Michael hochstat (@HochstatMichael) profile picture

data maker

Florida, USA
Joined January 2016
Photo by Michael hochstat (@HochstatMichael)Photo by Michael hochstat (@HochstatMichael)Photo by Michael hochstat (@HochstatMichael)Photo by Michael hochstat (@HochstatMichael)Photo by Michael hochstat (@HochstatMichael)Photo by Michael hochstat (@HochstatMichael)Photo by Michael hochstat (@HochstatMichael)Photo by Michael hochstat (@HochstatMichael)Photo by Michael hochstat (@HochstatMichael)Photo by Michael hochstat (@HochstatMichael)
Breaking Nvda update
55
Time to lower your data costs
1
93
Did Wall Street sell the market today to keep Anthropic above its IPO price? The distinction is important. This wasn’t necessarily about raising money to fund initial IPO allocations. It may have been about finding additional capital to absorb selling pressure in Anthropic’s aftermarket. Imagine institutions liquidating existing equity positions to prevent the largest IPO in history from breaking its offering price. The broader market absorbs the damage while the IPO holds. All the experts, all the research, all the data and all the AI models — and ultimately, it comes down to available liquidity. It is a plausible hypothesis, but confirming it would require institutional trading and order-flow data. Did Wall Street sacrifice the broader equity market to protect the biggest AI IPO in history?
1
414
Wall Street: From the O.K. Corral to Star Wars Take all the experts, all the data, all the research, all the alternative data and all the AI models. Put them in one place. Everyone is looking at similar information, chasing similar trades and heading for the same exit. We thought we were building an information advantage. Instead, we may have built the ultimate crowded trade. Welcome to the O.K. Corral — except the gunslingers have been replaced by algorithms. The drones have won. It’s Star Wars on Wall Street. Autonomous agents, quantitative models and automated trading systems competing against one another at machine speed. But even the most sophisticated machines cannot repeal the laws of supply and demand. More data does not create more buyers. More tokens do not create more capital. More intelligence does not create more liquidity. We have 21st-century technology fighting over an economic problem that is centuries old. The machines may have won the technological war, but who is left to buy the stock?
1
181
Sometimes, it’s just supply and demand. The equity market underestimated one of the most fundamental principles of economics. Too much stock supply chasing a finite pool of capital. IPOs, secondary offerings, insider selling, stock-based compensation and private equity exits all compete for investor dollars. Meanwhile, valuations assumed an almost unlimited appetite for equities. Companies can report record earnings and still see their shares collapse. Good fundamentals do not guarantee sufficient buying demand at elevated valuations. AI models, alternative data, sophisticated algorithms and billions of tokens cannot manufacture liquidity. When sellers overwhelm buyers, prices must adjust. Wall Street spent too much time calculating future growth and not enough time calculating who would buy all the stock.
96
The Crypto Treasury Race to the Bottom Reported unrealized losses: Michael Saylor’s Strategy ($MSTR): $14.1 billion on Bitcoin. Tom Lee’s BitMine ($BMNR): $10.1 billion on Ethereum. Combined: $24.2 billion. The figures still need independent verification, but the underlying financial risks deserve attention. What happens when the capital markets stop rewarding companies for accumulating cryptocurrency? The financial engineering that amplified the upside can amplify the downside: Collapsing premiums to net asset value. Increasingly dilutive equity issuance. Higher financing costs. Reduced access to fresh capital. Potential liquidity pressure. Unrealized losses do not automatically mean insolvency or forced liquidation. Debt obligations, cash reserves and financing structures will determine the real damage. Buying an appreciating asset with other people’s capital is not necessarily a sustainable business model. The real test begins when the asset depreciates and Wall Street stops writing checks. Financial engineering is not operating cash flow.
129
Michael hochstat retweeted
In Nazaré, Portugal, surfer Sebastian Steudtner rode a 26.2-meter wave, setting a new world record for big wave surfing.
149
1,205
148
9,232
714,949
Galaxy Digital: Is the end near—or is this a severe valuation reset? Galaxy Digital’s stock has fallen roughly 51% since late September, significantly more than Bitcoin’s 32% decline. My diagnosis: this is not an existential crisis. The selloff reflects four overlapping forces: 1. Crypto-market weakness
Bitcoin’s decline and reduced risk appetite have pressured crypto-linked equities. Galaxy trades like a leveraged crypto proxy, similar to Coinbase and Strategy. 2. Helios II uncertainty
Investors are no longer willing to fully value Galaxy’s future AI/data-center capacity before tenants, contract economics, power approvals and completion timelines are secured. 3. Leverage and capital intensity
Galaxy issued $3.5 billion of secured notes to finance Helios expansion. The project is funded, but higher debt and construction spending make the equity more sensitive to delays, rates and financing conditions. 4. Broader macro pressure
Higher discount rates and weakness across technology and risk assets have hurt both sides of Galaxy’s story: digital assets and long-duration data-center infrastructure. Importantly, the decline does not appear to reflect collapsing operations. Galaxy reported: •$323 million of Q3 net income •$455 million of adjusted EBITDA •$4.9 billion of cash, stablecoins and U.S. Treasuries •$69 million of Data Centers adjusted EBITDA The central question is no longer whether Helios has strategic value. It is whether Galaxy can convert that value into contracted, financed and deliverable cash flows without excessive dilution or execution delays. What could change sentiment? ✅ A major Helios II tenant agreement
✅ Clear ERCOT power and energization timelines
✅ Construction remaining on schedule and budget
✅ Stabilization in Bitcoin and crypto activity What would confirm the bear case? ⚠️ Further leasing delays
⚠️ Reduced power allocations
⚠️ Expensive new financing or equity issuance
⚠️ Construction overruns
⚠️ Persistent losses in the core digital-assets business Bottom line: Galaxy is facing a credibility and valuation test—not yet a survival test. The next major move will likely depend on Helios contracting progress and the direction of the broader crypto market. #GalaxyDigital #GLXY #Bitcoin #Crypto #DataCenters #AIInfrastructure #Investing
101
In the old days on Wall Street, during a market collapse, we used to say, “Don’t walk too close to the buildings.” But Wall Street has moved to the cloud. The buildings are still standing. The capital is disappearing into virtual infrastructure, inflated valuations, overleveraged balance sheets and promises of future profitability. Trillions in market value. Billions spent on data. Endless metrics. AI agents running 24/7. And when the market turns, none of it guarantees protection. The cloud is starting to look like nothing more than steam blowing in the wind. Technology changes. Financial gravity doesn’t.
2
517
Snowflake is melting. What happens to the private AI data companies? Snowflake has access to public equity and debt markets. Private companies depend on venture capital, strategic investors, acquisitions and eventual IPOs. As public valuations compress, private financing becomes increasingly selective. Consider the private financial-data and AI landscape: • Databricks — $190B reported valuation • AlphaSense — $7.5B reported valuation • Hebbia — $700M reported valuation • Rogo • Model ML • Daloopa • Brightwave • LinqAlpha • Finster AI • Fintool • Boosted.ai • 9fin These companies have different financial positions, but they face many of the same industry pressures: • Rising AI infrastructure costs • Expensive data licensing • Intensifying competition • Customer acquisition costs • Pricing pressure • Greater scrutiny of profitability • Uncertain exit opportunities The next chapter could involve down rounds, layoffs, delayed IPOs, consolidation and distressed acquisitions. Forget the promotional metrics: Tokens consumed. Queries executed. Agents deployed. Documents processed. Investors will increasingly want five numbers: Cash remaining. Monthly cash burn. Gross margins after AI and data costs. Paying customer retention. Months until another financing is required. Most private vendors do not publicly disclose enough information to evaluate these metrics. Recently funded companies have breathing room. Others may struggle to raise capital at their previous valuations. Capital markets are not completely closed, but the price of capital and the requirements for obtaining it are changing. Snowflake gets repriced every day. Private companies can hide behind their last funding valuations until they need more money. The next AI data industry shakeout may be driven by financial survival rather than technological superiority. #AI #PrivateEquity #VentureCapital #FinancialData #CapitalMarkets
93
Snowflake ($SNOW) is melting — and the bigger question is whether its business economics justify its valuation. The fundamental issues worth examining go well beyond the stock price: Debt: $3.75 billion in convertible notes issued in 2024, with potential shareholder dilution. Stock-based compensation: The difference between reported free cash flow and genuine shareholder economics. AI monetization: Is AI generating incremental profitable revenue or simply increasing infrastructure costs? Competition: Databricks, Microsoft, Google and Amazon. Consumption-based revenue: Customers can optimize their workloads and reduce spending without terminating contracts. Snowflake | November 20, 2026 $261.78 −2.52% in premarket trading Latest verified premarket quote, 9:14 a.m. ET 52-week high$384.56 Decline from high −31.9% Trailing GAAP EPS−$3.18 FY2027 product revenue guidance$6.07B Yahoo Finance +1 Wall Street is starting to cut targets Analyst Old target New target Bank of America $470 $400 UBS $500 $450 Oppenheimer $475 $475 Bank of America reduced its target today, citing growth expectations. NYSE: SNOW Live - Investing.com +1 The fundamental problem Snowflake’s latest product revenue grew 37% to $1.49 billion. The business is not collapsing. Its valuation is compressing as investors reassess AI profitability. More AI workloads do not necessarily translate into better margins. More consumption does not necessarily translate into better shareholder returns. Meanwhile, today’s broader technology selloff is adding considerable pressure. Reuters +1 December 2 earnings: What matters I would concentrate on operating cash flow, stock-based compensation, convertible debt, purchase commitments, cloud infrastructure costs and incremental AI margins. I would also reconcile the outstanding convertible debt against any subsequent financing before drawing a balance-sheet conclusion. The original $3.75 billion issuance is not necessarily the current outstanding balance. Snowflake doesn’t have to stop growing for shareholders to lose money. It simply has to grow less profitably than its valuation demands.
157