Podcast Summary: How I Invest with David Weisburd

Episode: E286 – How LPs Can Actually Find Alpha in Venture
Date: January 20, 2026
Guest: Abe (AngelList Consulting Researcher, Strawberry Tree Management)
Host: David Weisburd

Episode Overview

This episode reunites David Weisburd and Abe, discussing how institutional investors, especially LPs, can find genuine alpha in venture—despite the crowded field and widely held beliefs about efficiency and randomness in the industry. Abe draws on six-plus years of deep data analysis at AngelList, unpacking power law returns, market inefficiency myths, the paradoxes of seed-stage investing, and where quant strategies fail but still offer opportunity—especially in fund of funds construction for LPs.

Key Discussion Points & Insights

1. Revisiting VC Culture and Rationality

• Initial VC Skepticism to Unexpected Respect:
  • Abe describes beginning his career skeptical of VC decision-making, hoping data would “rationalize” odd behaviors. Instead, the breadth and objectivity of AngelList data have “radicalized” him toward respect for what VCs do and the structures that drive alpha.
  • “Just things that would be shocking to me to hear…is that, you know, actually venture capitalists are doing a pretty good job from the data.” [01:03, Abe]

2. The Unique Data Edge of AngelList

• No Survivorship Bias & Price per Share Transparency:
  • Abe emphasizes the value of having both the “dead” companies and precise share price trajectories—not just headline rounds.
  • This reveals that big winners (e.g., Anthropic’s $183B round) achieve greater outcomes not only by outperformance but by being less diluted than expected.
  • “The real significance…is that you can, you don’t need to make assumptions about…dilution…You get very, very different perspectives because so much the returns are driven by your highest performers.” [04:30, Abe]

3. Venture Capital is Not One Asset Class

• Three Distinct Regimes (Power Law Explained):
  • Seed: “Alpha less than two power law”—returns are so dominated by a handful of companies that broad exposure is optimal.
  • Series A: A “middle ground”—still fat-tailed, but requires more selectivity.
  • Series B and beyond: “Alpha greater than 3”—resembles public market return distributions.
  • “Venture capital is not an asset class. It is actually three different asset classes…Seed investing is its own thing…” [04:30–08:02, Abe]

4. Practical Implications for Seed Investing

• You Should Be an Indexer (‘Within the Credible Set’):
  • A paradox: You can rank investments, but the market is highly efficient (price almost always reflects perceived quality).
  • The best deals (repeat founders, big exits, high prices) may only be marginally better on an absolute basis compared to much cheaper ones—because returns drop off faster than price.
  • “There’s not just dollars lying on the ground for people to build quant models to like pick up...[research] is already priced in.” [13:51, Abe]
  • “Anything that’s above the threshold for investment quality you should have some exposure into. It’s really…contradictory to have those two things right.” [09:53, Abe]

5. The Wildness of Power Law Returns

• No Expected Value—Infinite Variance:

  • You can’t think in terms of “expected return”; seed returns are so fat-tailed, outcomes are fundamentally unpredictable (“escape” behavior in power law distributions).
  • “The correct answer is you’ve already made an error asking the question because you used the concept of expected value...you can’t even refer to the concept...” [14:32, Abe]

• You Need Broad Exposure:

  • Even for the best deal (e.g. Samsara), its chance of being one of the “hyper-winners” is low, necessitating broad indexing.
  • “The number of opportunities that meets this threshold is very large…so much of the returns…driven by the top handful of those thousand…” [17:55, Abe]

6. The Role of Signal, Pricing, and Active Selection

• Why Quant and Signal-Based Approaches Fail for Startups:

  • The market is “80–90% efficient” — most signals are priced in (e.g., MIT founders, repeat exits).
  • “Not like ‘oh, I found this signal that nobody else knows about’...if it’s in a deck, it’s 80–90% priced in.” [13:51, Abe]

• Startups as Quantum Observations (Ontic, not Epistemic, Uncertainty):

  • The very act of investing brings possible companies into existence—there is no perfect information.
  • “Literally what I think seed investors do is actually…by funding ideas with capital…actually create these particles…by deciding to go on this journey, they’re actually creating the universe.” [23:00–25:00, Abe]
  • Memorable: “The cheapest and most effective way to diligence a pre seed investment is to write a check.” [24:30, Abe quoting Kevin Lawless]

7. Adverse Selection and GP 'Style Drift'

• Investment Size as a Signal (Small Checks Outperform):

  • Found consistent data that small GP checks relative to their normal size are the best performers, because the VCs are being ‘squeezed’—they’re fighting to participate in the hottest rounds.
  • Conversely, big checks show little added performance—if a top VC isn’t taking that allocation, why is it available?
  • “Small checks are the best investments…because they're being squeezed down by better investors.” [38:00, Abe]

• Adverse Selection Domination:

  • The availability of allocation itself is a negative signal, unless you are a top domain expert or fund.
  • “Anyone who’s not [top investor], the fact that capacity exists for you is a negative signal about the company.” [41:10, Abe]
  • On 'Style Drift': If a GP moves outside their expertise, they see worse deals and are at risk of negative alpha.

8. Implications for LPs: Where to Find Alpha

• Direct Indexing vs. Funds of Funds:

  • If you could index all credible seed investments, you’d likely get top quartile returns (as the mean is historically around 75th percentile).
  • But nobody sees all deals; fund managers are essential due to specialization. LPs should “index” GPS with differentiated sourcing.

• Fees—Not as Significant as Access:

  • Management fees matter only when funds go very high-multiple; otherwise, fund selection is overwhelmingly more important.

9. Strawberry Tree’s Quantitative Fund of Funds Approach

• Why Quant Works for Funds (Not Startups):

  • “If you have, you know, positive quantitative signals, you can kind of directly convert those to alpha…It’s not priced out the same way [as startups].” [46:15–49:15, Abe]
  • Uses statistical models (e.g., Markov S over baseline: how many past investments are marked up vs. expected) to select GPs showing above-trend interim progress, not just DPI (cash back to LPs).

• Fund of Funds Target:

  • Strategy is to index the top 60-70 GPs based on noisy but positive correlation signals; seek 3.6x–6x net returns.
  • “Every fund we invest in has an above average chance of being an above average venture fund. That’s the level at which we’re willing to say.” [50:11, Abe]

• Negative Alpha as a Concept:

  • An early huge win by a GP can be a contra signal: subsequent larger, more generalist raises lead to style drift, loss of edge.

Notable Quotes & Memorable Moments

• On VC Rationality:

  • “Actually venture capitalists are doing a pretty good job from the data, actually a considerable amount of respect for them and for the work that they do.” [01:03, Abe]

• On ‘Indexing’ Startups:

  • “I think what’s interesting about seed is that you can do the research and figure out what the good companies are and you should still invest in virtually…everything that is [a] credible seed deal.” [09:53, Abe]

• On Power Laws:

  • “The correct answer is you’ve already made an error asking the question because you use the concept of expected value. And if you actually believe what I’m saying, you can’t talk about expected value. There is no expectation. It doesn’t exist.” [14:32, Abe]

• On Quantum Uncertainty in Investing:

  • “[Seed investing is] ontic, not epistemic…by funding ideas with capital to actually go and create these particles…they’re actually creating the universe. Which I think is like absolutely mind blowing and fascinating.” [23:00–26:00, Abe]

• On Signals and Pricing:

  • “If it’s in a deck, it’s 80, 90% priced in. You’re not uncovering the hidden mysteries…” [13:51, Abe]

• On Small vs. Large Checks:

  • “Small checks are the best investments…because they're being squeezed down by better investors.” [38:00, Abe]

• On Adverse Selection:

  • “Anyone who’s not [top investor], the fact that capacity exists for you is a negative signal about the company.” [41:10, Abe]

• On Evaluating GPs:

  • “Every fund we invest in has an above average chance of being an above average venture capital fund. That’s the level at which we’re willing to say.” [50:11, Abe]

• On DPI as a Bad Signal:

  • “DPI is not even relevant…as a future facing signal because it takes so long to assemble enormous DPI...are they still plugged in?” [57:27, Abe]

Important Timestamps

• [01:03–04:30]: Abe’s journey from VC skeptic to empath/respecter based on AngelList data
• [04:30–08:02]: Three-asset-class framework for VC: Seed, A, B+; power law explained
• [09:53–14:08]: Seed return paradox—practical advice on diversification vs. concentrated investing
• [14:32–19:12]: Why you can’t use expected value; statistical “wildness”
• [23:00–26:00]: Quantum mechanics and “creating” startups through investment
• [34:52–41:10]: Adverse selection, check size signaling, capacity as negative signal
• [45:55–53:48]: Why quant doesn’t work for startup picking, but can for fund selection
• [53:48–57:03]: Fund of funds strategy: process over outcome, key signals (mark-ups over baseline)
• [57:03–60:36]: Critique of DPI as signal; negative alpha and dangers of early big wins

Actionable Insights for LPs

• Don’t Over-index on Signal at the Seed Stage:

  • Everybody knows the “obvious” signals—they are already priced in.
  • Broad exposure to credible seed deals beats heroic concentration.

• Identify and Avoid Adverse Selection:

  • If you are getting unexpected access to a hot deal in a domain outside your expertise, consider that a negative signal.

• Fund of Funds May Offer Real (if Modest) Alpha:

  • Quantitative signals applied to GPS (not startups) can be meaningful, since GP pricing is much less differentiated and access is fundamental.
  • Focusing on interim signals (markups, graduation rates) is likely superior to backward-looking DPI.

• Be Wary of Style Drift in GPs:

  • Alpha is domain-specific; success attracting more money often leads to diminished returns if GPs expand outside their main circle of competency.

Closing

Abe remains open to collaboration and dialogue:

• Read more: AngelList Blog
• Contact: abe@angelist.com

“I think there is just some really fundamentally, intellectually interesting, weird, fascinating stuff about this asset class that I maybe didn’t expect when I started.” [60:57, Abe]

The episode leaves LPs with a nuanced, data-driven framework to navigate VC—one that values breadth, humility, and process over conviction in picking outliers.