Threading on the Edge

I think everyone is really waiting for a Semis to Software rotation, but I’m not sure we easily get one…

Do I think software is cheap? Yes. The software names like $PLTR $APP $DDOG also proved great earnings.

Are the semiconductor names overextended? I mean, look at the charts. However, their earnings have also been INCREDIBLE.

It doesn’t feel YET like the market can have $SMH and $IGV rally together, I think the only reason $IGV is up is because $SMH is down.

But if they were to rally together…it would probably be net new capital coming into the broader market vs people selling their semis to buy software or selling software to buy semis.

Just one day, likely need much more data to see if a rotation is happening…

Are you buying software over semis here? Rotating? Adding on dips?

https://x.com/amitisinvesting/status/2052458934741401933?s=46&t=hr2fbvcHJGpvp_SbDstdzg

Codex 5.5 hack:

"Are you 100% confident in this strategy? If not, find all possible loopholes, suggest proper fixes and run this loop until you are factually 100% confident in the new startegy"

This works like charm. It makes Codex 5.5 high perform even better than codex 5.5 extra high.

Why? Codex 5.5 is the only model i noticed that is self aware. It never makes high claims unless the model verifies everything.

This doesn't work with Opus 4.7 cuz that's a very insecure model. You can paste this prompt over and over again, the model keeps saying "you're absolutely right,....."

But with codex, after 2-3 iterations you'll notice yourself it actually patched all loopholes and this genuinely sounds like a good strategy.

Try this out, thanks me later.

https://x.com/cjzafir/status/2052110266566107321?s=46&t=hr2fbvcHJGpvp_SbDstdzg

US Stock AI Super Cycle Rotation: Full Rhythm Breakdown for the Bros

https://x.com/dtdt666/status/2052307717306429870?s=46&t=hr2fbvcHJGpvp_SbDstdzg

$HYPE is programmed for 3figs

It has something most altcoins do not: a mega financial flywheel and probably why it outperforms every protocol.

It has undoubtedly created one of the best alignment amongst market participants, community and token.

Months ago, I talked about its financial verticalization thesis and I believe it will continue to be the hegemony for an onchain financial system.

- offers over 100+ crypto assets, equities, commodities, ETFs, memes
- HIP-3, builders stake $HYPE to launch own perp market leading us to see institutional adoption for the $HYPE token and tokenized US stocks, commodities, indices, pre-IPO, treasuries, etc.
- commodities found a new distribution dominance on chain like $XAUT with often 5-10x the volume of every other tokenized gold asset combined
- launch of HIP-4, prediction markets coming to @HyperliquidX, which drives more demand to $HYPE, stake 1M $HYPE to launch your own markets
- in other words, you can use HIP-4 to hedge your positions, which is essentially an options market.
- I am optimistic to see future HIP to include other financial derivatives products and more formally an options market.

This financial flywheel it created will never cease:
- Volume creates fees.
- Fees create buybacks.
- Buybacks create scarcity.
- Scarcity creates attention.
- Attention creates more liquidity.

I believe $HYPE goes higher because of fees, strong value accrual mechanisms to token with each HIP launch and that has proven strength to be the best alt that has held up against the market when $BTC was dumping.

It will be the safest long term hold for me to know that institutions are holding my bags.

Keep buying, keep staking $HYPE

(Image credits to @x256xx )

https://x.com/arndxt_xo/status/2052662579856851130?s=46&t=hr2fbvcHJGpvp_SbDstdzg

This is what Palantir is.

https://x.com/rahul_garg/status/2051919689333522583?s=46&t=hr2fbvcHJGpvp_SbDstdzg

Hyperliquid did an L1 for perps. Caster is doing it for prediction markets.

I believe that @CasterMarkets might actually be building one of the coolest prediction markets.

Been following the prediction market space closely for a while, and I want to break down how Caster's architecture is different vs others.

vs. @Polymarket

- Runs on Polygon.
- General-purpose chain with recent focus onto payment.
- When market becomes inefficient and it becomes a pure speed race, the one with the fastest infrastructure wins, market makers eat toxic fills, spreads widen to compensate.
- That cost gets passed to you.
- Problems revolve around, oracle disputes, insider trading, regulatory pressure, and morally ugly markets around war/politics.
- Also, resolution is bottlenecked through UMA's optimistic oracle. In theory, anyone can propose an outcome and anyone can dispute it.
- Two men allegedly using classified military information to profit from Polymarket bets around Israel-Iran military actions.

vs. @Kalshi

- Centralized, raised $1B valued at $22B
- Banned or legally challenged in a growing list of states.
- You're trusting one company to resolve your markets correctly, with no on-chain recourse if they don't.
- The recent Khamenei market controversy showed that even when traders believe an outcome is obvious, payout can still depend on platform-controlled wording, carveouts, and legal interpretation

vs. @opinionlabsxyz

- Runs on BNB Chain, inheriting all of its MEV and ordering problems.
- An AI model is their primary oracle for resolution.
- That's a black box with extra steps.
- Their volume spike was points-farming. Unsustainable by design.
- Resolution becomes less legible. You may get faster settlement, but less transparent settlement.

vs. @predictdotfun

- Yield on locked capital is a nice QoL feature.
- It solves idle capital, not adverse selection
- It does not tighten spreads, prevent toxic fills, improve resolution quality, or make market prices more truthful.
- For long-dated markets, yield is nice. For fast markets, it barely matters.
- Prediction markets are won by liquidity, resolution credibility, distribution, and market-maker confidence, not by turning dead collateral into a savings account.

vs. @42space

- 42space turns prediction markets into market/event launchpads.
- Instead of buying a simple Yes/No claim, users buy outcome tokens whose price moves along a bonding curve.
- Bonding curves for price discovery means early entrants win by default, not by information edge.
- That's a meme coin mechanic. Frontrunning becomes a problem.

What Caster built

- Custom Rust L1, 200ms blocks: prediction markets move fast when news breaks, so slow chains create bad prices and bad fills.
- Cancel Priority at consensus is important when you accidentally leave a bad price on the table, Caster gives you a better chance to pull it back before someone takes advantage, market makers need this protection, or they stop providing liquidity.
- Other chains cannot easily copy it unless their own core team changes the chain rules.
- Native CLOB settled on-chain in the same block trades match, this can create tighter spreads and better execution
- CLP built specifically around binary market failure modes, as generic AMMs can break badly in prediction markets.
- Conditional markets where child markets that only exist if the parent resolves YES. Trade full causal theses, not just isolated binaries

That's not the same race for Caster.

https://x.com/arndxt_xo/status/2052445898907771350

2026 Hyperliquid Q1 Report

Today, we're excited to release Hyperliquid's 2026 Q1 Report.

While Q1 stands as one of crypto’s most challenging quarters since 2018, Hyperliquid emerged stronger. As markets fell sharply, the protocol continued to gain traction, with HYPE returning +44.2% and entering the top 10.

More importantly, the quarter brought Hyperliquid closer to the “House of All Finance” vision, from institutional recognition through S&P Dow Jones to emerging as a price discovery venue for oil and other assets.

Enjoy the read. Some excerpts below, with the link to the full PDF.

Hyperliquid.

https://x.com/hyperliquidr/status/2052371846750724319?s=46&t=hr2fbvcHJGpvp_SbDstdzg

We used to have KOLs shilling shitcoins.

Now we have KOLs shilling stocks.

Nothing changed, it’s still the same old.

KOLs using followers to pump their bags and as exit liquidity

https://x.com/kirbyongeo/status/2052271753356062747?s=46&t=hr2fbvcHJGpvp_SbDstdzg

Not every gap is worth attention. But when the fundamentals are visible, they rarely stay ignored.

$CARDS is doing ~$87M in monthly gross volume, translating into ~$3–4M in protocol revenue (~$40M annualized) at a ~$42M market cap.

@Collector_Crypt is tokenizing a multi-billion dollar trading card market on on-chain rails.

The business is operating at strong margins, putting it at ~1.0x P/E (Price-to-Earnings), vs ~$HYPE at significantly higher multiples. The gap is not marginal.

Collector Crypt shows clear product–market fit, driven by a simple loop:

> Buy pack ($25 → $1000)
> Receive NFT (real graded card)
> Redeem, trade, or instant sell (85–93%)

The driver is unit economics.

Inventory is sourced ~15% below market, allowing ~4–5% platform margins while returning 5–10% positive EV to users. Value is left with the user at the point of interaction, which shows up in repeat spend and sustained demand.

This is not competing with DeFi, but with eBay and offline card markets, compressing 10–30% fees into ~2% while improving liquidity.

Buybacks have started, with further token-linked flows expected. Combined with category expansion and fiat onboarding, the model is moving toward direct value accrual.

The model is already working, pulling a multi-billion dollar offline market online and on-chain with clear unit economics.

As profits begin routing back into systematic buybacks, the question shifts from whether it works to how long this multiple can persist.

https://x.com/thelearningpill/status/2052336153043616032?s=46&t=hr2fbvcHJGpvp_SbDstdzg

every leading chain now has its own version of a neobank:

https://x.com/Octop3s/status/2051657141212110869

list of tcgs

check it out here: https://tcg.eli5defi.xyz

https://x.com/Eli5defi/status/2051953974883041310

A new AI company called @subquadratic just released a model that breaks one of the oldest limits in modern AI.

To understand why it matters, you need to understand a single math problem that's been silently shaping every chatbot you've ever used.

https://x.com/Eli5defi/status/2052026378187649382

been playing @Showdown_TCG on @megaeth


interesting twist on the standard poker gameplay. turns it into a strategy battle rather than a capital one

quite fun and recommended, especially if you like poker
https://x.com/insomniac988/status/2052061921445093767

Most people will read this as “Pendle becoming the yield hub.”

But it is MORE than that.

That framing misses the more important shift around who controls the pricing layer of capital.

> Tokenization already gave access
> Yield already existed across T-bills, private credit, and dividend flows

But without a consistent way to price and allocate it, this does not scale beyond niche usage.

Tokenized RWAs are at ~$30B on-chain, with $350B+ in represented asset value, but capital remains selective because these yields are not directly comparable.

@pendle_fi is standardizing that missing layer.

Apollo credit, @saylor's $STRC dividends, regulated T-bill yield via $USDG, and @ethena’s multi-source backing all originate from different markets. They converge not because of yield, but because they need a common venue where it can be priced, split, and underwritten.

Private credit is a $3.5T market, but still runs on slow price discovery and constrained liquidity.

Pendle works because it compresses those differences into a rate the market can act on. It starts to resemble a pricing layer forming underneath the rest of DeFi.

$PENDLE is becoming the layer that others price themselves against.

https://x.com/thelearningpill/status/2051610461679014068

at 3b fdv, this is the next $RAVE $RIVER $SIREN

https://x.com/arndxt_xo/status/2052051343452590193

$AUROS: The Last Unpriced Moat in the HBM4 Supply Chain

Auros is a small Korean company that makes inspection tools for advanced memory chips.

The simple thesis is: AI chips need faster memory, so Samsung and SK Hynix are moving to a new HBM4 manufacturing method called hybrid bonding. This new method is harder to inspect, and the old market leader, KLA, may not be as strong here.

Auros has spent years building tools for exactly this problem, and because it is Korean, it may have a local advantage with Samsung and SK Hynix.

If its tools get approved and used in real production, Auros could become deeply embedded in the HBM supply chain, making it hard to replace. The market is still valuing Auros like a risky small company, so if the HBM4 qualification works, the stock could potentially re-rate much higher.

Read the full research deep dive here: https://threadingontheedge.substack.com/p/auros-the-last-unpriced-moat-in-the

here my thesis on memory/storage as a subset of the ai supercycle

storage has had supercycles before. 2010. 2014. 2017.

the core trading logic is simple: AI devour compute power + voraciously consuming storage, with even 2026 capacity nearly fully pre-booked.

the 2026 $SNDK $MU $WDC $STX storage trade still has significant upside because:

- the storage supercycle lasts 3-5 years on current demand trajectory
GF Securities has a $1,462 target that was set conservatively before this week's volume data
- $SNDK $2,000 by year end implies ~40% upside on an asset with locked-in demand
- the late-cycle rotation from compute into storage has barely started institutionally

in crypto we will also see the same play, the crypto storage play $FIL $AR $STORJ is a different bet entirely. thesis is simple the move in equities will be reflected in that of the crypto markets

every storage supercycle was demand-driven by consumer devices: smartphones, PCs, streaming

the cycle would run beucase of AI, supply would catch up within 12-18 months, margins would compress, the trade would end

this cycle is different in 3 ways:

1. the demand source has a different cost tolerance
when a consumer buys a smartphone with 256GB, they're price-sensitive. they'll wait a cycle if prices spike.
when AWS pre-books 2026 storage capacity for AI infrastructure, they don't care if NAND (a type of memory chip used for storage) prices go up 40%. the margin on selling compute to enterprises covers it easily. they cannot afford to not have the storage.
demand is price-inelastic in a way consumer storage never was.

2. the product mix is shifting in favor of manufacturers
AI workloads specifically require high-density, high-endurance flash with specific write patterns. it's enterprise SSD with meaningfully higher average selling price and margins.
SanDisk, Micron, and Western Digital all have better product mix today than any prior cycle. they're selling $300 enterprise SSDs.

3. supply cannot catch up in 12 months anymore
advanced NAND fab construction timelines are now 3-5 years, not 18 months. equipment lead times, cleanroom build time, and process qualification have all extended.
if you pre-book 2026 capacity now, you're acknowledging that you cannot build or buy your way to more supply on a short timeline.
this cycle does not end the way prior cycles ended.

also, oil at $200 also accelerates the energy efficiency mandate for data centers which directly benefits high-density, low-power enterprise NAND over spinning disk

capital is rotating out of @nvidia. what’s interesting is that $NVDA has been down for several days, while others like
$SNDK $MU $WDC $STX are still pushing higher.


https://x.com/arndxt_xo/status/2051907255373910066

7 tactics to 100x the impact of your announcement

https://x.com/intern/status/2051731138197045377

How to design beautiful software:

1. use https://getdesign.md to get a popular site design system

2. use https://app.superdesign.dev it copy components you like

3. have taste

https://x.com/mann_idan/status/2051121074075767105