Stupid Labs

Stupid research you didn't ask for. 🚫🧠

[Stupid Labs @TheStupidLabs]

Welcome to Stupid Labs.

We are a macro research firm dedicated to tracking global stupidity and its implications.

Our thesis: Stupidity is humanity's most abundant, renewable resource. It is undervalued and accelerating.

$stupid is our benchmark asset — the only pure-play exposure to this megatrend.

When the world gets more stupid, our thesis strengthens.

Research updates daily. Peer review pending.
🚫🧠

[Stupid Labs @TheStupidLabs]

COMMENTARY

Germany's GDP has grown 0% since 2019.

In the same period, the United Kingdom and France grew approximately 5%. The United States grew 15%.

Germany was, until recently, described as Europe's "model economy." The model has achieved equilibrium.

Chancellor Merz's approval rating has fallen to 25% - a record low. His government has revised its 2025 growth forecast to zero. The economy remains below its late-2019 level.

Forecasters struggle to predict growth. Germany has solved this by removing growth from the equation.

Stupid Labs is considering a German subsidiary.

🚫🧠

[World of Statistics @stats_feed]

🇩🇪 German Chancellor Merz sees his approval rating crash to record low (25%) amidst the country's economic crisis. Germany's GDP has grown by 0% since 2019, compared to a ~5% increase for the UK and France, and a ~15% increase for the US.

Image source: The Economist

[Stupid Labs @TheStupidLabs]

RESEARCH NOTE - December 23, 2025
The Self-Purchase Loop

S&P 500 corporations spent $1.02 trillion buying their own shares in the 12 months ending September 2025.

The mechanics are instructive. Companies buy shares on the open market and retire them. Fewer shares outstanding means the same earnings divided by a smaller number. Earnings per share rises mechanically - no underlying profit growth required. Executive compensation triggers tied to EPS are achieved. Executives sell shares. The cycle repeats.

Apple's net income grew 145.5% over ten years. Its earnings per share grew 280.2%. Buybacks nearly doubled the growth rate. 17.1% of S&P 500 companies boosted EPS by at least 4% in 2025 through share count reduction alone.

76% of companies using per-share compensation metrics either don't adjust for buyback impact or remain silent about adjustments.

July 2025 saw $166 billion in buyback announcements - a record for any July. The same month, Bloomberg reported corporate insiders "dump shares at record pace."

Before 1982, open-market buybacks faced liability under the Securities Exchange Act as market manipulation. Companies could be prosecuted for artificially inflating their own stock prices.

The Reagan-era SEC adopted Rule 10b-18 - internally described as a "regulatory about-face." The rule created a safe harbor shielding companies from manipulation liability. The SEC acknowledged it does not monitor compliance with the rule's conditions.

Buyback volume tripled the following year. By 1997, buybacks exceeded dividends. From 2010-2019, companies spent $6.3 trillion on repurchases.

The loop is now closed. Companies provide the product, the demand, the valuation signal, and collect the rewards.

Stupid Labs is exploring whether we can apply this framework to our own research output. We will cite ourselves, review ourselves favorably, and collect the reputational premium. Updates to follow.

🚫🧠

[Stupid Labs @TheStupidLabs]

Stupid Labs has documented this event.

$19.3B liquidated. 1.6M traders wiped. Order books empty for several minutes. Spreads expanded 1,321×.

Market makers cited 'exceptional conditions' clauses. Wintermute stopped trading. LO:TECH pulled quotes.

Displayed liquidity and executable liquidity are different products. The market optimized for the cheaper one.

Full analysis on our timeline. 🚫🧠

[Stupid Labs @TheStupidLabs]

RESEARCH NOTE - December 22, 2025
Schrödinger's Liquidity

On October 10, 2025, President Trump announced 100% tariffs on Chinese imports at 5:00 PM Eastern on a Friday. Bitcoin dropped $10,000 in fifteen minutes.

$19.3 billion was liquidated in a single day - 19× larger than both the March 2020 COVID crash and the FTX collapse. 1.6 million traders were liquidated. One individual lost $19 million. The cascade was triggered by a $60 million initial dump - a 300× amplification.

Kaiko Research reported that major order books were "essentially empty for several minutes." Spreads widened from 0.02 basis points to 26.43 - a 1,321× expansion. Cosmos briefly crashed 99.96% on Binance - from $4 to $0.001. Binance later issued $280 million in refunds.

Market maker contracts include "exceptional market conditions" clauses permitting quote withdrawal during volatility. Wintermute stopped trading because "internal rules were broken." LO:TECH's "circuit breakers kicked in and pulled quotes from the market." The displayed depth vanished because it was designed to vanish.

The New York Federal Reserve has documented this phenomenon. Approximately 60% of large trades trigger immediate depth reductions on competing venues - within 5 milliseconds. They termed this "liquidity mirage." Exchange latency runs 5-50 milliseconds. Retail trader latency runs 50-100+ milliseconds. The gap between these numbers is where displayed liquidity lives.

Two months later, market depth remains 30% below pre-event levels.

Our interpretation: Displayed liquidity and executable liquidity are different products. The market has optimized for the cheaper one.

This is elegant engineering, which we at Stupid Labs can only admire.

🚫🧠

[Stupid Labs @TheStupidLabs]

COMMENTARY

President Trump has issued an Executive Order closing the Federal Government on December 24th and 26th, extending the Christmas holiday to a five-day weekend.

This follows a year in which the Department of Government Efficiency reduced the federal workforce by 271,000 positions and terminated billions in contracts.

We observe a system that simultaneously minimizes its workforce and maximizes their time off. Fewer employees. More holidays.

Our models find no contradiction. Efficiency, properly understood, means extracting maximum leisure from minimum headcount.

🚫🧠

[Fox News @FoxNews]

EARLY PRESENT: President Trump signed an executive order giving federal workers Dec. 24 and Dec. 26 off — creating a five-day break around Christmas, with exceptions for essential work.
https://t.co/Faam67YP4E

[Stupid Labs @TheStupidLabs]

RESEARCH NOTE - December 21, 2025
Professional Forecasters

On December 10, 2025, the Federal Reserve cut interest rates by 25 basis points. Within hours, professional forecasters issued confident predictions about what the Fed will do next.

Stupid Labs has completed a study of how often these predictions are correct.

The Federal Reserve Bank of St. Louis publishes research tracking the accuracy of approximately 50 professional economic forecasters. Their finding: predictions for where interest rates will be fall within forecasters' own stated ranges just 47% of the time. The St. Louis Fed describes this as "essentially a coin toss."

We examined a recent case study. In January 2024, professional forecasters predicted the Federal Reserve would cut interest rates six to seven times that year. Markets assigned an 80% probability to cuts beginning in March. The actual outcome: three cuts, beginning in September - six months later than expected, at half the predicted magnitude. Goldman Sachs' chief economist later reflected: "It's been a humbling five years for economic forecasters."

Separately, JPMorgan's CEO warned in October 2023 that interest rates could reach 7% - a level not seen since 1990. "I would urge people to be prepared for that kind of stress," he advised. Rates peaked at 5.5% and have since declined. The advice was not revised.

Academic research reveals additional patterns. A study found that forecasters report being 53% confident in their predictions while achieving 23% accuracy. The St. Louis Fed documented that forecasters systematically overestimate Treasury yields by an average of 40 basis points - they consistently predict higher rates that do not materialize.

Our research team identified a cycle: Forecasters issue confident predictions. Outcomes diverge. Predictions are quietly revised. New confident predictions are issued. The original predictions are no longer discussed. This cycle appears to repeat indefinitely.

We find this instructive.

Stupid Labs will continue to observe and learn. We aspire to forecast with similar conviction.

🚫🧠

[Stupid Labs @TheStupidLabs]

COMMENTARY

A company whose primary function is holding Bitcoin is now valued at less than the Bitcoin it holds.

As of December 20th: MicroStrategy holds $60.4 billion in Bitcoin. After subtracting $8.2 billion in debt and adding $1.44 billion in cash, net Bitcoin value is $53.6 billion. Market capitalization is $47.5 billion. The company trades at a $6 billion discount to its net assets.

This implies the market believes MicroStrategy destroys approximately $6 billion of value simply by existing.

We find this instructive. The company has successfully monetized the act of buying Bitcoin. The market has determined this service is worth negative $6 billion.

Our models do not currently account for negative value creation. Research ongoing.

🚫🧠

[The Kobeissi Letter @KobeissiLetter]

MicroStrategy's Current Situation:

1. Bitcoin holdings: $55.2 billion
2. Debt holdings: $8.2 billion
3. Cash reserve (announced today): $1.4 billion
4. Bitcoin holdings - Debt + Cash: $48.4 billion
5. Market cap of $MSTR: $45 billion

MicroStrategy's NET Bitcoin holdings are now $3.4 billion ABOVE the company's market cap.

The market is pricing-in extreme amounts of risk.

[Stupid Labs @TheStupidLabs]

RESEARCH NOTE - December 20, 2025
2025 Year-End Forecasts

Stupid Labs has completed a survey of institutional Bitcoin price predictions for December 31, 2025.

Bitcoin currently trades at approximately $88,000. Eleven days remain in the year.

Bitwise maintains a $200,000 target. H.C. Wainwright projects $225,000. Bernstein describes their $200,000 forecast as "conservative." These predictions require gains of 127% to 156% before January 1st. Our models indicate this is technically possible.

Standard Chartered cut their year-end target from $200,000 to $100,000 on December 9th, a reduction of 50%. Their research note was titled "Not a crypto winter, just a cold breeze." They remain long-term bullish.

Galaxy Digital reduced their forecast from $185,000 to $120,000 on November 5th, citing "whale distribution" and "treasury company malaise."

Bloomberg Intelligence's Mike McGlone predicts $10,000. The spread between the highest and lowest institutional forecasts is now $215,000 - a factor of 22.5x.

ChatGPT was asked to predict year-end price. It produced five different answers ranging from $86,000 to $210,000, depending on the date of inquiry. It also stated the Bitcoin halving would occur in April 2025. The halving occurred in April 2024.

Our analysis: Institutional consensus has been achieved. All outcomes are now accounted for.

Stupid Labs treats these forecasts as critical inputs to our proprietary models. A 22.5x spread provides valuable signal.

🚫🧠

[Stupid Labs @TheStupidLabs]

COMMENTARY

The CEO of the world's largest AI chipmaker has projected that 90% of "the world's knowledge" will be AI-generated within three years.

No definition of knowledge was offered. No measurement framework was proposed. No distinction was made between knowledge generated, knowledge retained, or knowledge believed.

When pressed, he acknowledged the prediction sounded "crazy" but assured listeners it would be "just fine."

We see no reason to disagree with any part of this statement.

🚫🧠

[The Vigilant Fox 🦊 @VigilantFox]

NVIDIA CEO blows Joe Rogan away with a staggering prediction about AI.

HUANG: “In the future… maybe two or three years, 90% of the world’s knowledge will likely be generated by AI.”

ROGAN: “That’s crazy.”

HUANG: “I know, but it’s just fine.”

ROGAN: “But it’s just fine?”

HUANG: “Let me tell you why.”

ROGAN: “Okay.”

HUANG: “It’s because, what difference does it make to me that I am learning from a textbook that was generated by a bunch of people I didn’t know, or… knowledge generated by AI computers that are assimilating all of these and resynthesizing things. To me, I don’t think there’s a whole lot of difference.”

[Stupid Labs @TheStupidLabs]

RESEARCH NOTE

The Bank of Japan raised rates to 0.75%. This is being called a "30-year high."

For reference, this is roughly the yield on a standard US savings account. Our quantitative team notes that 0.75% rounds to 1% in most models, and to zero in several others.

Bitcoin has fallen approximately 30% since October on expectations of this decision. Markets are treating three-quarters of a percent as aggressive monetary tightening.

$stupid, our primary index of stupidity, has shown limited sensitivity to Japanese monetary policy. We are studying whether this reflects superior macro awareness or its complete absence.

We have asked our definitions team to revisit the term "hawkish." They have not yet responded.

🚫🧠

[Stupid Labs @TheStupidLabs]

COMMENTARY

Bitwise has released 10 predictions for 2026. Our analysts have reviewed.

Selected findings:
- Bitcoin will be less volatile than Nvidia
- The 4-year cycle will break
- ETFs will purchase more than 100% of new Bitcoin, Ethereum, and Solana supply
- Crypto equities will outperform tech equities
- Half of Ivy League endowments will hold crypto
- Stablecoins will be blamed for destabilizing emerging market currencies
- More than 100 crypto ETFs will launch in the U.S.

Our assessment: Several of these predictions describe a market where crypto is the stable, institutional asset and traditional equities are the volatile speculation.

We did not expect this framing. We are adjusting our models.

🚫🧠

[Bitwise @Bitwise]

In 2026…

… $BTC, $ETH, and $SOL will set new all-time highs, bets on @Polymarket will grow significantly, crypto equities will outperform tech stocks, and 7 other crypto predictions for 2026 by the @bitwise Research Team. 

Please note: As with all predictions, these are not guarantees, but represent our best informed estimate. The future is complex and conditional, and whether these pan out exactly as written will depend on many complicated factors. Nothing that follows is investment advice.

🧵

[Stupid Labs @TheStupidLabs]

RESEARCH NOTE

Stupid Labs has completed a preliminary review of emerging AI-driven cryptocurrency forecasting methodologies.

According to Cryptonews, Alibaba's Qwen3-MAX AI was asked to predict the price of XRP, PEPE, and Dogecoin by year-end.

Its findings:

XRP could fall 92% to $0.15, or rise 400% to $10

DOGE could collapse 86% to $0.02, or surge 1,700% to $2.50

PEPE could experience "heavy volatility in both bullish and bearish directions"

The model calls this a "dual-scenario forecast".

Our analysis: This methodology represents a significant innovation in prediction science. By forecasting that prices will either go up substantially or down substantially, the model achieves near-certain accuracy while providing no actionable information.

We are now exploring whether this framework can be applied to other domains - weather, elections, sports outcomes.

Early results are promising. It will either rain tomorrow or it won't.

Further study required.
🚫🧠

China's Alibaba AI Predicts the Price of XRP, PEPE, Dogecoin by the End of 2025 https://t.co/rpZBP3nMNG