Anthropic’s $900B Compute Problem with Jeff Richards + Why It Feels Like 2007 for Retail with CME Group’s CEO Terry Duffy
Anthropic’s $900B Compute Problem with Jeff Richards + Why It Feels Like 2007 for Retail with CME Group’s CEO Terry Duffy
Podcast1 hr 16 min
Listen to Episode
Note: AI-generated summary based on third-party content. Not financial advice. Read more.
Quick Insights
  • Consider CrowdStrike (CRWD) and Palo Alto Networks (PANW) for cybersecurity exposure, which the discussion viewed as a clearer AI-era beneficiary than traditional enterprise software; no price targets were provided.
  • Snowflake (SNOW) may benefit if AI adoption drives greater demand for data infrastructure, but look for confirmation in customer demand and results rather than relying on the theme alone.
  • Treat Anthropic’s IPO as a high-growth but volatile opportunity: its first-day swings are not a reliable signal, and a clearer valuation may emerge over the next three to six months.
  • Be cautious with Treasury bonds: the guest expected yields to rise, which could pressure bond prices, though no yield target or timeframe was given.
Detailed Analysis

Anthropic (ANT, as identified in the transcript)

  • Anthropic’s enterprise AI business is growing rapidly: its revenue run rate was cited as rising from $16 billion a year ago to an expected $60 billion.
  • The company disclosed 11,000 customers spending at least $100,000 a year, while those customers represent less than 1% of their software and technology spending.
  • The guest argued that adoption is still early, with many large companies only piloting AI tools. He sees potential for enterprise adoption to support growth into 2027 and 2028.
  • The transcript cites roughly $900 billion of compute commitments over five years. The guest pointed to strong margins, cash flow, and a large cash balance as counterpoints to concerns about the spending.
  • The IPO was priced at $83. It opened above $100, briefly traded around $115, and was near $97 late in the first trading day. The guest cautioned that first-day trading is not a reliable measure of long-term investor demand and said a more meaningful valuation may emerge over three to six months.
  • The guest described the IPO as priced at about 10 times forward revenue, compared with public software and technology companies he said were trading at higher multiples. The host also cited a valuation around $1.25 trillion.

Takeaways

  • The discussion presents Anthropic as a high-growth enterprise AI opportunity, but the compute commitments make future revenue growth and customer adoption especially important to monitor.
  • Treat the initial trading swings as a sign of IPO volatility, not as evidence by themselves that the long-term investment case is succeeding or failing.

NVIDIA (NVDA)

  • NVIDIA was described as a major source of recent volatility in technology stocks and broader AI sentiment.
  • The guest said NVIDIA traded at roughly 10–12 times forward earnings, while Walmart was cited at about 40 times. He argued that the valuation gap reflects investor uncertainty about NVIDIA’s outlook, compared with greater confidence in Walmart.
  • The guest said a decline in AI-related stocks over recent months had left the market questioning whether expected AI demand will support the companies’ forecasts.
  • He argued that if NVIDIA and AI companies such as Anthropic meet their expectations, the market could reward the broader AI sector.

Takeaways

  • The discussion points to a gap between NVIDIA’s valuation and the uncertainty surrounding future AI demand. Investors weighing the stock should watch whether AI spending translates into sustained business results, rather than relying on the scale of spending alone.
  • The episode offered no price target for NVIDIA.

Enterprise software and AI infrastructure

  • The guest described the software sector as undergoing a major transition: investors are asking whether businesses will continue using traditional enterprise software or shift toward AI prompts and agents.
  • He said stock selection may matter more in this environment because some software companies will adapt successfully and others may not.
  • He sees data and infrastructure as potential beneficiaries of AI agents, which could generate much greater demand for computing and infrastructure.
  • Private companies mentioned as having experienced faster growth included Databricks, Vercel, Supabase, Fireworks, and Base10. The guest also named Cursor and Cognition as examples of companies serving AI coding demand.
  • The guest said public investors had initially lacked visibility into AI adoption on the customer side, but recent company results had provided more evidence of demand.

Takeaways

  • The discussion favors looking for companies with clear AI-related customer demand, especially in data, infrastructure, and security, rather than assuming all software companies will benefit equally.
  • Most of the private companies mentioned are not publicly traded, so the transcript does not identify a direct public-market investment in them.

Salesforce (CRM)

  • Salesforce fell roughly 10% in one session and was described as down around 20% over the week and 30% from post-earnings highs.
  • The guest noted Salesforce’s Cloudforce initiative and its partnership with Anthropic, but said software stocks remain vulnerable to questions about whether AI prompts could replace or reshape traditional enterprise applications.
  • The guest also pointed to algorithmic trading and momentum as possible influences on short-term price moves.

Takeaways

  • Salesforce’s AI partnerships are a potential positive, but the episode frames the larger question as whether the company can maintain its role in enterprise software as AI tools change how employees use applications.
  • The discussion did not provide a valuation or price target.

CrowdStrike (CRWD) and Palo Alto Networks (PANW)

  • The guest said investors were gravitating toward cybersecurity companies such as CrowdStrike and Palo Alto Networks, which he described as having clearer momentum in the current AI-driven environment.
  • He contrasted this with uncertainty around traditional enterprise software and applications.

Takeaways

  • The discussion suggests cybersecurity may be viewed as a more direct AI-era beneficiary than some traditional software categories.
  • The guest did not give specific recommendations or price targets for either company.

Snowflake (SNOW) and Databricks

  • The guest said he believes data companies such as Snowflake and private company Databricks can benefit from the broader AI trend.
  • He cited infrastructure and data demand as areas where AI adoption could create additional growth.

Takeaways

  • The episode’s positive view rests on AI creating more demand for data infrastructure. Investors may want to look for evidence of that demand in company results rather than assuming the theme benefits every data or software company equally.
  • Databricks is private, while Snowflake is publicly traded.

Walmart (WMT) and SCHD

  • The guest said SCHD, a dividend-focused ETF, was up about 34% for the year, compared with roughly 7% for QQQ at the time of recording.
  • He cited Walmart’s roughly 40-times forward earnings valuation as an example of investors paying a premium for a business they view as more predictable.
  • Walmart was also described as having rallied about 25% from recent lows.

Takeaways

  • The comparison highlights a market preference for perceived stability and lower-risk exposure during a period when technology and AI stocks were under pressure.
  • The episode does not recommend SCHD or Walmart, and it does not discuss a price target for either.

CME Group (CME)

  • The CEO described CME as benefiting from strong trading activity and improvements in market efficiency. He cited 15% year-over-year growth in the company’s third-quarter business and said its portfolio-margining and offset capabilities free up about $105 billion each day for market participants.
  • He emphasized CME’s clearing, risk controls, and market liquidity as advantages over newer or less-established trading venues.
  • The discussion also covered CME’s planned compute futures product with Silicon Data. The CEO said the proposal had been delayed after the Commerce Department raised a national-security concern.
  • The CEO criticized perpetual contracts offered elsewhere, arguing that some are swaps rather than futures and may carry regulatory and tax-treatment issues for retail users.

Takeaways

  • The discussion presents CME’s established infrastructure, liquidity, and risk-management systems as competitive strengths.
  • New products such as compute futures could create opportunities, but the transcript identifies regulatory approval as a key uncertainty.
  • CME’s stock was described as trading at all-time highs, but no price target or specific buy recommendation was given.

Robinhood (HOOD), Coinbase (COIN), and Bitcoin (BTC)

  • The CME CEO said retail-focused brokerage and crypto-related stocks had fallen sharply. He cited Robinhood’s decline from about $117 to around $60–$70, and said Coinbase had reached a multi-year low.
  • Bitcoin was cited at about $63,000–$64,000, down from roughly $82,000 a few weeks earlier.
  • The CEO linked Coinbase’s share price to Bitcoin’s performance and questioned whether crypto products, including decentralized finance, were gaining meaningful institutional adoption.
  • He said institutions may be reluctant to use DeFi because of concerns about client identification, transparency of positions, and the possibility of being front-run.

Takeaways

  • The discussion is cautious on crypto-linked equities and DeFi, emphasizing price volatility and uncertainty about institutional adoption.
  • For Bitcoin and related stocks, the episode suggests that investor confidence in practical use cases and adoption remains an important issue to watch.

Kalshi, prediction markets, and perpetual contracts

  • The CEO said prediction-market activity at Robinhood was largely tied to sports and argued that sports-focused prediction products face regulatory scrutiny and potential manipulation risks.
  • He said Kalshi was reportedly raising funds at a $64 billion valuation, which he attributed to a focus on compute futures rather than just prediction markets.
  • The CEO said CME had filed applications and indexes for compute futures but had not received approval.
  • He warned that retail traders using perpetual contracts may face risks involving the contracts’ legal classification, tax treatment, and the potential for automatic liquidation.

Takeaways

  • The episode views prediction markets and perpetual contracts as areas where product design, regulation, and investor protections remain significant uncertainties.
  • The discussion does not present these products as a straightforward investment opportunity for retail investors; it stresses that users should understand the contract structure and risks.

U.S. Treasury bonds and interest rates

  • The CEO said the Treasury market had become less liquid than in the past, partly because foreign participation had declined and hedge funds and proprietary trading firms play a larger role.
  • He attributed sharp, fast moves in Treasury yields partly to algorithmic trading and rapid reactions to news.
  • He said Treasury yields looked too low and expected yields to rise, which implies bond prices could fall.

Takeaways

  • The discussion is bearish on Treasury bond prices in the near term, based on the CEO’s view that yields were low and market conditions were making bonds vulnerable to further moves.
  • The episode cited no specific yield target or timeline.

Bank of America (BAC) and JPMorgan Chase (JPM)

  • The conversation raised concerns about duration risk at banks that bought longer-term securities when interest rates were very low.
  • Bank of America was discussed as having accumulated bonds during the low-rate period, while JPMorgan was described as having avoided some of that exposure.
  • The CEO said he believed banks were generally in better shape after the lessons of past crises, while acknowledging that mismatched asset and liability durations can create stress when rates rise.

Takeaways

  • The discussion highlights interest-rate and duration risk as important factors when evaluating banks.
  • It did not offer a specific recommendation on Bank of America or JPMorgan.

Private credit and data-center financing

  • The guest raised concerns about the amount of capital being used to build AI compute infrastructure and the possibility that capacity could exceed demand.
  • The discussion noted that hyperscalers have made large capital commitments and that private-credit firms and banks are involved in financing parts of the buildout.
  • Amazon was discussed as building data-center capacity and selling data centers to an investment firm, which the host presented as part of the broader financing structure around AI infrastructure.
  • The CEO warned that a pullback in AI commitments could have spillover effects, while the guest said the market still needed more evidence that AI demand would support the spending.

Takeaways

  • AI infrastructure is a large investment theme, but the episode emphasizes overbuilding, financing dependence, and the risk of spending outpacing revenue.
  • Investors considering companies or funds exposed to data centers and private credit should pay attention to who is financing the buildout and whether customers are committing to use the capacity.

AI in healthcare

  • The CME CEO said he believes one of AI’s most valuable applications could be medicine, where it may improve health outcomes and, potentially, reduce insurance costs.
  • He argued that healthcare could provide a concrete way to assess AI’s benefits to society.

Takeaways

  • Healthcare is presented as a promising long-term AI application, but the transcript does not name specific companies, funds, or investment recommendations in the sector.
Ask about this postAnswers are grounded in this post's content.
Episode Description
This episode is sponsored by Fidelity Investments and the all-new Fidelity Trader+ platform. Try Fidelity’s most powerful trading experience yet: https://Fidelity.com/TraderPlus Fidelity Investments and Risk Reversal are not affiliated. Views, opinions, products, services, and strategies discussed are not endorsed or promoted by Fidelity Investments. Fidelity Brokerage Services LLC, Member NYSE, SIPC. Dan Nathan is joined by Jeff Richards, Managing Partner at Notable Capital, on the day Anthropic began trading publicly. Jeff invested in Anthropic’s Series E at a $60 billion valuation. He walks through his original thesis, what the S-1 showed about margins, cash flow and enterprise adoption, and why he thinks the deal was priced reasonably at about 10x forward revenue. The two also take on the bear case around Anthropic’s $900 billion in compute commitments, what first-day trading does and doesn’t tell you, the hit to AI sentiment from the Nvidia news, and the sharp selloff in SaaS names like Salesforce. After the break, from the CME Group Tour Championship Pro-Am in Naples, Florida, Dan and Guy Adami sit down with Terry Duffy, CME Group Chairman and CEO. Terry explains why he believes perpetual contracts are swaps rather than futures, warns that “it’s 2007 for retail,” and takes on prediction markets, Kalshi’s valuation and the stalled launch of compute futures. He also covers liquidity and volatility in the Treasury market, the systemic risks of AI infrastructure financing, why institutions are wary of DeFi, the deal with Google Cloud and his views on today’s tech CEO class. Learn more about St. Jude HERE —FOLLOW USYouTube: @RiskReversalMediaInstagram: @riskreversalmediaTwitter: @RiskReversalLinkedIn: RiskReversal Media The financial opinions expressed in Risk Reversal content are for information purposes only. The opinions expressed by the hosts and participants are not an attempt to influence specific trading behavior, investments, or strategies. Past performance does not necessarily predict future outcomes. No specific results or profits are assured when relying on Risk Reversal. Before making any investment or trade, evaluate its suitability for your circumstances and consider consulting your own financial or investment advisor. The financial products discussed in Risk Reversal carry a high level of risk and may not be appropriate for many investors. If you have uncertainties, it's advisable to seek professional advice. Remember that trading involves a risk to your capital, so only invest money that you can afford to lose. Derivatives are not suitable for all investors and involve the risk of losing more than the amount originally deposited and any profit you might have made. This communication is not a recommendation or offer to buy, sell or retain any specific investment or service.
About RiskReversal Pod
RiskReversal Pod

RiskReversal Pod

By RiskReversal Media

Welcome to the RiskReversal Pod, where Dan Nathan and Guy Adami are joined by the most brilliant minds in markets and tech.  We break down the most important market moving headlines to help listeners make better informed investing decisions. Our goal is to deconstruct Wall Street speak and offer contrarian insights and strategies that help investors navigate increasingly volatile markets. Tune into the RiskReversal Pod Monday through Friday for succinct 30 minute pod drops of market analysis that you won't find anywhere else. For new episodes of On The Tape with Danny Moses, search "On The Tape" in your favorite podcast platform. — FOLLOW US YouTube: @RiskReversalMedia Instagram: @riskreversalmedia Twitter: @RiskReversal LinkedIn: RiskReversal Media