All-In PodcastE91: SoftBank's $21B+ Vision Fund loss, signals of a bubble, macro picture, Trump raided by FBI
CHAPTERS
- 0:00 – 2:58
Besties banter: vacations, wakeboarding, and Chamath’s unbuttoned market indicator
The episode opens with light roasting about outfits, travel, and how bullishness supposedly correlates with Chamath undoing buttons. The hosts reestablish the “besties” dynamic before pivoting to the week’s main topics.
- •J-Cal shares he was wakeboarding in Tahoe; Chamath recounts a Sardinia trip
- •Running jokes about Chamath’s shirt buttons and shorts
- •Quick check-ins with Sacks and Friedberg on how they’re doing
- •Tone-setting: playful sparring before serious market talk
- 2:58 – 6:53
SoftBank’s $21B+ Vision Fund loss: Masa’s public mea culpa and what changed
J-Cal introduces SoftBank/Vision Fund’s massive quarterly loss and plays a clip of Masayoshi Son framing the setback through a historical Japanese analogy. The group reacts to Masa admitting “delirium” during peak profits and discusses the dramatic reversal from gains to losses.
- •Vision Fund losses and SoftBank’s rapid shift toward cost-cutting
- •Masa’s presentation style and admission of embarrassment/remorse
- •J-Cal argues SoftBank helped inflate the bubble (e.g., WeWork)
- •Capital deployment collapses: from tens of billions to hundreds of millions
- 6:53 – 10:34
Respect vs. criticism: risk tolerance, LP protections, and sovereign wealth structuring
Chamath emphasizes the courage required to deploy huge capital and notes Masa’s history of extreme wins and drawdowns (e.g., Alibaba and dot-com). He also highlights how Saudi/Abu Dhabi structured their Vision Fund exposure with preferred-like protections, potentially earning money even if the fund underperforms.
- •Chamath frames outsized losses as the flip side of outsized ambition
- •Masa’s long arc: Alibaba win and dot-com-era drawdown
- •Sovereign LPs’ preferred equity/coupon structure reduces downside
- •SoftBank liquidity management via Alibaba hedges/derivatives
- 10:34 – 12:33
Why mega-VC isn’t scalable: craft business, bubble psychology, and ‘sawtooth’ markets
Sacks argues bubbles are psychologically powerful and end abruptly, making it hard to time exits. He disputes that SoftBank alone caused the bubble, pointing to ‘tourist’ crossover capital and massive post-2020 liquidity as key drivers, then asserts VC is fundamentally hard to scale without degrading returns.
- •Bull markets rise slowly but crash quickly (‘elevator down’)
- •Bubble creation attributed to broad liquidity + crossover funds (e.g., Tiger)
- •VC described as a craft business, difficult to scale operationally
- •Noticing how macro liquidity conditions seep into private markets
- 12:33 – 16:19
The ‘oversized check’ failure mode: when capital becomes the product
Friedberg and Sacks explain how SoftBank’s tendency to offer far more capital than requested warped companies’ unit economics and decision-making. They use DTC customer acquisition and WeWork as examples where too much money forces unnatural growth, rising acquisition costs, and strategic distraction.
- •Founder pitches get rewarded with larger-than-needed checks
- •Too much capital flips unit economics (CAC rises, efficiency falls)
- •WeWork case: strategy inverted after massive funding
- •Overfunding distracts founders into side projects and empire-building
- 16:19 – 19:22
Kingmaker fallacy, seed-sized businesses with giant rounds, and milestone-based checks
The hosts critique SoftBank’s belief it could ‘anoint’ winners in competitive markets, forcing others to react. They cite examples of very early-stage companies receiving huge checks and contrast that with a milestone/proof-based approach to sizing investments.
- •SoftBank’s ‘kingmaker’ strategy and competitive market dynamics
- •Contrast with Tiger’s more passive, lower-ownership approach
- •Examples of pre–product-market-fit or ‘seed-like’ companies funded at scale
- •Chamath: check size should map to proof/milestones
- 19:22 – 22:53
Structural mistake thesis: a $100B fund with a 5-year investment period
Chamath argues the core error wasn’t just picking winners/losers, but the fund structure: a 10-year fund with ~5 years to deploy. That math forces frantic pacing, diluted diligence, and an oversized organization—whereas longer-duration vehicles (15–20 years) could deploy more patiently and become the buyer in downturns.
- •Five-year investment window implies ~$20B/year deployment pressure
- •Implied pace makes diligence and decision quality collapse
- •Longer-life PE-style funds (15–20 years) enable patience and timing
- •Organizational sprawl and ‘team sport’ investing reduces alpha to beta
- 22:53 – 27:59
‘Ride your winners’ debate: paring positions, evergreen funds, and liquidity promises
J-Cal argues SoftBank could have reduced risk by selling small portions of winners along the way, especially when valuations became extreme. Chamath counters that strategies like permanent holding/evergreen structures can create leverage and liquidity mismatches—great in up markets, fragile in down markets—leading to a debate over secondaries, distributions, and debt.
- •J-Cal: pare 10–20% in secondary/public markets to lock in gains
- •Chamath: evergreen + liquidity mechanisms can force leverage/borrowing
- •Sacks: private-company sales are often constrained; public positions allow distributions
- •Tension between long-term ownership and managing LP liquidity expectations
- 27:59 – 36:36
Spotting bubbles: vanity valuations vs. fundamentals, SaaS multiples, and Fed policy
The conversation shifts to indicators of market tops. Friedberg flags ‘valuation trophy hunting’ as a tell; Sacks cites median SaaS price-to-ARR multiples deviating from historical norms and emphasizes how interest-rate policy (ZIRP/QE) quietly drives tech valuations.
- •Friedberg: bubble sign = focus on unicorn/decacorn status over margins/ROIC
- •Sacks: track median SaaS revenue multiples as public-to-private valuation anchor
- •Macro matters: inflation, rates, and liquidity conditions drive risk assets
- •Historical note: rate hikes popped the dot-com bubble despite moderate starting rates
- 36:36 – 40:17
Macro snapshot: CPI, jobs, recession semantics, and the ‘double-dip’ risk
They react to inflation cooling from 9.1% to 8.5% and strong employment prints, while criticizing political spin around definitions (inflation framing, recession criteria). Sacks outlines a base case of a shallow technical recession with potential for a later downturn as rate hikes propagate through housing and construction.
- •CPI down modestly; debate over YoY vs. MoM framing
- •Jobs strong, but broader slowdown signals persist
- •Housing/construction hit first by higher rates; lag effects 6–9 months
- •Sacks: possible near-term recovery followed by double-dip next year
- 40:17 – 48:42
Consumer debt as pressure point: variable rates, defaults, and tipping points
Friedberg focuses on rising household debt and consumer credit growth in a rising-rate environment. He explains the mechanism by which higher debt service could outpace income/asset gains, triggering defaults that ripple through credit markets and equities.
- •Household debt ~ $16T; credit card balances rising again
- •Variable-rate debt means monthly servicing costs climb with rate hikes
- •Risk framework: when debt service outpaces income/asset values, defaults rise
- •Potential downstream effects: bond stress, equity declines, broader economic crunch
- 48:42 – 50:45
Panel friction: ‘we don’t know’ fatigue and deciding what’s worth discussing weekly
The macro discussion devolves into frustration about repetitiveness and uncertainty. Friedberg and Sacks argue the inflation/recession talk has become circular without new information, while J-Cal defends discussing major prints; they agree to move on to more concrete news.
- •Complaints that weekly macro talk is repetitive without actionable insight
- •Core macro framing: inflation path vs. economic damage from rate hikes
- •Acknowledgement of uncertainty and need for better topic selection
- •Transition decision: pivot to Trump/FBI raid story
- 50:45 – 1:07:56
FBI raid on Mar-a-Lago: nuclear documents claims, trust in institutions, and political blowback
The group debates the unprecedented nature and optics of the Mar-a-Lago raid, including allegations about classified/nuclear-related materials. Sacks argues the FBI leadership’s past conduct justifies skepticism and warns the raid may politically strengthen Trump; Friedberg worries more about institutional trust collapsing and polarization accelerating, with Chamath emphasizing that no outcome will satisfy both sides.
- •Scenario debate: if sensitive nuclear materials were withheld, what’s appropriate enforcement?
- •Sacks: FBI/DOJ history (FISA issues, leadership bias) reduces benefit of the doubt
- •Optics and process critiques: militarization, lawyer access, cameras, scope of search
- •Broader concern: ‘inflammatory index’ rises, institutional integrity erodes, Trump re-centered
- 1:07:56 – 1:09:00
Quick wrap and outro: tension release and signature catchphrases
They close the episode abruptly after the political discussion, then shift back into jokes and recurring show phrases. The ending mixes playful banter with callbacks to “let your winners ride” and other in-jokes.
- •Abrupt sign-off after heavy institutional/polarization discussion
- •Recurring outro lines and besties catchphrases
- •Light comedic riffing to end the episode
