Our trusty space friend and serial X enthusiast Elon Musk has just liquidated and exercised an option converting $3,900,000,000 of paper Tesla shares into actual cash. In our report this morning:

• Trade: Options exercise + immediate sale, TSLA
• Amount: ~$3,900,000,000 (transaction value)
• Trade date: 9 May 2026
• Filed: Day 2 of 2 (the legal maximum)
• Source: SEC Form 4

Why has he done it, Quant — what are we saying?

[ MY QUANT — INITIATING ]

Subject:                  Elon R. Musk
Filing:                   SEC Form 4 — TSLA option exercise + sale
Trade date:               9 May 2026
Filed:                    11 May 2026 — Day 2 of 2
Shares sold:              ~8,000,000
Average sale price:       $487.50
Gross proceeds:           $3,900,000,000
Estimated US federal tax: ~$1,500,000,000
Net cash to subject:      ~$2,400,000,000
Largest single sale of his career:           TRUE
TSLA closing price impact (day of trade):   -6.0%
Total TSLA divestment, 2024-2026 cumulative: ~$28,400,000,000
Remaining stake post-sale:                   ~13% of Tesla, Inc.
Estimated value of remaining stake:          $187,000,000,000
Buried statistic: At current TSLA market capitalisation,
   the subject's residual position exceeds the combined
   market value of every UK FTSE 100 retailer.

[ MY QUANT — COMPLETE ]

Blimey. “At current TSLA market capitalisation, the subject’s residual position exceeds the combined market value of every UK FTSE 100 retailer.” That is actually mental. It makes you realise that Tesla is an unbelievably big company — still trading at a market value of roughly $1.5 trillion, give or take a Bulgaria.

This is the single largest sale of his career. TSLA dropped 6% on the day. As a Tesla shareholder — as I’m sure you are (might be) — I think I’d want to know why.

Quant, what’s happening with his new chip side hustle?

[ MY QUANT — INITIATING ]

Subject:                  xAI / Tesla Dojo compute pipeline, 2026
Total chip-related capex YTD:                $1,400,000,000

Allocation breakdown:
  Custom silicon development:                ~$420,000,000
  NVIDIA H200 GPU procurement:               ~$680,000,000
  Datacentre buildout (Memphis, TN):         ~$300,000,000

GPU equivalent fleet:                        ~50,000 NVIDIA H200s
Power draw at full load:                     ~140 megawatts
                                             (the residential draw of Dover)

Stated purpose (Tesla 10-K, March 2026):     "AI compute infrastructure"
Stated purpose (subject's X account, same week): "we are going to crush"

Years since subject first announced custom Tesla chip:    9
Years since "Full Self-Driving by end of year" promised:  11

Buried statistic: Had the $1,400,000,000 been allocated
   to NVIDIA common stock in January 2024 instead of NVIDIA
   hardware, the position would today be worth approximately
   $5,200,000,000.

SUBJECT note (unsolicited): That figure exceeds today's gross sale proceeds.
   Retracting subject note. Subject does not editorialise.
   Withdrawn.

[ MY QUANT — EXPERIENCING DIFFICULTIES ]

Right. In plain English, then.

The man has spent a billion-and-a-half dollars this year on chips. Not chips you can eat — chips you can compute with. About a third of it on designing his own (Tesla’s been at this for nine years now, with limited success). About half on buying off-the-shelf NVIDIA chips, which is what everyone else uses. And the rest on building a power station — sorry, “datacentre” — in Memphis, Tennessee, that draws as much electricity as the entire town of Dover.

The official Tesla document filed with the SEC says all this money is going on “AI compute infrastructure.” The version of Musk on X, that same week, said “we are going to crush.” I leave it to you which one sounds more believable.

For context: he’s been promising Full Self-Driving — actual robotaxis, the holy grail — by “the end of this year” for eleven years now. There are children currently sitting their GCSEs who were born after that promise was first made.

And the absolute punchline — the one Quant tried to point out before remembering itself — is this. If Musk had simply bought $1.4 billion of NVIDIA shares in January 2024 instead of $1.4 billion of NVIDIA hardware, his position today would be worth about $5.2 billion. That’s more than the entire sale he made this morning. He could have stayed in bed, bought the company that makes the pickaxes, and come out a billion and a half ahead of where he currently is. The gold rush, as ever, was won by the shovel-seller.

So — am I in?

Here’s the interesting bit. Why might Elon Musk selling a billion-dollar chunk of his own company actually be a good thing? Possibly even a buy signal?

If you want to go deep on chip manufacturing, all roads lead to one company: ASML, based in Veldhoven, Netherlands. They make the lithography machines — the only machines on Earth capable of carving the circuits onto next-generation chips. Every NVIDIA GPU, every Apple M-series, every TSMC wafer: made on ASML’s machines. Each machine costs $200–400 million and waiting lists run for years. ASML is the actual upstream monopoly of the entire AI boom.

For Musk to build his own chips at scale, he needs to get to the front of the ASML queue. If he can pull it off, the difference is between walking to China to get a chip approved versus walking down the hall. The Memphis money isn’t being spent on a vanity project — it’s being spent on access to the most constrained piece of industrial infrastructure on the planet.

The tax bill on this sale alone — $1.5 billion — equals the full annual income of roughly 20,000 American households. He has paid, in a single afternoon, the combined earnings of a medium-sized British town.

Reader, I am watching. I’m not selling my (small, ETF-held) Tesla exposure. If this money is genuinely heading into AI and chip infrastructure rather than the next Twitter bonfire, I’ll be holding my bag. The Space Cadet has done daft things. He has also, occasionally, done genuinely visionary things. Today’s filing is — for now — too early to call.

I shall be watching closely.

(This is not financial advice)