$AXTI may not just be a meme retail play..

JP stocks levered to optics doing well today too

$6834.JT
$5801.JT

I have no strong views on the current Iran/SoH crisis. While Fintwit has now magically transformed into geopolitical/oil experts (ex-software/SaaS experts), I’m aware enough to recognise that this isn’t my domain.

I can see both sides of the argument:
- Bear case is Trump didn’t plan for this, and now can’t wiggle out of this as he’s set off irreversible dominoes that has been in the making for more than 3 decades.
- Bull case is both sides are incentivised to off-ramp after a sufficient display of power; Trump needs to save the midterms (time is running out), and Iranians want to buy time to complete their nuclear program.

Given the volatility and uncertainty, yes I agree that the responsible thing to do wrt managing money is to degross to a manageable quantum.

That said, sentiments are overwhelmingly negative right now. The uncertainty now sets up a future catalyst path, as they will one day resolve back into certainty again.

From my limited experience, bulls REALLY love to climb a good wall of worry. A wall of worry works because it keeps sidelined investors from piling in all at once, thus saving incremental bidders to drive the market higher over time as certainty gradually resolves.

NVDA GTC day! I'm focusing mainly on implications for CPO, NAND and CPU.

Built one of my largest short positions in $TSLA @ $396

- robotaxis story increasingly commoditised by Waymo and Uber partnerships
- robotics/Optimus story; if Tesla can’t move EVs at a premium, what makes us think that they can sell Optimus at a premium? And if premium doesn’t work, then I don’t think they stand a fighting chance against China in the mid/low-end markets
- moving into fabs? Not sure when that actually materialises given Musk’s timeline esp with the complexity of building leading-edge fabs
- most imptly, it’s trading near 200x fwd p/e. Yes the stock has never traded on fundamentals/valuation but instead more on vibes, and exposure to Musk. So how much of that Musk’s premium gets diluted when SpaceX IPOs at $1.25tn? $1.25tn.. that cashflow has to come from somewhere.. Not to mention how Musk treats Tesla shares like his personal piggy bank..
- also memory headwinds
- also Musk <> Trump’s close relationship makes this a good hedge for Trump losing midterms

Lumentum also released their target financial model, which surpasses even my own model estimates (I think also other buyside bogeys)

- this target does not include the additional $5b revenue potential from their new fab
- target model did not extend into CY28, which is ironically when scale-up CPO really ramps. so what could CY29/30 look like with a scale-up ramp, if CY27/28 looks like this without a scale-up ramp? the photonics wave could be just starting

$LITE

Bought $GLD @ $408 and $SLV @ $63. On the other side of a forced seller is a happy buyer.

Have also had good luck trading with Silver in the past, so if it would be so kind as to let me buy the bottom for the third time…

Optics names have been bid hard these 2 weeks

Small (embarrassing) story: a month ago, I was very confident on the optics narrative going into GTC and OFC. So I bought lots of LITE, TSEM, AAOI, AXTI and some Jap optics names. Blinded by greed, I became overly concentrated / levered.

You can see where this story is going. Then came the U.S. strike on Iran, and the SaaS short squeeze… etc etc. Around the same time, my broker also increased the initial margin due to the heightened volatility of the largest winning stocks of the year, many of which are my largest positions like SNDK LITE. As a result, my buying power decreased and not only could I not buy more of the optics names when they corrected, but I was also forced to delever many names near the bottom.

If you asked me what I learnt from that episode, I would say nothing; because I already knew not to be so concentrated/levered, and also believed a SaaS short squeeze was coming. But this is a classic example of having the right idea, but the wrong execution. If I was not a forced seller 3 weeks back, I would be far above my YTD ATH today with the optics bid.

I’m not too hard on myself, mistakes happen and I am learning from it. Thankful also that I did manage to build back my positions in optics before OFC, though not to the amounts that I had had — a conscious decision to not be overly concentrated again (and thus by extension, to underperform).

It’s a dilemma right, like there’s alpha in concentration (when you’re right), but if you’re heavily concentrated and right, then you inevitably become heavily levered to momo, especially if you’re not playing some niche turnaround events with long duration. Many view momentum as inherently risky/bad/volatile, but I think it’s not so clear cut.

So is the right approach to trim as the market realises your thesis and price rises? Or should you double down PRECISELY because your thesis is being proven right? What’s the right balance? Still figuring this out but I believe there is no right answer - only one that fits to your own personality, duration and risk mgmt.

Anyway, no moral of the story, but just thought to share that sometimes you can be very right and yet very wrong at the same time; the market (which is an extension of the collective personalities and moods of the participants) is a funny thing

Put on large shorts on $COIN @ $200 and $CRCL @ $125

For the same reasons as before, think the bounce up the past month gives us a beautiful short setup and also a good hedge for Trump’s midterms.

If this is crypto under the most pro-crypto (read: grifting) admin, can’t imagine it under the Democrats. Also fundamentally, I think these 2 biz are terrible, and I really don’t think there’s moats here.

For Circle, aside from the CLARITY Act, Tether may be IPOing and has finally engaged a big four auditor for their reserves for the very first time in history. I see stablecoins endgame as either 1) USDT wins most share eventually, or 2) BB banks issue their own stables.

$MARA also announced sale of BTC yesterday and stock went up +DD% on news. I don’t think this will go unnoticed amongst the treasury companies and miners..

So why are these 2 extremely cyclical & low moat businesses valued at > 45x / 85x fwd p/e and people still think these are cheap? Agentic commerce?? Retail distribution?? SoV?? Lol

There’s news that Borat Margaret Sagdiyev from Kazakhstan is thinking of joining too… 🤔

https://x.com/moneyfetishist/status/2039587329535304168?s=46

It’s gna be ok

Well so the first model trained at scale on Blackwell is finally out.. Mythos showing a massive step function gain against previous gens, just months apart. Now how will models trained on Vera Rubin, Feynman look like..?

Maybe the default expectation is that everyone will be eventually made redundant. Feels like there’s maybe only 2 years left to hoard as much hard assets as possible. Time to strap in, get down and dirty using AI, max out your rate limits, and don’t forget to treat your LLMs better