Jan 29, 2024 · 57m · news
Ryan Akkina: How MIT Builds Their Venture Fund Portfolio & How MIT Approach Direct Investing | E1109 · 20VC with Harry Stebbings
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this interview, Ryan Akkina of MITIMCo discusses how MIT builds its premier venture capital portfolio, evaluates fund managers, and executes high-speed direct and co-investments. He shares vital lessons on GP-LP dynamics, direct deal structuring, and managing endowment liquidity across changing market cycles.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Harry holds 26.3% of the talking time here. How this is scored →
Harry as informed peer 4.8 Guest teaching 5.0 Guest disagreement 1.0 Harry pushing back 3.0
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 24:03 Ryan defends backing undisciplined 2021 managers
Ryan counters Harry's sharp criticism of LPs being lenient on 12-month deployment cycles by pointing out that blacklisting 2021 deployers would leave LPs with no established managers left to back.
Hardest push from Harry ▶ 6:44 Harry challenges MIT's brand power in securing accessHarry explicitly challenges Ryan's assertion that access is hard, arguing that MIT's elite reputation should allow them to sit back and secure allocations easily.
Biggest teaching moment ▶ 39:13 Ryan breaks down shale energy mechanics and decline risksRyan provides a masterclass on horizontal shale drilling, energy debt structures, and rapid decline rates, explaining why continuous balance sheet reinvestment creates severe structural downside risk.
Harry holds his own ▶ 33:43 Harry attacks endowment compensation and incentive flawsHarry forcefully calls out the structural flaw in university endowments where investment professionals achieve massive 7-8x returns on $120M direct deals without receiving proper carried interest.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Ryan's Journey to Fund Investing and MIT | 1 | 1 | 0 | 0 | Ryan outlines his career trajectory from electrical engineering at Stanford to management consulting and eventually endowment investing at MIT. Harry sets up the episode with a open conversational invitation. | |
| The Evolution of the Venture Landscape | 5 | 3 | 1 | 3 | Harry quotes Doug Leone on venture transitioning from a high-margin boutique model to a low-margin commoditized asset class. Ryan agrees that the influx of institutional capital makes this shift permanent. | |
| Securing Access and Emerging Managers | 6 | 5 | 2 | 6 | Harry explicitly pushes back on Ryan, arguing MIT's top-tier brand should allow them to sit back and secure top allocations. Ryan clarifies that late allocation sizes in premier funds are often too small to justify. | |
| The "See, Pick, Win, Service" Evaluation Framework | 6 | 4 | 2 | 6 | Harry challenges Ryan's See/Pick/Win/Service evaluation framework by pointing out top-performing GPs who have terrible founder NPS. Ryan acknowledges the nuance, explaining that pure picking works in seed but winning is critical at Series A. | |
| Evaluating Track Records and Human Factors | 5 | 5 | 1 | 2 | Harry brings up perspectives ranking track record as secondary to qualitative factors. Ryan details how likability and partner diversity drive founder selection, citing Sequoia's historic partner composition. | |
| Manager Mistakes and the Luck Factor | 3 | 5 | 1 | 2 | Ryan explains how early luck creates a virtuous feedback loop that builds long-term GP credibility. Harry asks how LPs evaluate success in years 1-3, and Ryan admits it takes 3-5 years to see genuine portfolio inflection. | |
| Why Venture Funds Go Sideways | 5 | 5 | 1 | 3 | Harry quotes Neil Mehta regarding why top funds derail, prompting Ryan to highlight arrogance, asset bloat, and lost drive. Harry presses on the LP challenge of scaling down commitments to mega-funds. | |
| Case Study: Backing Green Oaks and Neil Mehta | 3 | 5 | 0 | 0 | Ryan shares the backstory of backing Green Oaks Fund I, revealing MIT almost passed due to lack of established brand until glowing founder references changed their minds. Harry listens and facilitates the narrative. | |
| Handling Strategy Shifts in Venture Funds | 4 | 6 | 1 | 2 | Ryan details MIT's three-bucket LP portfolio architecture, ranging from $50M-$150M core checks down to $1M co-investment sourcing checks. Harry asks clarifying questions on position sizing and strategy drift. | |
| Venture Deployment Pace and Market Cycles | 7 | 5 | 3 | 7 | Harry strongly criticizes LPs for being too forgiving when GPs deployed funds in 12 months during 2021. Ryan counters that if LPs blacklisted undisciplined 2021 deployers, they would have no established managers left to back. | |
| Market Timing and the Public-Private Interplay | 6 | 5 | 1 | 2 | Harry cites 20VC data showing valuation entry prices down 43% from peak bubble levels. Ryan draws parallels to public market mandates, explaining MIT's steady 5% annual institutional payout mandate. | |
| The Venture Liquidity Bottleneck | 4 | 6 | 1 | 1 | Harry explores options for solving the pre-IPO liquidity bottleneck. Ryan explains that secondary LP buyers demand high return thresholds, making holding positions economically superior to selling at steep discounts. | |
| Downside Protection and Structured Notes in Directs | 6 | 6 | 2 | 6 | Harry criticizes LPs who attempt direct investments without proper rigor, calling it the 'messy middle'. Ryan responds by detailing MIT's deep on-the-ground due diligence for Coupang in South Korea. | |
| Direct Portfolio Allocation & Sizing | 7 | 6 | 1 | 6 | Harry directly challenges traditional endowment compensation models for lacking carry on huge direct hits like Coupang. Ryan candidly agrees, admitting institutional risk aversion limits performance-based incentives. | |
| Designing the Ideal Family Office Incentive Model | 5 | 6 | 1 | 3 | Ryan outlines how he would structure compensation at a family office to align skin in the game. He emphasizes that MIT's institutional speed allows them to win direct co-investments during high-pressure deals. | |
| Case Study: Quick Decision-Making during the Bank Collapse | 4 | 7 | 0 | 1 | Ryan details participating in Rippling's emergency round during the SVB collapse over a weekend, then shares a comprehensive post-mortem on a past oil and gas investment loss. | |
| Investing Style Evolution and Errors of Omission | 4 | 6 | 0 | 1 | Ryan reveals passing on OpenAI when Sam Altman left YC as his single biggest error of omission. He also outlines MIT's overall portfolio risk parameters, targeting a drawdown beta under 0.75. | |
| Fundraising Advice, LP Base Construction, and Investor Relations | 6 | 5 | 1 | 4 | Harry offers his 'lines not dots' rule for fundraising and challenges whether IR teams create unnecessary barriers between LPs and GPs. Ryan defends skilled IR teams for protecting GP investment focus. | |
| Evaluating Investment Exposure in China | 6 | 6 | 1 | 5 | Harry questions LP willingness to accept modest returns on bloated multi-billion dollar mega-funds. Ryan agrees that 3x returns are far less probable at scale and states MIT will scale out if excess returns disappear. | |
| Quick-Fire Questions: Lessons, Regrets, and Future Outlook | 3 | 4 | 0 | 1 | Harry conducts a quick-fire round covering hard tech, LP pro-cyclical tendencies, passing on Founders Fund, and Neil Mehta's key strengths. Ryan offers measured reflections on time management and independent thinking. |
Statements from this episode (56)
Stebbings: MITIMCo has over $30 billion in assets under management
“He is a member of the global investment team at the MIT Investment Management Company with an AUM of over thirty billion dollars.”
Akkina: Fund managers become arrogant after success, leading to worse decisions
“Frankly, if people have a spell of success, sometimes they become arrogant, right? They start to make worse decisions and treat people worse. When things are going well, you're never as smart as you think.”
Stebbings: Venture Capital Will Never Return to a High-Margin Boutique Model
“I don't think we ever go back to that boutique high margin business. I think we've forever private-equitized or hedge fundized. Our business has changed so seismically, and there's not a going back, just like hedge funds are the asset allocation industry that …”
Akkina: Venture capital is a secular growth industry despite cycles
“I think this is something that only goes one direction, and of course there will be cycles, like right now in things to You know, having a bad spell the past few years, it's become a little less popular and there's a little less competition, but I think that i…”
Akkina: Down markets weed out uncommitted emerging VC fund managers
“During times like this, when it's much harder to raise capital, you see which new firms really want it badly, and the ones who don't are disappearing, right? So I think that's one thing. And of course, look, at least for some period of time while we're in a do…”
MITIMCo backs Sequoia, Andreessen Horowitz, Green Oaks, Y Combinator, and Initialized Capital
“So we're lucky to be in a lot of the big names you would expect, you know, Sequoia, Kleiner Perkins, Andreessen, et cetera. And then, you know, there's some newer things as well that we've done in the past 10 or 15 years, like say, Green Oaks or Y Combinator i…”
Akkina: Venture fund evaluation requires assessing see, pick, win, and service
“It's see, pick, win, right? We need people who can see the best deals or at least a very large volume of deals, right? Hopefully the best deals are inside that set. Then we want them to be able to pick the best ones to try and invest in. And then finally they …”
Akkina: Unliked venture managers can succeed at seed, but rarely at Series A
“So I think in seed, that's probably more doable than in series A, for instance, right?”
MITIMCo evaluates GPs primarily on winning deals and servicing portfolio companies
“I would say we tend to spend the most time thinking about the winning and servicing angle, which kind of feed on each other. Because I think, you know, look, there's at least several hundred people in the industry at any given time, GPs I'm talking about, who …”
MITIMCo Biases Toward New Fund Managers With Angel Track Records
“Although, I would say when we back a new fund, we do probably have a bias towards people who at least have some kind of angel track record.”
Akkina: Sequoia's Dominance Came From Partners Appealing to Different Founder Profiles
“An interesting observation I heard about Sequoia once actually is that one of the things that made them so successful was there was this period where they had Mike Moritz and Jim Getz and Doug Leone all kind of in their prime all very impressive VCs in their o…”
Akkina: VC fund success cannot be judged in first three years
“Honestly, I would say in the first three years, you usually can't tell. Right. It would be some time between three and five years where we start to see whether they're, some of their companies really inflected. And we don't necessarily care whether something h…”
Akkina: VC funds go sideways by growing too big too quickly
“Well, one thing is I think sometimes firms grow too big too quickly, right? And that forces them out of whatever their sweet spot was.”
Akkina: MITIMCo's default strategy is loyalty to existing fund managers
“Our default is always to be loyal to the people who we're already invested in, right?”
Akkina: Every venture firm eventually declines due to botched generational transitions
“Every firm, no matter how great, has a half-life, right? No firm is going to be great forever. They'll eventually botch a generational transition or something, if they even try.”
Akkina: Green Oaks is one of MITIMCo's top fund relationships over 10 years
“And I remember actually, you know, my boss, Seth, and I decided, you know, let's, let's leave this and look at fund two or something. And the next, but I slept on it, and then the next day I decided to call my boss, Seth, back and say, you know, I think we sho…”
Akkina: MITIMCo conducts thousands of reference calls during manager diligence
“And you know, we do thousands of reference calls, right?”
Akkina: MITIMCo supports strategy shifts if communicated early and purposefully
“I mean, we want people to do what they say, but if they discover the original strategy is not likely to work, then obviously we want them to evolve, right? So I think it's about, You know, communicating that early if you are gonna shift your strategy and doing…”
Akkina: MITIMCo commits $50M–$150M across 6–8 core U.S. venture funds
“So today, at least in the U.S. Venture portfolio, we probably have Maybe six to eight core relationships we would refer to them as where we're writing checks of like 50 to a hundred fifty million per fund.”
Akkina: MITIMCo writes $10M–$20M checks for mid-tier venture funds
“And then after that, we have another bucket where the check sizes are typically 10 to twenty million per fund.”
Akkina: MITIMCo writes $1M checks to emerging managers without track records
“And then the, a third bucket we have now is things where basically we'll write a million dollar check because maybe it's extremely early and the person had no track record, not even an angel track record, let's say, or we're doing it to work with them to sourc…”
MITIMCo scaled annual private market commitments down from $3B to $1B
“You know, at peak we probably were putting three billion out the door in one year, and now we're probably down to about a third of that, so maybe a billion dollars.”
Akkina: LPs Avoid Criticizing Top VCs Due to Scarce Fund Capacity
“The tension there, right, is the top VCs have scarce capacity relative to, You know, their fund size relative to the universe of LPs who might like to invest there's a big mismatch there, right? And so I think LPs often feel reticent to criticize people too ha…”
Akkina: Virtually Every Venture Firm Was Undisciplined in 2021
“We all know that, that we were all undisciplined for some period of time, right? It's not a secret, and you know, every, every, virtually everyone did it.”
Stebbings: 20VC Entry Prices Are 43% Lower Than 18 Months Ago
“When we compare our entry price now versus 18 months ago, it's 43% lower. And, like, of the last nine companies we've done, that is held true.”
Akkina: VCs should not be criticized for deploying capital during bubbles
“In the public market, if a manager went all to cash in their portfolio, we wouldn't like that either. We would say they're trying to time the market, right? So similarly, I don't think you can criticize VCs for deploying money during a bubble period, right?”
Akkina: MIT Endowment Remains Liquidity-Constrained Until Major Unicorns Like Stripe Exit
“We're less liquidity constrained now than we were say at the beginning of last year, but the major issue still continues to be that there's all these big companies like Stripe, for instance from the last cycle that have not had liquidity events yet. And so unt…”
Akkina: MIT endowment must pay out roughly 5% of assets annually
“Normally we know we have to pay out about five percent Of our assets per year to support our institution. That never really changes.”
Akkina: MITIMCo almost never sells fund stakes on secondary markets
“We'll certainly think about it. We almost never do it because we're not willing to do it at a low price, right? I mean, the reason these secondary funds exist is obviously to get high IRRs themselves, right? And so, generally speaking you know, if we're sellin…”
Akkina: Direct co-investments offer higher conviction than blind-pool fund commitments
“I personally find it easier in a lot of cases to write a big check to a co-invest than to a blind pool fund, right? Because no matter your conviction in someone, when you're writing them a blank check, you know, we don't know what's gonna happen, right? And as…”
Akkina: MITIMCo structured most of its Coupang investment as senior notes
“Most of the money we invested in that was in this special structured note where we felt it had very protected downside, but managed to keep a lot of the upside as well.”
MITIMCo invested $120M in Coupang, returning 7-8x at distribution
“So in, you know, in the case of Coupang, that was a hundred twenty million co-investment for us. We made, I think, a distribution price is seven to eight X or something like that.”
MITIMCo was constrained to a small Snowflake pre-IPO allocation
“I mean, there's another opportunity where we co-invested in Snowflake before it IPO'd, and in that case, we were only able to get a few million bucks allocation, even though, given what we knew about the company, we would have invested a lot more if we had the…”
Akkina: Traditional Endowment Incentives Fail to Reward Risk-Taking
“The traditional endowment or foundation, I think part of the reason they're not good at this stuff is they don't have an incentive to be right, and people, when they don't have an incentive to take risks, they're not going to want to stick their necks out.”
Akkina: Family offices should require investment staff to co-invest personal capital
“I, you know, if I think, if I had a blank sheet of paper, if I were running a family office, let's say, I would probably do something where, one, people would be required to invest a lot of their own money in the investments the firm was making. You know, eith…”
MITIMCo Declined Co-Investments From Respected GPs at Peak Tech Bubble
“I would give ourselves some credit for, at the peak of the bubble, we said no to several co-investments where, ah, you know, people that we really respect and still respect were pounding the table to do these things, and we said no.”
Few University Endowments Can Execute Large Direct Investments Quickly
“We're probably one of the few LPs, certainly one of the few university endowments who can react to one of these things very quickly in, in big size.”
Neil Mehta Completed Rippling Funding Deal Over Weekend During SVB Collapse
“With Neil, for instance, I think there was a big article maybe last year about the deal he did in Rippling, right? And that came together over a weekend, right? During this bank collapse.”
MITIMCo approved a Rippling co-investment within hours on a Friday night
“Within a few hours, Friday night we made the decision to back that deal.”
Akkina: Peak valuation of failed oil investment exceeded $1 billion at 5-7x
“I mean, this thing was worth, I think over, you know, well over a billion dollars at peak. And at peak it was probably like a five to seven X for us.”
Akkina: Avoid businesses needing annual balance sheet reinvestment or having high churn
“You have to be very careful with businesses where you have to reinvest the balance sheet every year. You have to continuously reinvest the customers, right? So other parallels I sometimes notice, things like gaming, where maybe you have a very high churn rate …”
Akkina: MITIMCo recovered roughly one-third of capital on worst direct deal
“I think on that particular deal, we still recovered about a third of our capital or something like that, so it wasn't a zero”
Akkina: Generating consistent public market alpha is harder than venture
“I just find it much easier for me and my personality to I think generate alpha on the private side. I mean, the public side is always becoming more competitive. You think, if you think venture is competitive, try making alpha versus you know, the NASDAQ, right…”
Akkina: MITIMCo Passed on an Early Opportunity to Invest in OpenAI
“I mentioned we're investors in YC, and when Sam left YC, around then we also had an opportunity to invest in OpenAI, and we didn't do it. So that's definitely an error of omission that I think about a lot.”
Akkina: VC fund managers should treat fundraising like enterprise sales
“I think you have to treat the process like an enterprise sales cycle, right? Which means you need to think of it as a numbers game, and you need to have a lot of funnels.”
Akkina: VC managers should not force LP evaluations too early
“It's good to build relationships, but you don't necessarily want to force people to evaluate you too early. Because I think particularly with larger institutions, it can be the case that, you know, once they evaluate you once, they may not be willing to look a…”
Akkina: VC funds should limit single LP concentration to 10-20%
“Well, yeah, I think you probably don't want to exceed 10 to 20%, except in very unusual circumstances.”
Stebbings: In-house VC IR teams weaken GP-LP relationships and direct deals
“I find IR teams respectfully, and there's nothing against them, but they just create a layer that removes from the human relationship with the GP. I think it also impedes your ability to get good directs, because you lose that messaging with Neil, or messaging…”
Akkina: MITIMCo deliberately minimizes time demands on portfolio fund managers
“And of course, something we try to do as LPs is not be overly demanding in, The time we take from people, right? You know, we're always very conscientious of the fact that we want most of these people to be spending most of their time out on the field, right? …”
MITIMCo is significantly reducing its investment exposure in China
“We're definitely, I think like many others, we're doing a lot less there now. I mean, we're not gonna categorically not do it at all yet. But we are reducing it, and to the things that we keep doing, we're, you know, we're very careful about whether they're pe…”
Akkina: Delaying economic sharing causes venture firms to lose top talent
“They're thoughtful about how they share economics and, you know, they don't wait too long to do that. And I think one of the reasons a lot of firms sometimes lose some of their best people is because they don't do that quick enough.”
Akkina: MITIMCo will scale out of VC funds that become too large
“If a firm gets too big and we're no longer confident that we can make exceptional returns over time, we're going to have to scale out.”
Akkina: LP interest in late stage, solo VC, and crypto will return
“Two years ago everyone wanted to do lots of late stage and solo capitalists and crypto, and then, you know, a year later they all think it's dead and don't want to do that ever again. I think most of those things will come back.”