E54
E54: Spread trading big tech, capital allocation, Zillow's misfire, Progressives suffer setbacks
2021-11-06 spoken.md · speaker-labeled ▶ watch ← E53 all episodes E55 →
4
ideas born
11
ideas moved
14
captures · 4 voices
2
dissenting
+182.3
conviction added
-97.1
decay · 90 silent
Every number here is replayed from score_events — the same ledger the pool ranks on. Decay is what the 90 ideas nobody mentioned gave up this week; it applies only when an episode is processed.
Tier crossings
conviction thresholds crossed by this episode — 65 / 45 / 15 · ideas born here show where they landed. why 65 isn't always green
Kill dates that landed since E53
0 hit · 1 partial · 0 miss — windows that closed after 2021-10-30 and up to 2021-11-06, auto-scored against price data and never hand-set. verdict · R · α
| idea | verdict | R | α | closed |
|---|---|---|---|---|
| 📈 Prop 22 IC+ model becomes the national gig-work architecture | PARTIAL | +10.4% | -27.4 | 2021-11-04 |
Who moved the board
each voice's force on conviction this episode — supports and opposes, weighted exactly as the replay applied them · share = % of this episode's movement
What got argued (11 ideas)
ordered by how hard each idea moved · quotes are verbatim from the transcript, timestamps deep-link into the episode
Zillow Offers did not fail on execution alone: the Zestimate was built as lead-gen where an inflated price is the feature and accuracy is a bug, so underwriting home purchases off it guaranteed overpaying on a 3% margin of safety, and the new business was fundamentally competitive with the media/broker franchise it was supposed to feed. Friedberg adds that Zillow called itself a market maker while actually speculating on model-projected prices with nobody doing ground truth. The besties treat this as a lasting strategic wound at a company a Silicon Valley legend runs, not a fixable quarter.
plays
ZG ·primary
evals 2022-11-06
they were trying to be a, quote, market maker in the business. And it turned out that what they were really doing was being more of a speculator, right? ... one could argue this was flawed from the beginning. The execution clearly was way off. ... the maturity of a Silicon Valley entrepreneur who's crushed it so many times doesn't mean that they're perfect. And this was a big miss.
it speaks to two big mistakes that Zillow made. The first is that catching a wave of disruption is very difficult when you have an old line business that is fundamentally competitive with the new line of business. ... Zestimate over time basically calcified this idea of price inflation. So the accuracy was never a goal. ... the open door margin on average is about 10%. The Zillow margin on average was 3%. So they had a much more razor thin margin of safety here, which again speaks to pricing.
Forecasting house price and volatility and then selling the asset at a defined margin of safety is a software problem, Opendoor solved it first (roughly 10% average margin against Zillow's 3%), and like Google against Yahoo those technical gains compound every year into an unassailable head start. Chamath discloses 3-4% personal ownership and Sacks discloses a seed position; both argue Zillow's failure is proof the moat is real rather than proof the category does not work.
plays
OPEN ·primary
evals 2022-11-06
Look, I am one of the largest individual shareholders of that company. I think somewhere between 3% and 4% of the business I own. ... I suspect we will look back and the ability to accurately forecast price and volatility and the ability to sell these assets at a defined margin of safety in a predictable way is something software can solve, and it seems that Open Door is the first. It doesn't necessarily mean it will be the only one, but it has an enormous head start now, and as long as they can keep iterating, those gains will compound. ... I just think this is a wonderful business run by an incredible CEO.
I was a seed investor in Open Door because Ravoy invited me to the seed round, so I'm a little biased here, but I think it's pretty obvious what happened. ... Zillow tried to copy them. The business is much more operationally complicated than they realized. ... That's what frequently happens in a big company, tries to copy a smaller sort of upstart company.
Jason's proposed pair trade: Airbnb is a high-margin, well-run, still-growing travel platform while the airlines are low-margin and badly run, so long ABNB against short airlines should pay out over the coming year. Chamath rejects it on the merits - the two legs sit in different capital pools with different owners and are not correlated enough to net out market risk, so it is 'getting cute' rather than hedging. Only the long leg is listed as a play so the scorer reads the direction Jason argued.
plays
ABNB ·primary
evals 2022-11-06
I just came up with one. What about Airbnb versus the airlines? ... Well, tell me why Airbnb, which is an incredible margin business that's incredibly well run and growing, versus airlines, which are horribly run, and low margin, why wouldn't that be a good spread trade?
You're talking about smaller, rinky dink ideas. ... No, that's a stupid idea, and I'll tell you why. If you're going to put these things on, go to where the deepest liquid markets are, because those are the safest. ... They're completely different businesses with completely different motivations, with different capital pools, with different people that own the stock. There's no point trying to get cute on these things.
▲ PAVE
🏛️ Infrastructure trillions get captured by the contractors
closed 16
CONTESTED
▲ +15.3
29.5 → 44.8
I think what's going to happen in the wake of this election is that this infrastructure bill is going to sail through. Because one of the crazier things that the progressives were doing was holding that bill hostage. ... Because a lot of those programs are going to be popular in a state like Virginia.
Chamath's second worked example of a deep-liquidity spread: inside the trillions of dollars of auto market cap, be long the EV natives (Tesla, Lucid, Rivian) and short the traditional automakers. Offered as an idea that 'makes sense' and 'could be a trade' rather than a position he discloses, so conviction is low - but the direction is explicit. Rivian is excluded from the plays because it did not list until 2021-11-10, four days after this episode.
plays
TSLA ·primary
LCID
evals 2022-11-06
▲ AMRS
🤖 Biomanufacturing capacity is the bottleneck the US should own
unresolvable 17
▲ +13.1
19.2 → 32.2
27 million tanks made out of plastic. You could probably get that stuff produced, you know, a couple billion dollars. Find a piece of land that's 25 by 25 miles, that's near some water and put a nuclear power plant there. And you could suck up all the CO2 in the atmosphere. ... If we want to invest in infrastructure, this is the sort of thing that both solves climate change, creates jobs, and has extraordinary economic return potential built into it.
The question is, what do you short against it, so that you can take up the market volatility and you can play spread? ... That's a spread trade that over the next four or five years, where if you expect a lot of market volatility, it makes sense to maybe put some of this kind of stuff on.
▲ GOOGL
📈 Asset-light mega-cap software is the inflation hedge
closed 0
CONTESTED
▲ +10.7
29.3 → 40.0
I'm still long Microsoft and Google and short the rest of Big Tech. So I'm fine with it. ... Meaning you can very comfortably short Apple, Facebook, Amazon, Netflix and be long Microsoft, Google. So as a spread trade, it's the best risk parity trade on the internet right now. ... when I tweeted that tweet out, to complete the picture, I was actually long something against my proposed short. And I put it on in pretty meaningful leverage. And it's worked out really well because I was playing the spread. ... Another obvious one, so obvious is within big tech, figure out which ones you want to be long, which ones you want to be short. That's a spread trade that over the next four or five years, where if you expect a lot of market volatility, it makes sense to maybe put some of this kind of stuff on.
▲ KRBN
🛢️ Carbon permits and carbon tariffs put a real price on emissions
closed 13
CONTESTED
▲ +9.9
31.0 → 41.0
we're now seriously considering carbon tariffs. ... I think this is the most disruptive thing in the capital markets and geopolitics that can happen in the next 10 or 20 years is an effective carbon tariff, which is to say that when a good or service enters the borders of a country, they will levy some tax that they think represents its drag on the environment. ... Yeah, that's happening. I think that's coming. So I think that, you know, the combination of tariffs and these transfer payments is going to create a real economic incentive for folks to make these kinds of big technological leaps. So I'm pretty bullish on all of that.
▲ TEAM
🤖 Bottom-up SaaS is the dominant mode of business software
closed 0
CONTESTED
▼ -8.5
8.5 → 0.0
I felt this when I was on the board of Slack. When Microsoft put its gun sights on us, we always thought that they could not outcompete with us. And what it turns out is, when you have a massive distribution advantage, feature parity is enough, and you can actually be slightly less than good enough on the features, because distribution and bundling and packaging overpower a customer's desire to adopt a product. ... the only real long-term protective solution for Slack shareholders was to basically get bought by Salesforce so that you could be part of a bigger whole. ... This past week, Microsoft decided to go after Notion, and it's going to be, I think, a very similar story, where, you know, once they decide to sort of go after this product experience, they only need to be 80% as good, and then the distribution and bundling and packaging will take care of the other 20%.
▲ QQQ
🤖 Own the top 30% of global tech and hold it five to ten years
closed 6
CONTESTED
▲ +1.2
91.0 → 92.2
You have to remember, when Sequoia distributed Google, every single partner got Google shares. Now, had they held those shares, they wouldn't be saying any of this. ... he told me he had never sold a single share of Amazon that was distributed to him, and he had only sold a handful of shares of Google only to fund future capital requirements at Kleiner at the time. ... I think probably what Sequoia was saying is, man, I wish I had not sold my Google shares when they were distributed to me
Episode digest
written during extraction and stored in data/extractions/ep054.json — the auditable source of truth, including everything market-adjacent that did not earn a capture
LABEL CHECK: CLEAN - the rare 2021-era episode where the diarization holds up. Roster count Jason 112 / Chamath 85 / Sacks 54 / Friedberg 43, all four hosts present and self-identifying: label-Jason does his own third-person intro ('I am a pasty, white, old, Greek man'), says he had 'Glenn from Redfin on the pod' and delivers the 'On behalf of the dictator Chamath Palihapitiya' sign-off; label-Chamath names himself in the intro bit ('My name is Chamath... six foot two'), cites Social Capital, the Slack board seat, IPOB and 3-4% personal ownership of Opendoor, and the newborn on his chest; label-Sacks is broadcasting from the NYSE floor for Bird's listing, 'the first check I wrote as a VC to lead a round at Kraft'; label-Friedberg talks Production Board, tweeting the CO2-to-starch paper and does the science corner. Every addressed-by-name test passes (Jason 'Yeah, Friedberg, can we hear yours?' -> Friedberg 5:58; Friedberg 'Chamath, do you think that the Google long Netflix short play' -> Chamath 7:58; Jason 'Sacks, you've got the background of the New York Stock Exchange' -> Sacks 17:37; Jason 'Chamath, what are your thoughts on this' -> Chamath 22:42; Jason 'To wrap up this Zillow story, Sacks' -> Sacks 44:36; Chamath 'Why, Friedberg, why did they open source it?' -> Friedberg 1:18:35). No merged turns, no swap, no rotation. MARKET CONTENT: the headline 'spread trading big tech' is not a new thesis - it is Chamath doubling down one week later on software-quality-inflation-hedge-2021 (born E53), and this time it is his own money: 'I'm still long Microsoft and Google and short the rest of Big Tech', the borrowed hedge-fund trade is now named as long GOOGL+MSFT against short AAPL/FB/AMZN/NFLX, he says he put it on 'in pretty meaningful leverage' and it has 'worked out really well', and the reasons are the same inflation/supply-chain asymmetry as E53 (Apple has 'severe headwinds with respect to inflation pressures... and margins and supply chain issues', Amazon pricing power, Facebook regulatory). New framing worth keeping: he tells you the horizon out loud - 'a spread trade that over the next four or five years' - and the whole reason he prefers spreads to naked longs is that he expects a lot of market volatility, which is a light reinforcement of his own E22 volatility call. Two spin-off spread ideas got coined: Jason's long-Airbnb / short-airlines pair, which Chamath called 'smaller, rinky dink' and 'a stupid idea' on correlation and liquidity grounds (a rare clean scoreable disagreement), and Chamath's own long-Tesla/Lucid/Rivian / short-traditional-autos example, offered at 'that could be a trade' conviction. Chamath's MSFT bull case doubles as an argument AGAINST Sacks' bottom-up-SaaS thesis: from the Slack board he learned that 'when you have a massive distribution advantage, feature parity is enough', Slack's only exit was Salesforce, and Notion is next - captured as an oppose on bottom-up-saas-dominant-2021. ZILLOW: the misfire got a full post-mortem four days after Zillow Offers was shut down (stock -37% on the week, -70% from its February peak, 7,000 homes for sale at ~$2.8B, 25% of staff cut). Two new ideas rather than one, because the besties made two different claims: Opendoor wins iBuying because pricing-at-a-margin-of-safety is a software problem whose gains compound (Google-vs-Yahoo analogy), with Chamath disclosing 3-4% ownership and Sacks disclosing a seed position; and separately that Zillow's failure was structural, not execution - Zestimate was lead-gen where 'accuracy is a bug', 3% margin against Opendoor's 10%, and Friedberg's point that Zillow claimed to be a market maker while actually speculating off model-projected prices with nobody doing ground-truth work. REINFORCED ELSEWHERE: carbon tariffs - Chamath repeats that an effective carbon tariff is 'the most disruptive thing in the capital markets and geopolitics that can happen in the next 10 or 20 years' and that 'it's coming', plus the developing world's $1.3T/yr ask, straight into carbon-pricing-markets-expand-2021; Friedberg's science corner on the Chinese cell-free enzymatic CO2-to-starch paper (10x corn's efficiency) is his bio-manufacturing thesis with a new proof point, and his 27-million-tanks / 25-by-25-miles / one-nuclear-plant build-out is the same capacity-is-the-constraint argument as E31; Sacks predicts the bipartisan infrastructure bill now 'sails through' post-election, which is the trigger condition for the contractor-windfall idea; and Chamath's John Doerr story (never sold a single distributed Amazon share) reinforces top-tier-tech-decade-hold. DISCLOSED POSITIONS: Chamath long GOOGL+MSFT vs short the rest of big tech with meaningful leverage; Chamath 3-4% of Opendoor via cash plus the IPOB SPAC merger; Chamath invested in Syndica (Solana-style Web3 RPC/compute infrastructure); Sacks led Bird's Series A (BRDS listed the day of taping at ~$2.4B, six-month lockup, he stays on the board a year, bias is to distribute); Sacks seed investor in Opendoor. DELIBERATELY NOT CAPTURED: Friedberg never states a directional view on the big-tech spread - he asks about it and explains the mechanics (rates are low so carry is cheap, market beta nets out), so he is not counted as a second voice on it and the idea stays single-voice; Chamath's Web3-plumbing/OSI-stack thesis and his Syndica investment have no public instrument in Nov 2021; Bird is skipped because BRDS was delisted in 2023 and can never be priced; the DAO segment is Sacks brushing off something that does not exist yet ('I'll let you know when it actually gets here') with no instrument; the whole progressive-setback block (Youngkin sweep, Minneapolis rejecting defund, 44-51 GOP seat gain forecast, CRT/school-choice fight, Sacks' 'Republican decade') is politics with no instrument except the infrastructure-bill line; Friedberg's detracking argument that removing exceptionalism moves the whole curve left and hands the win to China/India is real but has no tradeable expression; and Sacks/Chamath saying Trump is 'nowhere in sight' and still banned from social was considered as a bearish datapoint for trump-spac-dwac-2021 and rejected as too indirect. Meta note: the pod openly blows up at 44:36 - Sacks calls it 'one of the stupidest f***ing episodes you've ever done' and Chamath agrees, which is why the back half is all politics and science.