Inside the Balancing Act Over Google’s Compute Crunch
The Takeaway
- Google forms executive council to allocate scarce computing capacity internally.
- Compute shortage creates internal tensions, impacting Cloud, AI, and core products.
- Alphabet plans $91B-$93B capex this year to boost capacity, nearly double 2024.
Earlier this year, Google convened a group of its top executives to address what some staffers see as an existential problem: the company doesn’t have enough computer chips for everyone who wants them.
A shortage of computing capacity was creating internal tensions, as leaders of Google’s varied product areas struggled to agree on where Google should invest its resources. The gathering was the first meeting of a new council of senior executives aimed at streamlining decision-making and evening the playing field between different departments, according to a person familiar with the situation. The group has the job of deciding how to allocate Google’s computing capacity between Google Cloud, products such as search, and AI research unit Google DeepMind, according to people familiar with the process.
On the council are Google Cloud CEO Thomas Kurian; DeepMind chief Demis Hassabis; Koray Kavukcuoglu, DeepMind’s CTO and Google’s chief AI architect; Nick Fox, head of Knowledge and Information (which includes products such as Google Search and ads); James Manyika, who oversees a group of researchers and experimental product initiatives; and Chief Financial Officer Anat Ashkenazi.
The group considers factors including projections of how much revenue Cloud expects to generate, as well as how much Google’s products are expected to grow, according to two of the people. But those forecasts are often imprecise, and the demand for computing capacity far outstrips the supply, the people said, forcing the executives to make hard choices. They sometimes forward decisions to Pichai and Alphabet’s board for a final stamp of approval, the people said.
In the balance are key businesses, including Google’s ability to continue developing state-of-the-art AI models, which require endless amounts of computing to keep improving, especially as new models from rival labs constantly leapfrog each other in terms of abilities. At the same time, Google Cloud—which has emerged as a major growth driver for the company—requires access to more and more capacity to serve customers and maintain its expansion.
And then Google must have the capacity to deliver its own sprawling network of products, seven of which—including Search, Chrome and Gmail—have over 2 billion monthly active users. Advertising running alongside those products accounts for the bulk of the company’s revenue.
In a statement, a Google spokesperson said, “We have a rigorous, ongoing process that ensures our compute resources are allocated to the most important priorities, balancing today’s customer and user needs along with our long-term investments to advance research and innovation.”
To meet demand for capacity, Google, like other big tech firms, has aggressively ramped up its capital expenditures on chips and servers. Pichai said on the company’s most recent earnings call that Alphabet expects to spend $91 billion to $93 billion this year in capex, nearly twice as much as in 2024.
But because of the long lead time it takes to build data centers and manufacture computer chips, capex doesn’t have an immediate impact on capacity. Google’s chip capacity now is based on its spending in previous years. In 2023, for example, Google spent a comparatively puny $32 billion on capex.
Inadequate Capacity
Tensions over computing capacity have been an issue at tech companies for years, but they have intensified in the past two years as the needs of AI have escalated. All of the major cloud firms have said they don’t have enough capacity to meet customer demand.
Ashkenazi told analysts on Google’s last earnings call in October, “While we have been working hard to increase capacity and have improved the pace of server deployments and data center construction, we still expect to remain in a tight demand/supply environment in Q4 and 2026.”
Before Google created the new computing capacity allocation group, Machine Learning Strategy and Allocation, a much larger body involving leaders from across Google’s product areas that sometimes struggled to come to a consensus, made those decisions. The new group still receives input and recommendations from the MLSA group, according to two people familiar with the process.
The new group’s slimmed down membership provides a window into which divisions—AI, Cloud, and Knowledge and Information, Google’s ads cash engine—have the most weight within Google.
In 2025, Google expected to allocate around half of its computing capacity to Cloud, Ashkenazi said at a Morgan Stanley conference this spring.
Google’s computing capacity planning is a constantly evolving process, according to people familiar with the situation. Even after broad buckets of computing capacity are allocated to different divisions, computing needs change as products launch and catch on with users. Google’s computing needs could change next year as well: for example, if it makes a major AI research breakthrough, if a big Cloud customer comes in or if a competitor makes an advance that forces it to respond.
Other potential pitfalls lie in wait when it comes to actually using the chips. Any snags in Google’s complicated supply chain could result in longer waits for chips becoming usable, forcing Google to adapt again.
One silver lining for Google is that its tensor processing units—the in-house chips it uses to train its AI models and also rents to some Cloud customers—are getting better at the same time as its large language models are becoming more efficient. That means Google can deliver better AI performance without as many chips.
For rank-and-file employees, computing capacity allocation looks different on the ground.
The Cloud division bases its decisions on how to allocate chips largely on what will make the most money for Google, according to two people with knowledge of the process.
Within Google DeepMind, Kavukcuoglu and vice presidents under him make many of the computing capacity allocation decisions, according to two other people with knowledge of the process. Researchers are supposed to spend the majority of their time working on one main project, for which DeepMind will allocate computing capacity.
But there are always workarounds. Star individual contributors can end up working on multiple projects at once, ending up with various pools of compute they can draw from, the people said. If researchers run out of compute, they can try to borrow from other teams, either trading the implicit promise of a future tit-for-tat or offering other resources such as debugging help, three people said.
Research Calls I
-
Upgrades:
- Bain Capital Specialty Finance (BCSF) upgraded to Outperform at Keefe Bruyette, tgt $16
- Cummins (CMI) upgraded to Overweight from Equal Weight at Barclays, tgt $546
- Generac (GNRC) upgraded to Overweight from Equal Weight at Wells Fargo, tgt $195
- PACCAR (PCAR) upgraded to Overweight from Neutral at JPMorgan, tgt $133
- ServisFirst Bancshares (SFBS) upgraded to Outperform from Market Perform at Hovde Group, tgt $89
- Stryker (SYK) upgraded to Outperform from Market Perform at Citizens, tgt $440
- ZIM Integrated Shipping (ZIM) upgraded to Hold from Sell at Fearnley, tgt $20
- Bain Capital Specialty Finance (BCSF) upgraded to Outperform at Keefe Bruyette, tgt $16
-
Downgrades:
- AECOM (ACM) downgraded to Equal Weight from Overweight at Barclays, tgt $100
- AGCO Corporation (AGCO) downgraded to Underweight from Equal Weight at Barclays, tgt $93
- Air Products & Chemicals (APD) downgraded to Equal Weight from Overweight at Wells Fargo, tgt $250
- Allegiant Travel (ALGT) downgraded to Outperform from Strong Buy at Raymond James, tgt $98
- Birkenstock Holding plc (BIRK) downgraded to Hold from Buy at Williams Trading, tgt $51
- Celanese (CE) downgraded to Equal Weight from Overweight at Wells Fargo, tgt $45
- Eagle Materials (EXP) downgraded to Hold from Buy at Truist, tgt $210
- Eastman Chemical (EMN) downgraded to Equal Weight from Overweight at Wells Fargo, tgt $70
- Elevance Health (ELV) downgraded to Hold from Buy at Deutsche Bank, tgt $320
- LinkBancorp (LNKB) downgraded to Market Perform from Outperform at Hovde Group, tgt $9.25
- Lockheed Martin (LMT) downgraded to Neutral from Overweight at JPMorgan, tgt $515
- LyondellBasell (LYB) downgraded to Equal Weight from Overweight at Wells Fargo, tgt $45
- Lyft (LYFT) downgraded to Underperform from Neutral at Wedbush, tgt $16
- AECOM (ACM) downgraded to Equal Weight from Overweight at Barclays, tgt $100
-
Others:
- AeroVironment (AVAV) initiated with an Overweight at KeyBanc, tgt $285
- Arthur J. Gallagher (AJG) initiated with an Overweight at Morgan Stanley, tgt $300
- Bristow Group (VTOL) initiated with a Buy at Texas Capital, tgt $52
- Capital Clean Energy Carriers (CCEC) initiated with a Buy at BTIG Research, tgt $25
- Century Aluminum (CENX) initiated with a Buy at Texas Capital, tgt $42
- Circle Internet Group (CRCL) initiated with a Neutral at H.C. Wainwright, tgt $85
- CoreWeave (CRWV) resumed with a Buy at Citigroup, tgt $135
- Dnow (DNOW) initiated with a Buy at Texas Capital, tgt $19
- Energy Fuels (UUUU) initiated with a Buy at Texas Capital
- enCore Energy (EU) initiated with a Buy at Texas Capital
- Firefly Aerospace (FLY) initiated with a Sector Weight at KeyBanc
- Insmed (INSM) assumed with a Buy at Truist, tgt $202
- Intuitive Machines (LUNR) initiated with an Overweight at KeyBanc, tgt $20
- Karman Holdings (KRMN) initiated with an Overweight at KeyBanc, tgt $80
- Kratos Defense & Security (KTOS) initiated with an Overweight at KeyBanc, tgt $90
- LivaNova (LIVN) initiated with an Overweight at KeyBanc, tgt $81
- Oculis Holding AG (OCS) initiated with an Overweight at JPMorgan, tgt $38
- Parsons (PSN) initiated with an Overweight at Barclays, tgt $70
- PennyMac Financial (PFSI) initiated with a Buy at Jefferies, tgt $160
- Redwire (RDW) initiated with a Sector Weight at KeyBanc
- Rocket Companies (RKT) initiated with a Buy at Jefferies, tgt $25
- SAB Biotherapeutics (SABS) initiated with a Buy at Guggenheim, tgt $15
- Septerna (SEPN) initiated with a Strong Buy at Raymond James, tgt $38
- Synaptics (SYNA) initiated with an Overweight at Wells Fargo, tgt $95
- Terrestrial Energy (IMSR) initiated with an Overweight at Cantor Fitzgerald, tgt $12
- UR-Energy (URG) initiated with a Buy at Texas Capital
- UWM Holdings (UWMC) initiated with a Hold at Jefferies, tgt $5
- Walker & Dunlop (WD) initiated with a Buy at Jefferies, tgt $75
- AeroVironment (AVAV) initiated with an Overweight at KeyBanc, tgt $285
Gapping down
In reaction to earnings/guidance:
In reaction to earnings/guidance:
- NTNX -14.7%, NKE -9.4%, SCHL -1.3% (also authorizes new $150 mln share repurchase program)
Other news:
-
NVA -16.3% ( announces pricing of public offering)
-
SAVA -14.4% (no longer expects to initiate a proof-of-concept clinical trial for simufilam in TSC-related epilepsy in the first half of 2026 as previously disclosed )
-
CGTX -2.7% ( also entered into an Open Market Sale Agreement )
-
ALT -2.6% (announces that pemvidutide achieved key measures of success at 48 weeks in IMPACT Phase 2b MASH Trial; Conference call to be held today at 8:00 a.m. ET)
-
GLPG -2.3% (topline results from two Phase 3 enabling studies),
-
ENVA -1.4% (and Grasshopper amended merger)
-
CGTX -1.3% (files for $300 mln mixed shelf offering; also enters sales agreement up to $75 mln of stock),
- BANX -1.3% (announces $14.9 million registered direct offering )
Gapping up
In reaction to earnings/guidance:
In reaction to earnings/guidance:
-
WGO +7.4% ( beats by $0.25, beats on revs; guides FY26 EPS in-line, revs in-line),
AVO +5.1% (also Founder & CEO Steve Barnard to become Exec Chairman; COO to become new CEO),
NNE +2.5%, FDX +2.1%
Other news:
- MIAX +4.4% (MIAX to sell 90% of MIAXdx to HOOD)
-
WYFI +13.6% (announces 40 MW colocation agreement)
-
LNKB +6.5% (BHRB and LNKB to merge)
-
CMCL +6.4% (: Zimbabwe government amends proposed changes to the royalty and tax regimes )
-
VLRIS +2.2% (& Viva announce the formation of a new Mexican Airline Group to accelerate the growth of air travel and connectivity in Mexico )
-
OCUL +2.2% (expects to provide an update on the timing of SOL-1 data, which remain on track for the first quarter of 2026)
-
TAC +2% (Sheerness Unit 1 will be temporarily mothballed)
-
TM +1.7% (aims to begin selling U.S. vehicles in Japan from 2026)
-
LEU +1.5% (launches commercial LEU enrichment activities )
-
WULF +1.4% (financing for previously disclosed 168 MW computing joint venture),
-
IVR +1.1% (increases dividend; also plans to switch to monthly dividends in Jan),
- IBN +1% (provides initial public offering of subsidiary, ICICI Prudential Asset Management Company, update)
>>> Up
* Abivax price target raised to $175 from $150 at Guggenheim
* Cicor Tech Raised to Buy at Baader Helvea; PT 195 Swiss francs
* Cicor Tech Raised to Buy at Baader Helvea; PT 195 Swiss francs
* ZIM Integrated Shipping Raised to Hold at Fearnley; PT $20 (+)
>>> Down
>>> Down
* Arcadis Cut to Neutral at UBS; PT 40 euros
* Birkenstock PT Cut to $60 from $80 at Jefferies
* Computacenter Cut to Hold at Peel Hunt; PT 3,000 pence
* Foresight Solar Cut to Hold at Jefferies
* Rieter Cut to Neutral at UBS; PT 3.30 Swiss francs
* Santander Cut to Hold at DZ Bank; PT 10 euros (+)
* Schweiter Cut to Reduce at Baader Helvea; PT 229 Swiss francs
* Schweiter Cut to Reduce at Baader Helvea; PT 229 Swiss francs
* Talenom Cut to Accumulate at Inderes; PT 3.40 euros
>>> Initiation
>>> Initiation
* Also Rated New Outperform at Oddo BHF; PT 272 Swiss francs
* Axfood Rated New Buy at Handelsbanken; PT 327 kronor (+)
* EMS-Chemie Rated New Neutral at Oddo BHF; PT 560 Swiss francs
* Ependion Rated New Hold at Danske Bank Markets; PT 118 kronor (+)
* Husqvarna Rated New Neutral at SB1 Markets; PT 50 kronor
* Magnum Ice Cream Rated New Buy at Kepler Cheuvreux
* Sandoz Group Rated New Neutral at Oddo BHF; PT 57 Swiss francs
* Sandoz Group Rated New Neutral at Oddo BHF; PT 57 Swiss francs
* Technip Energies Rated New Buy at Goldman; PT 40 euros
>>> Call
>>> Call
* Goldman Says Next Phase of Market Rally to See Higher Volatility
* Goldman Sees Equity Gains in 2026, Although Lower Than This Year
Asian stocks followed gains in US equities as cooling inflation backed the case for Federal Reserve interest-rate cuts. The yen fell as traders bet the Bank of Japan will stay cautious despite raising rates Friday as expected. The MSCI Asia Pacific Index rose 0.6%, with technology giants such as SoftBank Group Corp. and Tencent Holdings Ltd. among the biggest contributors. The S&P 500 climbed 0.8% Thursday, while the tech-heavy Nasdaq 100 rose 1.5%. Gains were helped by a solid outlook from giant Micron Technology Inc., easing concerns over artificial intelligence spending and valuations. The US inflation data bolstered sentiment despite caveats over the data tied to the recent government shutdown, with markets focusing instead on the slowest increase in consumer prices since early 2021. The print boosted investor confidence and bolstered Treasuries on renewed expectations of Fed rate cuts. The yen dropped against all its Group-of-10 peers and Japan’s benchmark bond yield climbed to 2% for the first time in almost two decades after the BOJ raised its benchmark rate to the highest level since 1995. All 50 economists surveyed by Bloomberg had predicted that outcome. BOJ Governor Kazuo Ueda will elaborate on his thinking behind Friday’s policy decision and the future rate trajectory at a press briefing starting at 3:30 p.m. Tokyo time. The slowdown in US inflation came with some data issues. The Bureau of Labor Statistics couldn’t collect prices throughout October because of the temporary government shutdown, and it started sampling later than usual in November. Swaps are implying about 20% odds of a Fed rate cut in January, with a reduction fully priced in by mid-2026. Traders are also sticking with their call that the central bank lowers rates twice next year. In commodities, crude headed for a second weekly loss despite tensions around a US naval blockade of sanctioned tankers calling at Venezuela, with futures weighed down by expectations for a global surplus. Brent has slipped more than 2% this week. Precious metals remained in favor. Platinum was near $1,930 an ounce, on track for a seventh day of gains and close to the highest level since 2008. The surge has come as the London market shows signs of tightening, with banks parking metal in the US to insure against the risk of tariffs. US After Hours NKE -9.4% lower on earnings; FDX +2.1% higher on earnings; LNKB +6.5% higher on merger with BHRB; WSJ reports that OpenAI financing round may push valuation toward $830 bln.
Nikkei +1.03 Hang Seng +0.65% CSI +0.33% Shanghai +0.33% Shenzen +0.97%
Eur$ 1.1718 CNH 7.0356 CNY 7.0412 JPY 156.08 GBP 1.3372 CHF 0.7952 RUB 79.8855 TRY 42.8074 WTI$ 56.02 -0.13% Gold 4,322 -0.27% BTC 87,050 +1.68% ETH 2,919 +3.05% SOL 122.8580 +2.12%
S&P -0.08% Nasdaq +0.à8% EuroStoxx -0.21% FTSE -0.39 Dax -0.19% SMI -0.19%
- Tech Billionaires Pledge $1 Billion for CERN’s Next Breakthrough
- ‘Swiss Cheese’ CPI Report Raises Doubts About US Price Data
- ECB Officials Say Cycle of Rate Cuts Is Most Likely Over
- Goldman Says Next Phase of Market Rally to See Higher Volatility
Keep an eye on :
Keep an eye on :
- ABVX US : Abivax to be Added to Nasdaq Biotechnology Index
- Altice : Patrick Drahi Creditors Are Suckers for Punishment: Chris Hughes
- AMZN US : Amazon, Google Among 24 Firms Joining US AI ‘Genesis Mission’
- AZN LN : Niowave Expands AstraZeneca Deal to 10-Year Actinium-225 Supply
- AZN LN : Syneron Bio Raises Nearly $100m in Series A and A+ Rounds
- BIOAB SS : EQT Life Sciences Acquires 347,500 Shares in BioArctic
- BMW GY : BMW Extends Share Buyback With Second Tranche
- COTY US : Coty Is Said to Sell Remaining 25.8% Stake in Wella to KKR
- ACA FP : French AMF Fines Caceis Bank €3.5M for Failings Related to H20
- EQT SS : EQT Life Sciences Acquires 347,500 Shares in BioArctic
- RF FP : Eurazeo Agrees to Sell About €260M of Assets to Third Party
- FAGR BB : Fagron Makes Two Acquisitions For ~€55m, Gets New US License
- FDX US : FedEx CEO’s Sweeping Network Overhaul Is Starting to Bear Fruit
- GLPG NA : Galapagos TYK2 Inhibitor Succeeds in a Study, Misses in Another
- HBH GY : Hornbach Holding 3Q Adjusted Ebit EU27.3M Vs. EU34.6M Y/y
- IBAB BB : IBA Buys ORA for Total Consideration €15M-€20M
- IPN FP : Ipsen Says Phase II Falkon Study Didn’t Meet Primary Endpoint
- LiNDEX FH : Lindex Explores Separation of Department Store Business
- MATAS DC : Matas Appoints Mette Uglebjerg as Group CEO Starting 1 May 2026
- BMPS IM : Italy Didn’t Influence Paschi’s Bid on Mediobanca, Minister Says
- NEXI IM : Nexi Rejects TPG’s €1 Billion Bid for Digital Banking Unit
- NEXI IM : Nexi Rejects TPG’s €1 Billion Bid for Digital Banking Unit
- NKE US : Nike 2Q Revenue Beats Estimates: Snapshot --> -4% in after hours
- NKE US : Nike’s Weakness in China, Converse Spur Growing Concern
- NOBA SS : Noba Bank Buys DBT Capital for SEK403m in Cash
- NVDA US : House Democrats Introduce Bill to Ban US AI Chip Sales to China
- PIRC IM : Sinochem’s Pirelli Stake Down to 34% After Bond Conversion
- PRX NA : Just Eat to Redeem 2026, 2028 Convertible Bonds on Jan. 19
- PRY IM : Prysmian to Proceed With Channell $200M Earn-Out Payment
- PUM GY : Puma Secures €500M Bridge Loan, Confirmed Credit Lines of €108M
- RNO FP : *RENAULT RAISED TO INVESTMENT GRADE BY S&P
- RWE GY : PGE Signs Conditional Deal to Buy Offshore Wind Project From RWE
- RWE GY : *RWE, INDIANA MICHIGAN POWER COMPANY SIGN LONG-TERM PPA
- SCR FP : Scor Renews Contingent Capital Program Through 2028
- TSLA US : Elon Musk’s SpaceX bought tens of millions worth of Cybertrucks Tesla can’t sell - Electrek
- TSLA US : Tesla’s ‘Musk Premium’ in Focus With SpaceX IPO in View
- TMUS US : SpaceX, AT&T Spectrum Licenses Plan Queried by Warren, Democrats
- VWS DC : Vestas Soars 74% in 2025 on Order Growth, Easing Policy Fears
- WYFI US : *WHITEFIBER SHARES JUMP 15% ON DATA CENTER PACT WITH NSCALE
>>> Up
* Abivax price target raised to $175 from $150 at Guggenheim
* Cicor Tech Raised to Buy at Baader Helvea; PT 195 Swiss francs
>>> Down
* Cicor Tech Raised to Buy at Baader Helvea; PT 195 Swiss francs
>>> Down
* Arcadis Cut to Neutral at UBS; PT 40 euros
* Birkenstock PT Cut to $60 from $80 at Jefferies
* Computacenter Cut to Hold at Peel Hunt; PT 3,000 pence
* Foresight Solar Cut to Hold at Jefferies
* Rieter Cut to Neutral at UBS; PT 3.30 Swiss francs
* Schweiter Cut to Reduce at Baader Helvea; PT 229 Swiss francs
* Schweiter Cut to Reduce at Baader Helvea; PT 229 Swiss francs
* Talenom Cut to Accumulate at Inderes; PT 3.40 euros
>>> Initiation
>>> Initiation
* Also Rated New Outperform at Oddo BHF; PT 272 Swiss francs
* EMS-Chemie Rated New Neutral at Oddo BHF; PT 560 Swiss francs
* Husqvarna Rated New Neutral at SB1 Markets; PT 50 kronor
* Magnum Ice Cream Rated New Buy at Kepler Cheuvreux
* Sandoz Group Rated New Neutral at Oddo BHF; PT 57 Swiss francs
* Sandoz Group Rated New Neutral at Oddo BHF; PT 57 Swiss francs
* Technip Energies Rated New Buy at Goldman; PT 40 euros
>>> Call
>>> Call
* Goldman Says Next Phase of Market Rally to See Higher Volatility
* Goldman Sees Equity Gains in 2026, Although Lower Than This Year
Which genius from history would have been the best investor?
FTAV goes hedge fund headhunting through the ages
With hedge fund founders peppering the Forbes list of billionaires, top traders getting paid $100mn, and even interns being offered $35k a month, you can maybe see why it’s become a common complaint that hedge funds suck talent away from the rest of the economy.
But money, if you believe the hype, is a neutral medium measuring societal contributions. So maybe the best and brightest should trade their lab coats for gilets, notebooks for spreadsheets, and just generally stop trying to change the world for so-called “good”?
This got us wondering which hedge funds might’ve lured away geniuses of yesteryear. Would Beethoven be at Balyasny, Mozart at Millennium and James Prescott Joule at Jane Street? That in turn led to the question of whether scientific or artistic brilliance ever translated into investing success.
It’s late December, so it’s the season for some vibes-based counterfactual whimsy. But this is Alphaville, so our vibes-based whimsy will be rooted by cool academic papers based on detailed trading records of our subjects. Luckily, there’s no shortage of such papers. So let’s begin.
Sir Isaac Newton
Newton made professor of mathematics at Cambridge at the age of 26, and in Philosophiæ Naturalis Principia Mathematica, he established the laws of motion, of universal gravitation, formulated infinitesimal calculus and basically laid out the dominant scientific consensus that stood for centuries. So it seems a slam dunk that today he’d be bid away from academia by one of the top quant prop shops.
But would his genius translate into trading smarts? We don’t need to guess. Newton lived and invested through the South Sea bubble — one of the most famous financial manias of all time. And in the careful work of Professor Odlyzko of the University of Minnesota, we can see quite how badly he did.
It started with such promise. Newton bought shares in the South Sea Company back in the foothills of the bubble, and by the beginning of 1720 had amassed around 10,000 shares worth ca £300mn in today’s money (indexing to GDP rather than CPI), together with a further £19,000 of gilts (worth about £570mn today).†
Moreover, Newton cashed out the bulk of his South Sea position in April and May 1720 when the company’s stock price was closing in on £400. That translated into profits of over 100 per cent. With a £20k gain banked (or £600mn in today’s money), his total portfolio value was taken to maybe just north of £40k (£1.2bn in today’s money). Alex Gerko eat your heart out.
But as mania turned to delirium Newton found himself dragged back in by his inner YOLOer — trading his cash and gilts into South Sea shares at the market peak mid-1720.
It looks like the physicist knew he was trading a bubble, but thought he was trading it well. Newton’s famous quote — “I can calculate the movement of stars, but not the madness of men” — wasn’t some reflection on the bursting of the bubble, but an answer to a question posed at the peak as to where the mania would end. Gotta keep dancing.
By 1721 his entire portfolio consisted of shares in the South Sea Company, with a total value of around £20,000.
Using Odlyzko’s figures, had Newton ploughed all his money into South Sea shares at the end of 1719 and just left it untouched, he would’ve done great. Fine, the share price rose almost 10-fold and then collapsed, but by August 1723 on a split-adjusted basis, South Sea Co shares were trading around 35 per cent higher than their November 1719 value. Newton’s fortune meanwhile had dropped by around 38 per cent.
Still, despite outsized losses Newton died an immensely wealthy man with an estate valued at around £30,000 — almost a billionaire in today’s money.
So while we can imagine XTX Markets, G-Research, Renaissance Technologies, Two Sigma or DE Shaw all engaging in a high stakes bidding war for the OG maths genius, we can also see Newton’s sheer appetite for risk might make Andurand Capital a better fit.
Winston Churchill
Churchill played a pivotal role in saving Europe from fascism and won the Nobel Prize for literature, but his relationship with money was less laudable. OK, it was chaotic.
Born in Blenheim Palace, son of the chancellor of the exchequer, and grandson to the 7th duke of Marlborough, Churchill wasn’t exactly a product of the school of hard knocks. But the man could spend with sufficient aplomb to be almost permanently in serious debt. As such, Churchill got to work making money through a combination of [checks notes] journalism and book-writing.
But what about his investment record? Leaving ministerial office in May 1929 he faced £5,250 of overdue bills, was massively overdrawn, but expected earnings of £12,700 from writing — largely in the form of a mammoth advance for a biography of his ancestor, the first duke of Marlborough. Rather than settle debts, Churchill instructed his publishers to send their Marlborough cheques straight to his broker, which he spivved into stocks and departed for North America on a speaking tour.
According to David Lough, Churchill’s financial biographer, he found himself intoxicated by America’s moneymaking opportunities and gripped by investment fever. He picked up stakes in small exploration companies, oilfields, furniture, retail, gas and electric companies. He began to trade on margin, and within four months was sitting on earnings of £22k (£13.8mn in today’s money).
As the market rolled, so Churchill’s trading turnover tended towards the Citadellian. During the nine trading days ending 18 October he traded $620k worth of stock (£78mn adjusting for contemporary exchange rates and inflating with UK GDP). Did it work?
No. Churchill waited until he saw his wife in person back in the UK before telling her that he had lost the equivalent of all his advances for the Marlborough book before he’d written a word.
The following year, despite carrying an overdraft with an increasingly angsty Lloyds Bank for £13,700 (£8.7mn today), the senior partner of his brokerage felt compelled to advise him to please stop buying worthless “gambling stocks” and by the middle of 1930 he’d lost a further £7k (£4.5mn today) on financial markets. Nancy/Paul Pelosi he was not.
We got in touch with Lough, whose excellent book on Churchill’s chaotic finances gives us almost all of what we know about the details. He told Alphaville:
Churchill never had the spare money to be a successful investor. He had a limited number of US stocks which he would range trade, often on margin. He had to be in and out within a few days. He got his information from the International Herald Tribune which he had flown over from Paris to Croydon Airport, then by rail to Westerham and taxi to Chartwell [his home]. The secretaries knew which stocks to look for and wrote them out a piece of paper which his valet took into meetings on a silver tray.
Churchill needed to be financially rescued a number of times throughout his life, two of which were arranged by the Financial Times’ very own Brendan Bracken. So it’s hard to think of a hedge fund that would have stolen Churchill away from his career in politics and journalism. Though maybe — just maybe — Churchill could’ve charmed Bill Hwang into providing him some seed money.
Charles Darwin
By 1873 Darwin was famed as the pre-eminent naturalist, biologist and geologist of his age. He is credited as the father of evolutionary biology, responsible for the unifying theory of the life sciences. But could the man trade stocks and bonds?
Responding to a questionnaire sent to selected Fellows of the Royal Society asking to list mental qualities that contributed to his success, Darwin answered that he had no special talents. Well, “none, except for business as evinced by keeping accounts, replying to correspondence, and investing money very well.”
Professor Janet Browne picked through the numbers and found Darwin a careful and successful investor, writing mortgages on Shropshire property and riding the wave of railway companies, canals and dockyard schemes into the 1860s before wisely flipping his portfolio into consols in the mid-1860s (there was a massive crash in 1873).
His account books show that he began married life in 1839 with a marriage bond of £10,000 from his father, £573 in the bank and £36 in his pocket — a total of around £55mn in today’s money. Marriage bonds, explains Browne, were not realisable assets but intended to provide income during the holder’s lifetime and to be passed to children at death. By 1881 Darwin had grown his capital to £282,000, excluding earnings from his books (£652mn today).
Achieving an annualised real growth in capital of 8.6 per cent per annum over 42 years looks impressive — although we’re not entirely sure how much his capital was swollen by two large inheritances.
We’re therefore not ready to accredit with Darwin with Buffett-level investment acumen, but his big trade out of railroads stocks and into gilts is enough to have delivered stellar returns and looks symptomatic of the kind of successful macro punting smarts that Soros or Brevan Howard look for.
JMW Turner
Turner is best known today for his oil paintings, watercolours, and the Tate’s contemporary art prize that bears his name. But he also made time for some light fixed income arbitrage trading.
According to Bank of England records, Turner bought his first gilt at the age of 19. Sure, back in the 19th century anyone who was anyone kept an eye on the gilt market. But according to — again — Andrew Odlyzko, Turner seriously over-indexes on this front.
Following the Napoleonic Wars, British debt-to-GDP was running close to 200 per cent. Were people bothered? Absolutely. London, we are told, was rife with chatter predicting a catastrophic debt crisis.
Back in the early 19th century debt was overwhelmingly in the form of perpetuals — so-called Consols. The thing about perpetuals is that while they definitely need servicing, they never need repaying. How then to pressure the government to start paying down debt and avert financial Armageddon? A bunch of parliamentarian debt-hawks came up with a cunning wheeze: create Terminable Annuities — streams of cash flows that ran out in thirty years — into which existing bondholders could switch.
But while almost all government debt was in the form of perpetuals, 3 per cent of the debt stock was in the form of so-called Long Annuities — non-perpetual debt with a 30-year term and a face value worth maybe 50 basis points more than the new TAs. Contemporary press accounts described TAs and LAs as identical, but bond geeks worked out you could pick up LAs below par, deliver them into the exchange and make a quick 340 basis point turn on the capital.
As Odlyzko puts it:
This was a totally risk-free exchange, one stream of payments from the British government for another, almost identical. There were no fees. All an investor had to do was to appear in the office of the Commissioners for the Reduction of the National Debt in Old Jewry Street, which sold the new annuities, sign the papers, then walk two blocks to the Bank of England to carry out the transfer of LA to the government, and then go back to complete the transaction.
Today such an arb looks fanciful, but in 1829 it took six weeks for the weight of money to close it down. Less than a tenth of the stock of LAs were delivered into the exchange, but Turner managed to jam about £10k through the books. In today’s money that’s a £66mn fixed income arb trade.
OK, he’s got the trading smarts, but what about his character? The then contemporary artist John Constable famously wrote of Turner that “he is uncouth but has a wonderful range of mind”. And Sir Walter Scott wrote to a friend that “Turner’s palm is as itchy as his fingers are ingenious and he will, take my word for it, do nothing without cash and anything for it. He is almost the only man of genius I ever knew who is sordid in these matters.”
Solid Jane Street hire.
John Maynard Keynes
Mathematician, geopolitical strategist, economic adviser and general father of macroeconomics, Keynes had it all. Moreover, we don’t need to guess whether his true calling lay in HFT, macro-punting, fixed income arb, or long-short equities. Because from 1922 he managed the endowment for King’s College Cambridge, and the records have been meticulously pored over by a host of academics.
Warren Buffett, George Soros and David Swensen each cite Keynes as a role-model, and biographers offer some good colour. But David Chambers, Elroy Dimson and Justin Foo provide the numbers.
Keynes took over a portfolio entirely invested in bonds and made the heretical move to trade this almost entirely into stocks — an asset class traditionally the territory of individual investors. The result was a portfolio that averaged over 15 per cent per annum investment returns for 25 years.
This wasn’t just beta surfing either. Keynes outperformed UK stocks by 521 basis points a year, and suffered only six years of underperformance over the period — four of which were in his first seven years as manager. (For context, Warren Buffett’s Berkshire Hathaway has outperformed the S&P 500 index by a still-impressive 138 basis points per annum over the last 25 years.)
The power of compounding meant that while £100 invested at the start of Keynes’ tenure as manager of King’s College’s endowment would have grown to £506 if invested in gilts, or £855 if invested in UK stocks, it would have ballooned to a whopping £2,893 if invested with Keynes by the time of his death in 1946. 🥂
Did he take outsized investment risks to achieve this outsized reward? Looking only at annual returns, Alphaville threw together an ex post efficient frontier. This suggests that yes, Keynes’ portfolio was more volatile than an index-tracking strategy. But this additional volatility was more than compensated:
How about investment style? In the 1920s he took a top-down approach and the authors find no evidence of any market-timing ability. But as Keynes wrote in 1934:
As time goes on, I get more and more convinced that the right method in investment is to put fairly large sums into enterprises which one thinks one knows something about and in the management of which one thoroughly believes.
And Chambers, Dimson and Foo’s quantitative analysis of his portfolio finds that from the early 1930s Keynes’ style evolved into a bottom-up stockpicker happy to run substantial active risk.
It also looks as though Keynes became a less active trader over time. The portfolio’s average annual turnover fell from 55 per cent a year in the 1920s to 30 per cent in the 1930s, and to only 14 per cent in the 1940s.
OK, high conviction qualitative bottom-up stock picking with relatively low turnover that compounds serious excess returns over decades? He’d have a prime risk-taking seat at Chris Hohn’s TCI Fund.
George Frideric Handel
Alphaville was excited to discover that the 18th century German-British composer’s financial records had been analysed in detail by Professor Ellen Harris.
Harris documents Handel’s bank annuities, as well as his holdings in the Royal African Company — responsible for shipping more slaves to the Americas than any other company. You’d think that this would provide rich pickings from which a performance track record might be assembled.
However, she argues — quite convincingly — that because the accounts in which these investments were registered were opened following major musical works, and because the accounts are no sooner opened in Handel’s name as closed by him, what we are actually seeing are payments in stock/ bonds by benefactors (notably the Duke of Chandos) rather than investment decisions.
This reminds Alphaville of that time George HW Bush took Global Crossing stock as payment for a speech, but either forgot or chose not to cash it. The share price at one point leapt 175x to take the speech’s value to over $14mn but then came crashing back down as the firm spiralled into bankruptcy.
Anyway, the Messiah looks safe. No hedge fund job for Handel.
What about the rest?
We’re conscious that this has been a very male, British-centric, and rather incomplete list of geniuses. But we’ve got to work with what we’ve got.
Alphaville has decided that stories of Albert Einstein losing his 1921 Nobel Prize money by punting them in stocks and getting wiped out in the 1929 Wall Street crash is an urban internet myth. As far as we can ascertain he handed the bulk of the money to his ex-wife as part of a divorce settlement which she used to buy Huttenstrasse 62, a five-storey apartment block in Zurich.
William Shakespeare looks to have ploughed his income into Stratford farmland rather than a stock portfolio. Jane Austen, despite writing in meticulous detail about the role that money and wealth played in early 19th-century society, was never paid enough to start investing. Goethe seemed happy to live off his inheritance rather than seek to build it further. Leonardo da Vinci made a good living as an artist and teacher, but we’ve found no documentation around his finances. And it looks like Marie Curie gave most of any money that came her way to advance scientific research rather than boost her capacity to buy flashy toys.
The idea that money is some neutral medium measuring societal contributions is, of course, a nonsense. And Alphaville is glad that it doesn’t take a genius to see that this is transparently the case.