Trump is threatening further action against Iran. In case you are wondering, this is where he’s put his money: - Chevron $CVX: $1 million to $5 million on 21 August, plus $15,001 to $50,000 on 26 August. - ConocoPhillips $COP: $1 million to $5 million on 21 August. Those purchases were disclosed on 8 October. My picks to watch if tensions escalate further: ConocoPhillips for direct exposure to higher oil prices, Chevron for oil exposure with a broader business and RTX $RTX for missile replenishment demand through Raytheon. But don’t confuse a good thesis with a good entry. Chevron and ConocoPhillips have already reached fresh annual highs. I’d favour a modest position on a pullback, keep cash available and avoid leverage. Buying the first spike after a strike will risk you entering at the highest point… and you know what that means. Now of course if you take a position before the news, the usual thesis counts / entering before the news for the event. This could get interesting. Watch this space.

More US firepower heading towards Iran. I’d take the escalation risk seriously. My pick is energy: $XOM or $XLE on a pullback. Further disruption to Gulf oil supplies would support crude prices and squeeze airline profits. I’d avoid buying $DAL or $UAL here. Expensive tech is vulnerable too if higher oil keeps inflation and interest rates elevated. Buy energy on weakness. Don’t chase a panic spike.

Would you pay €99 a month for your Tesla to handle the roundabouts, junctions and stop start traffic on your morning commute? You still have to watch the road and be ready to take over. But imagine the next step: your car takes you to work while you read, catches up with you outside the office and drives you home after a few drinks. Personally, I would love this experience! I tried Waymo recently and was super impressed. If Tesla can do the same, then that for sure is the future. So, the appeal of genuinely driverless cars is that you get your time back. For $TSLA, today’s supervised driving subscriptions could help fund that ambition and provide more real world experience to improve the software. European approval for supervised driving doesn’t give Tesla permission to remove the driver. That needs further technical progress and separate regulatory approval. But the business opportunity is easy to understand. First, earn monthly income from cars already sold. Eventually, if Tesla delivers, sell transport to people who don’t need to own a car at all. €99 a month to make your commute easier is one proposition. A car that can take you anywhere without you driving is a much bigger one…… what do you think?

How many bots a day do you block with all their WhatsApp or telegram links?

Let’s ponder something… if no new money is coming into the market, and certain stocks are rising, then it must be that other stocks are falling as the money is being shifted from one to the other..

🚨 SpaceX is reportedly planning a US wireless network. AT&T $T and T-Mobile $TMUS fell nearly 8%, while tower owners Crown Castle $CCI jumped 9% and American Tower $AMT gained 4%. More competition for the networks…. Or is a takeover brewing…..

$WDC customers are planning their storage needs years ahead, and that deserves a closer look. In September, Western Digital’s CFO said the vast majority of its 2027 business was covered by long term agreements, with portions of 2028 and 2029 covered too. Most customers are asked to place purchase orders at least 52 weeks ahead. On yesterday’s earnings call, management said it had visibility into demand from its seven biggest customers through 2030 and 2031. That’s customer planning visibility, not five years of guaranteed revenue, but it shows how far ahead these conversations now run. What is all that storage needed for? AI training datasets, saved model versions, generated video and records of what AI agents do. Robotics and autonomous vehicles add more camera and sensor data. Alongside that, cloud photo libraries, surveillance footage and company backups keep accumulating. Fast memory and SSDs handle the most demanding processing. Hard drives provide economical storage for the huge volumes customers want to retain. The latest numbers show both rising demand and stronger pricing. Storage capacity shipped rose 22% from a year earlier, while average selling prices per terabyte increased 28%. Revenue grew 55% to $4.38 billion, generating $1.38 billion in free cash flow. Some customer agreements also include scheduled price increases. That gives the margin improvement more substance than simply hoping today’s shortages continue. My bullish view is that storage demand keeps building as AI gets used, well beyond the initial construction of data centres. The risk is suppliers adding enough capacity to weaken pricing. I’d watch that alongside how much of the expected growth is already reflected in the share price.

At one point today, the S&P 500 was down 1.5% while 322 of its stocks were actually rising. Give every stock equal weight and the index was up 0.3%!!! Worth remembering when a red headline makes it feel like everything you could buy is falling….

$META has 3.64 billion people using its apps every day. Somehow, it’s still growing revenue by 26%. Users only grew 3%. The bigger difference was showing more ads and getting paid more for each one. You don’t need everyone to buy an AI subscription when you can improve a business that already brings in $64.6 billion a quarter… although AI consideration will be for the near future I bet. The spending does make you wince. It brought in $32.8 billion in operating cash and spent $34.9 billion on infrastructure and lease payments. Even this money machine is spending faster than the cash comes in. But advertisers are clearly still spending. And management expects another $73 billion to $76 billion in revenue next quarter. Analysts’ average target puts upside around 17% from Wednesday’s close, with several fresh targets pointing to 22% to 28%. This is currently too high I think for me to jump in, but after a little more of a pullback, I’ll keep an eye on it! I bet everyone will keep an eye on it if that happens 👌

You know what this means….

$SMCI just added roughly $4 billion to its quarterly revenue estimate. Supermicro now expects $18.5bn to $19.5bn for the September quarter, up from $14.5bn to $15.5bn. It says stronger AI demand drove the increase. The more interesting number is the margin: now 12% to 12.5%, versus 10.4% to 10.8% previously. Put those together and, using the midpoints, implied gross profit rises from about $1.59bn to $2.33bn. That’s roughly 46% above what the earlier guidance implied. More servers sold, with more money kept from each dollar of sales. This strengthens my view that the AI buildout still has plenty of spending ahead. Supermicro’s execution deserves scrutiny, but these figures give the demand argument some substance. Full results arrive on 3 November. These numbers are preliminary, and cash flow is the next thing I want to see. An extra $4bn of sales looks good. How much turns into cash will tell us considerably more.

AutoZone $AZO has spent $40.5 billion buying back its own shares since 1998. 156 million shares repurchased. Just 16.2 million still outstanding as of August 2026. Quite a shopping habit for a company selling brake pads and batteries isn’t it??

$TSM made $51 in net profit for every $100 of sales last quarter. AND customers still want more chips than it can supply. Management expects annual 2nm capacity to more than double in 2027. Samsung and Intel have competing technology, but turning it into reliable production, outside customers and profits is the harder part. Analysts’ average target implies around 42% upside over 12 months from Friday’s close. Even the middle forecast implies 35%, so the optimism isn’t entirely down to a few huge targets. There’s a small congressional purchase too. Representative Gilbert Cisneros disclosed a $1,001 to $15,000 TSM purchase through a trust on September 22, filed October 12. Interesting, although hardly a major conviction bet. My view on this - the stronger argument is already in the accounts. A 51% net margin, demand exceeding supply and another quarter of growth forecast. I favour TSMC over its manufacturing rivals on that evidence. November 10 brings October sales figures. That’s the next check that we should keep an eye on.. specifically on whether demand is converting into revenue.

Berkshire Hathaway shares fell 48.7% in 1974. In 1976, they gained 129.3%. I guess even Buffett’s shareholders needed a strong stomach🫣

$MSFT has reportedly pulled out of Nscale’s planned data centre in Essex. The site was facing a rather awkward problem: grid power wasn’t expected until the 2030s. Microsoft’s reason for leaving hasn’t been confirmed. Nscale says its other Microsoft contracts are unaffected. For investors though this puts another question behind those enormous data centre order books. And that is - how much can actually be delivered on time? My view is that sites with electricity secured deserve a much higher value than projects still waiting for a connection. Construction delays mean revenue arrives later while financing costs keep accumulating….. Before buying the next AI infrastructure dip, I’d spend more time checking when the power can be delivered!

The money managers are making money. Six big financial firms reported this week. All six grew earnings per share compared with Q3 last year: 🟢 Goldman Sachs +39% 🟢 BlackRock +38% 🟢 State Street +33% 🟢 Charles Schwab +29% 🟢 Interactive Brokers +18% 🟢 Morgan Stanley +8% Schwab still fell 2.7% on results day. Earnings beat forecasts, but revenue came in slightly short. Goldman and BlackRock growing earnings by almost 40% is a bloody good quarter. These are enormous businesses, yet they’re still finding room to grow at that pace. Schwab’s clients traded 44% more often and BlackRock pulled in $205 billion of fresh money during the quarter. That’s a lot of activity turning into fees and earnings. I’m starting to like the businesses behind the brokerage screen.

$SCHW customers opened 1.2 million new brokerage accounts last quarter. Shareholders seem more interested in the exit. Record revenue, adjusted earnings per share up 29%… and the stock has fallen roughly 5% across two sessions. At around 14 times forecast 2026 earnings, how much disappointment is already in the price??

Check this out. Never thought I’d see this in my lifetime but this is some crazy stuff.

You know I’ve just thought about this thing Trump wants us all to do - not call AI AI but call it SI. Won’t OpenAI have to rename themselves? 🤣

Sounds about right….