, founder love, and a small initial bet
Ryan recounts nearly passing on Greenoaks Fund I due to a conservative bias toward established brands. After sleeping on it, he pushed for a small initial commitment—a decision that became one of MIT’s most successful relationships. Key inputs included exceptional founder references (notably Coupang’s founder), an initial thesis, and a gut-level bet on the people.
- •Initial internal hesitation due to lack of “brand” at fund launch
- •Decision shifted to a smaller starter bet to earn relationship exposure
- •Founder reference calls were unusually strong and differentiating
- •Importance of underwriting people and adaptability, not just stated thesis
- 21:21 – 23:36
Building MIT’s venture fund portfolio: core vs. mid-tier vs. option bets
Ryan breaks down MIT’s venture fund position sizing into three buckets: a small set of core relationships with very large checks, a broader middle bucket with moderate checks, and a newer, very small-check bucket for extremely early managers or strategic access to co-invest. He also shares how deployment levels swing with the cycle, shrinking meaningfully from peak years to today.
- •Core relationships: ~6–8 managers, ~$50M–$150M checks per fund
- •Middle bucket: ~20 managers, typically ~$10M–$20M per fund
- •New micro-bucket: ~$1M checks for very early managers or co-invest access
- •Annual private commitments are cyclical; recent pace is materially lower than peak
- 23:36 – 25:42
The 2021 speedrun: compressed deployment timelines, LP leverage, and learning discipline
They discuss managers compressing deployment into ~12 months during the bubble and the myth of temporal diversification. Ryan agrees more discipline was needed but highlights the practical constraint: if LPs punished everyone for 2021 behavior, there would be few managers left to back. The focus becomes intellectual honesty about mistakes and improved process going forward.
- •Compressed deployment is a key regret area from the bubble period
- •LPs hesitate to criticize scarce-capacity top firms
- •Discipline is difficult when “everyone did it,” making exclusion impractical
- •Best response: candid post-mortems and process correction, not denial
- 25:42 – 28:44
Liquidity, denominator effects, and the long wait for exits
Ryan explains how improved public markets reduce pressure, but the bigger issue is delayed liquidity from large private companies like Stripe. MIT must manage mandated outflows (e.g., ~5% spending rule) and models liquidity carefully, but the last cycle created unusual strain: fast fundraises plus slow distributions. They also consider risks when IPO lockups expire and long-tenured investors seek exits.
- •Denominator effect pressure has eased somewhat as publics recovered
- •Primary constraint remains lack of liquidity events from mega private holdings
- •Endowment spending requirements add planning complexity, managed via modeling
- •IPO/lockup dynamics may pressure post-IPO prices as pent-up selling emerges
- 28:44 – 29:33
Secondary sales: why MIT rarely sells fund positions (price vs. necessity)
Harry asks about selling fund stakes to create liquidity. Ryan says MIT considers it but almost never sells because discounts often imply the buyer is underwriting a high IRR—suggesting MIT should hold instead. Only a truly forced liquidity situation would justify selling at unattractive prices.
- •Secondaries often clear at prices that transfer upside to the buyer
- •MIT sells only if pricing is compelling or liquidity need is acute
- •Holding is preferred when the buyer’s return math looks too attractive
- •Liquidity management aims to avoid forced sales
- 29:33 – 32:53
Why directs matter: conviction, structures, and how MIT underwrites co-investments
Ryan explains why direct/co-investing is strategically valuable: it diversifies tools for alpha, and it can be easier to write large checks into known companies at reasonable valuations than into blind pools. He highlights the importance of strong downside protection via structuring (e.g., senior notes) and doing real diligence—often by building familiarity with companies well before a deal appears.
- •Directs offer higher-conviction deployment than blind-pool commitments
- •Structured instruments (e.g., senior notes) can protect downside while keeping upside
- •Best outcomes come from a ‘prepared mind’—knowing company/founder ahead of time
- •MIT emphasizes real underwriting to avoid the ‘messy middle’ of half-diligence
- 32:53 – 33:43
Allocations and speed: sizing direct bets and moving fast on time-compressed deals
They discuss wide variation in direct check sizes—from a few million when constrained to $120M+ when conviction and structure support it (e.g., Coupang). Ryan notes that sometimes MIT can decide in hours because of trust in the GP, prior knowledge of the company, and excellent structure. Discipline still matters: they’ll say no on pricing, even if it risks fewer future looks.
- •Position sizing ranges widely based on certainty, liquidity, and access
- •Coupang example: large check with strong outcome; Snowflake example: limited allocation despite desire
- •Fast decisions rely on trust, prior diligence, and attractive structure
- •Saying no can risk relationship friction, but discipline and NPV thinking prevail
- 33:43 – 38:59
Incentives and governance: why endowments struggle with risk-taking—and a better model
Harry challenges whether endowment incentives match the value created (especially on directs). Ryan agrees incentives often aren’t aligned, crediting culture and leadership for enabling risk-taking at MIT. If designing a family office, he’d require meaningful personal co-investment from team members and introduce performance incentives above a hurdle, while acknowledging governance constraints in institutions.
- •Endowment compensation rarely mirrors the economics of standout wins
- •Culture and CIO leadership can substitute for direct financial incentives
- •Family office blueprint: meaningful personal capital at risk + performance incentives over hurdles
- •Governance aims to prevent “free options” on institutional capital, limiting carry-like pay
- 38:59 – 45:55
Learning from mistakes and evolving as an investor: omissions, OpenAI, and portfolio risk controls
Ryan shares a painful direct-investing mistake in oil & gas, emphasizing brittleness in leveraged, commodity-exposed models and the danger of businesses requiring constant reinvestment. He also highlights that biggest mistakes are often omissions—passing on OpenAI due to structure and technical uncertainty, where a small bet on a special founder might have been right. He closes with portfolio-level targets (public/private mix, cash levels) and drawdown beta as a risk lens.
- •Oil & gas loss: leverage + commodity exposure + operational brittleness can spiral quickly
- •Lesson: be wary of models demanding continuous balance-sheet reinvestment (parallels in other sectors)
- •Biggest regrets are often omissions; OpenAI pass driven by structure and perceived technical speculation
- •Portfolio controls: ~50/50 public-private, 5–10% cash, drawdown beta target ~0.75
- 45:55 – 52:51
Fundraising advice, LP base construction, and firm durability (including China and transitions)
Ryan advises managers to treat fundraising like an enterprise sales funnel and to build relationships without forcing premature evaluation by picky institutions. They discuss LP concentration limits, IR tradeoffs, and how top firms still raise despite a tougher market while emerging managers struggle. The conversation also covers reduced China exposure due to regulatory and sensitivity concerns and what strong generational transitions require: mentorship and timely economic sharing.
- •Fundraising is a numbers-driven funnel; relationship-building matters, but timing evaluations is critical
- •LP base: avoid over-concentration; too many LPs increases relationship-management burden
- •IR can work if it preserves GP time while maintaining access and transparency
- •China: reduced activity, heightened sensitivity and regulatory uncertainty
- •Generational transition: mentorship and earlier, fairer economic sharing to retain talent
- 52:51 – 57:59
Quick-fire: hard tech, pro-cyclicality, independent thinking, and time as the scarce resource
In quick-fire, Ryan describes becoming more open to hard tech after Tesla/SpaceX-style outcomes. He critiques LP pro-cyclicality and emphasizes humility about ‘crazy’ ideas being right but mistimed. He also notes MIT speaks limitedly with other LPs to stay independent, highlights time allocation as the core constraint, and shares that the hardest part of the job is disappointing many managers.
- •Shift toward more optimism on hard tech investing
- •LP behavior is overly pro-cyclical; narratives swing too fast
- •Staying independently minded can mean limiting LP-to-LP feedback loops
- •Best advice: humility in good/bad times; time is the scarcest resource
- •Hardest job element: saying no and disappointing long-term relationships
