Martin Shkreli on building a Bloomberg killer and the future of Wall Street data

Summary

Martin Shkreli is building a Bloomberg competitor, and his structural case for why it works stands up better than the pitch sounds.

Bloomberg was built for fundamental long/short investors. When Mike Bloomberg became mayor in 2002, quant funds were taking over Wall Street. Today, 85–90% of the top 15–20 hedge funds are quants, yet Bloomberg offers no meaningful quant toolset. Firms like Jane Street and Two Sigma built their own infrastructure from scratch. Only about 5% of Jane Street and Two Sigma staff use Bloomberg, treating it as an entry-level tool.

Shkreli's opportunity moves clients up the power curve. Fundamental funds could operate more like quants. Quants could avoid building infrastructure themselves. He frames it as an AWS model where the same tools available to Citadel become accessible to smaller funds at lower utilization.

Revenue traction

The product has not formally launched, but Shkreli says the company already runs at millions of dollars in annualized revenue. Earlier AI startup attempts like an AI doctor or text-to-speech tools generated nearly zero revenue. Trading tools sell themselves. TradingView, estimated at around $300 million in revenue, proves the market pays for this category. TradingView ranks in the top 100–200 websites globally despite being a niche charting product.

Bloomberg's defensibility

Shkreli rejects the common claim that Bloomberg's moat is its chat network. The social layer matters but does not explain why quants largely abandoned the terminal. Bloomberg's defensibility is real, but Shkreli argues a well-run financial information company should be worth a trillion dollars, not the roughly $100 billion Bloomberg likely commands today. He estimates Bloomberg generates $13–14 billion in revenue, nearly all margin, and could fetch $200 billion in a sale or float.

Shkreli also speculates that Mike Bloomberg may be significantly wealthier than Forbes suggests. Bloomberg reinvested LP dividends into a family office that anchored KKR, Carlyle, Warburg Pincus, and other major private equity funds over decades. Shkreli estimates Bloomberg's total wealth at $400–500 billion, well above public estimates, though this reflects his own inference rather than disclosed data.

Quantum computing

Shkreli is bearish on quantum computing stocks. Current quantum machines run at around 100 kilohertz compared to 5 gigahertz for classical chips. The best current system is IBM's 135-qubit processor. The only algorithm delivering meaningful computational advantage is Shor's algorithm, which factors large numbers and moves complexity from exponential to polynomial time. Outside of Shor's, no equivalent breakthrough justifies current valuations of public quantum companies. Shkreli calls the stocks worthless and attributes their market caps to retail investors pattern-matching to Nvidia's run from zero to $4 trillion.

Bitcoin's cryptographic exposure

Shkreli argues that roughly 15% of Bitcoin—specifically coins mined by Satoshi using an early P2PK format that exposes the x-coordinate of the elliptic curve public key—cannot be migrated to quantum-secure systems the way modern wallets can. He is pursuing three parallel approaches to eventually crack those keys: brute force, dependent on future chip improvements; a mathematical attack on elliptic curve cryptography, a field he describes as a barren wasteland with perhaps 10 serious global experts; and quantum computing itself. He dismisses the third path working anytime soon but keeps a small team on it.

AI may accelerate the mathematical cryptography work. Demis Hassabis at DeepMind proved Navier-Stokes, which Shkreli cites as a benchmark for what AI can do on unsolved math problems. AI could help non-expert cryptographers do PhD-level work on elliptic curve vulnerabilities.