YC Root AccessThe Startup Powering Billions In Trades Every Day
CHAPTERS
- 0:00 – 1:29
OneChronos in plain terms: an exchange enabling atomic, multi-asset trades
Jared introduces OneChronos as a new stock exchange already handling a meaningful slice of daily U.S. equities volume. The key capability is allowing institutional traders to bundle multiple legs into one “all-or-nothing” transaction, avoiding costly execution risk.
- •OneChronos functions like a new Nasdaq-style equities venue
- •Already ~0.3% of U.S. equities daily volume; billions of dollars traded
- •Core innovation: sophisticated institutional trades that were previously impractical
- •Example: buy Tesla while shorting GM/Ford as one atomic package
- •Framing: “overnight success” driven by years of esoteric math and persistence
- 1:29 – 2:46
Co-founders’ origins: long friendship, tinkering, and early hacking
Kelly and Steven describe meeting in middle school and bonding over computers in suburban Colorado. Their early experimentation (and occasional school discipline) foreshadows the technical curiosity that later shaped the company.
- •Met in middle school; close friends through high school
- •Few distractions besides computers—learned by tinkering
- •Early “computer incidents” framed as harmless/white-hat pranks
- •Longstanding trust lowered co-founder risk
- •Stayed connected while studying/working on opposite coasts
- 2:46 – 3:23
Auction theory influences: seeing markets as an engineering problem
Kelly recounts being influenced by Caltech economics and auction theory (Preston McAfee). That experience reframed markets as computational design problems where mechanisms can be engineered for better outcomes.
- •Undergrad exposure to auction theory and market design
- •Shift from physics mindset to computational economics
- •Key idea: you can engineer economic outcomes via mechanism design
- •Auctions seen as CS/math problems, not just traditional econ
- •Seed of applying richer auction formats to real markets
- 3:23 – 6:13
From hedge funds and cybersecurity to a market-structure insight
The founders connect their professional backgrounds: Kelly looking for signals in markets and Steven detecting cyber threats in large datasets. They realize modern electronic markets resemble distributed databases, with latency-induced “stale reads” creating arbitrage conditions.
- •Kelly learns market plumbing/portfolio-manager needs at a hedge fund
- •Steven applies data tooling to cybersecurity anomaly detection
- •Insight: exchanges resemble distributed databases with consistency issues
- •Latency races map to database anomalies (writes/reads on stale data)
- •Initial goal: design an exchange immune to latency race conditions
- 6:13 – 7:05
The long runway: 2011 ideas, 2016 commitment, 2022 launch
They trace OneChronos from early conceptual work around 2011 to quitting jobs and pursuing it full-time in 2016. Even after YC, the exchange required years of execution before real trading began in 2022.
- •Earliest conceptual timestamp around 2011
- •Went full-time and became a real company in 2016
- •Exchange launched mid-2022 after years of building
- •Illustrates “overnight success” built over a decade
- •Launch timeline shaped by both engineering and market adoption hurdles
- 7:05 – 8:38
YC rejection, MVP pressure, and choosing to build it “the right way”
Kelly describes an early conversation after an initial YC rejection: pressure to compress scope into a fast MVP. They found an MVP path, but as understanding deepened, investors and partners reinforced the need to build the full system correctly because they’d only get one credible launch.
- •Rejected from YC initially; continued dialogue with Jared/YC
- •Challenge: shrink ‘5 years/$10M’ vision into ‘6 months/$500K’ proof
- •They preserved an MVP path while continuing toward full build
- •Investor conviction grew as the model became clearer
- •Decision: maximize launch success rather than rush a partial rollout
- 8:38 – 9:36
Two deep-tech problems: real-time optimization + usability of combinatorial auctions
What began as a single hard scaling problem became two distinct deep-tech challenges. Beyond running complex optimizations fast enough for markets, they also had to make a foreign mechanism—combinatorial auctions—understandable and usable for Wall Street participants.
- •Problem #1: scaling optimization/matching computations to market speed
- •Problem #2: making combinatorial auctions accessible to market participants
- •Combinatorial auctions were (and remain) conceptually unfamiliar in equities
- •Product work needed alongside research/engineering breakthroughs
- •Building the mechanism wasn’t enough—adoption required comprehension
- 9:36 – 12:55
Regulatory gauntlet: educating regulators and building compliance operations
Steven explains the many operational and regulatory tasks required to become a trading venue, including extensive education on a novel matching model. They also highlight how exchanges usually use price-time priority, while OneChronos intentionally removes time priority from matching.
- •Regulated environment adds extensive non-obvious requirements
- •Significant effort educating regulators on a new matching mechanism
- •Contrast: standard exchanges use price-time priority (limit order book)
- •OneChronos matches on price/volume (time only for auction admission)
- •Compliance details range from mundane (office printer) to major (cloud automation)
- 12:55 – 15:14
Launching and cold-starting liquidity: from 200 shares to a hockey stick
They discuss the hardest non-technical hurdle: bootstrapping a two-sided market where trades only happen if participants connect and trust the venue. Early adopters joined despite minimal liquidity, and activity ramped from test trades and 200-share days to meaningful volume by late summer 2022.
- •Cold start/liquidity acquisition described as the ‘mother of all problems’
- •Early participants joined for first-mover reps and client signaling
- •Day-one production: ~200 shares, largely plumbing and reporting validation
- •Onboarding includes settlement/reporting workflows beyond matching
- •Volume inflection: late Aug/Sept after late-June first production trade
- 15:14 – 17:56
Six years of persistence: being “right” but told it would never work
Jared probes how they stayed motivated through years without trades and widespread skepticism about market entry. The founders emphasize that critics often agreed the mechanism was correct, but doubted feasibility—fueling determination to solve the distribution and adoption barriers.
- •Skeptics said execution was impossible, not that the idea was wrong
- •Validation: repeated feedback that it was ‘the coolest thing’ in market structure
- •Hard part: breaking in, getting connectivity and liquidity, and overcoming complexity
- •Founders’ goal was solving the problem, not merely “starting a startup”
- •Mindset: keep investing time because time (and iteration) was what they needed
- 17:56 – 20:32
Why combinatorial auctions matter: exposure problem and “ships in the night”
Kelly explains the mechanism-design rationale through the history of FCC spectrum auctions, where complements/substitutes make sequential auctions inefficient. Combinatorial auctions let participants express portfolio-level intent directly, reducing strategic underbidding and enabling trades that otherwise never surface.
- •FCC spectrum auctions as the canonical use case (complements/substitutes)
- •Sequential auctions create strategic behavior and inefficiency
- •Exposure problem: fear of ending with an unhedged/partial bundle
- •Combinatorial bids express true price+quantity across a package
- •Solves ‘ships passing in the night’: hidden offsetting interests become matchable
- 20:32 – 21:42
How OneChronos changes execution: algorithms vs atomic portfolio constraints
Steven contrasts traditional execution—multi-leg algos that juggle risk as prices move—with OneChronos’ ability to encode constraints like “all-or-nothing.” Jared reframes it as a new API for trading: multi-symbol orders plus constraints as first-class primitives.
- •Traditional approach: leg-by-leg execution via algos, hoping prices cooperate
- •Execution risk arises when one leg moves while filling the other
- •OneChronos enables atomic, all-or-nothing multi-leg transactions
- •Generalizes to multiple symbols plus interaction constraints
- •Reframed as a new trading API/primitive for institutional workflows
- 21:42 – 23:17
Company today: a ~40-person, engineering-led team and small-team philosophy
They describe OneChronos’ current scale—about 40 people—running a high-volume exchange. Hiring spans from non-finance engineers to veteran market-structure builders, unified by first-principles thinking and a deliberate preference for small, high-leverage teams.
- •Team size ~40 despite significant market impact
- •Talent mix: zero-finance backgrounds through veteran market-maker engineers
- •Engineering-led product culture focused on primitives and APIs
- •Third co-founder from Goldman; YC network helped key connections
- •Emphasis on per-capita output (WhatsApp-style leverage)
- 23:17 – 26:57
Future vision: smart markets across asset classes—and even compute markets
The founders argue the approach generalizes beyond U.S. equities to FX, European equities, and many real-economy markets where non-price constraints matter. The conversation extends to GPU/compute allocation as a fragmented, inefficient marketplace with strong complements/substitutes—potentially a fit for combinatorial mechanisms.
- •Strategy: expand to additional asset classes/markets with modular teams
- •Thesis: applies wherever complements, substitutes, or non-price constraints exist
- •Markets today often approximate real-economy activity due to mechanism limits
- •Compute/GPU allocation discussed as inefficient and fragmented
- •Combinatorial auctions could unlock idle capacity and better matching for compute
- 26:57 – 33:55
Career risk and building for Wall Street: opportunity cost, regret minimization, and rigor
They close on whether to leave high-paying finance jobs to start uncertain companies and what it takes to build for Wall Street. The founders emphasize aligning with a mission you deeply care about, minimizing regret, and adopting a critical-systems engineering mindset where you can’t ‘move fast and break things.’
- •Opportunity cost is real; only worth it if you strongly believe it should exist
- •Working first can provide staying power and financial resilience
- •Regret-minimization framing and belief there will ‘always be a job’
- •Money viewed as secondary to meaningful problem-solving (beyond family needs)
- •Wall Street demands critical-systems rigor; trust collapses if you cut corners
