Doug Colkitt - The Evolution of Markets - [Invest Like the Best, EP. 255] — Summary & Takeaways · Colossus (Invest Like the Best / Business Breakdowns)

At a Glance

Former Citadel high-frequency trader and CrocSwap founder Doug Colkitt has transitioned from an HFT trader to a builder of DeFi infrastructure. The main thread of this episode: how the mature mechanisms of traditional electronic markets (e.g., NASDAQ) provide an evolutionary blueprint for DeFi, and how single-contract DEXs can solve the fragmentation and high friction issues of current AMMs. Doug Colkitt argues that concentrated liquidity is the future direction of DEXs, but current Uniswap V3's high gas costs prevent liquidity providers from engaging in active price discovery like traditional HFTs, whereas a single-contract architecture can reduce LP gas costs by 50-70%, thereby unlocking true active market-making behavior.

Theme 1: The Profit Mechanism of Traditional HFT — "A Coin Flip with 51% Odds"

Doug Colkitt argues that the stable profitability of HFT does not come from winning every trade, but from statistical advantages at an enormous scale.

Mechanism Breakdown: If a coin has a 51% chance of landing heads and 49% tails, a single bet carries high risk; but after 10,000 flips, the win rate becomes almost certain. This is exactly how HFT works — "they're making money on 51% of their trades, but doing 10,000, 100,000 trades a day," and the law of large numbers makes daily profit and loss highly predictable.
Data Chain: Traditional HFT generates annualized returns of approximately 20-30%, with Sharpe ratios reaching above 30 (far exceeding traditional asset management). However, capacity is extremely limited — "Whether your returns are 500% a year or 2,000% a year... capacity constraints are by far the biggest limit." The entire U.S. HFT industry generates annual revenue of roughly $2 billion to $20 billion (depending on the definition).
Key Distinction: Not all HFT firms are market makers (liquidity providers). In fact, the most profitable strategies are often liquidity takers — "they immediately cross the spread," exploiting cross-market spreads (e.g., S&P futures in Chicago, constituent stocks in New York) for statistical arbitrage.
Falsification Condition: HFT profitability is not guaranteed — "it's definitely hard to get to the point where you're profitable at all." But once the profitability threshold is crossed, daily profits become almost certain.

Theme 2: The Evolution of AMMs — From "Zero to Infinity" to "Concentrated Liquidity"

Doug Colkitt argues that the core innovation of AMMs lies in enabling non-professional participants to provide liquidity, but classic AMMs suffer from extremely low capital efficiency, making concentrated liquidity an inevitable evolutionary direction.

Historical Context: Vitalik Buterin proposed the AMM concept (constant product curve), and Hayden Adams commercialized it by founding Uniswap. The 2020 DeFi Summer saw explosive growth for Uniswap due to its "permissionless" nature (allowing any token to be listed without exchange approval) — "the whole space has grown probably more than a thousand percent just the past year."
Mechanism Breakdown: In classic AMMs, liquidity providers (LPs) allocate capital uniformly across a price range from "zero to infinity." When a trade occurs, the ratio of the two assets in the pool changes, automatically adjusting the price. LPs earn trading fees (typically the majority of the 0.3% fee) but face "impermanent loss" — "the pool's always rebalancing in the wrong direction."
Key Innovation: Curve was the first to address capital efficiency — designing a special curve for stablecoin pairs (e.g., USDT/USDC), "it kind of pretends, oh, we have more capital than we do." Uniswap V3 introduced concentrated liquidity — LPs can choose their own price range (e.g., $0.80 to $1.20), requiring capital only within that range, but necessitating active position management.
Core Contradiction: Concentrated liquidity is theoretically more efficient, but current gas costs are too high for LPs to frequently adjust quotes — "you might cancel 99% of your orders... that means to do one trade, I have to do 100 different actions." The result is that "people aren't moving these positions around... the positions go out of range and then they forget about it a week later."

Theme 3: Single-Contract DEX – Bringing HFT Logic into DeFi

Doug Colkitt argues that CrocSwap's single-contract architecture is key to solving the high friction of concentrated liquidity, reducing LP gas costs by 50-70% and enabling active market making.

Technical bottleneck: Ethereum's 24KB size limit per smart contract forces existing DEXs (e.g., Uniswap) to treat each liquidity pool as an independent contract. This creates significant friction—if an LP wants to adjust quotes (e.g., moving from a 0.3% fee pool to a 0.05% fee pool), they must "take my liquidity out of the higher pool, burn my position... transfer those tokens to the other pool," incurring substantial gas costs and taxable events.
CrocSwap's solution: Place all pools within a single contract, using internal data structures to track different pools. "The single contract DEX can just say, okay, this is how much net out the position," allowing execution of arbitrarily long operation chains with net settlement at the end.
Quantitative improvements: Trader gas costs reduced by 10-20% (by holding collateral directly on the DEX, avoiding back-and-forth token transfers); LP gas costs reduced by over 50%, reaching 70% in certain scenarios. Tax implications—avoiding taxable events caused by intermediate tokens (e.g., ETH) entering and leaving addresses.
Engineering challenges: Implementing full functionality within the 24KB limit—"a lot of pain and suffering, basically, in engineering." Borrowing the HFT optimization approach of "hot path vs cold path" (making 99% of common operations faster, even at the expense of performance in 1% of edge cases).
Future expansion: The single-contract architecture enables "cross margin" (cross-market margin)—"all the capital's at a single place," supporting leveraged trading and perpetual contracts, a key area where DeFi has yet to catch up with centralized exchanges.

Theme 4: The Future of DeFi—The "Lego Blocks" of Global Capital Markets

Doug Colkitt argues that the ultimate form of DeFi is to enable real-world assets (stocks, bonds) to trade on-chain, potentially merging with centralized exchanges through a "sponsored pool" mechanism.

Long-term Vision: Within 5-10 years, "there's going to start being real assets start trading on DeFi." Using Turkey as an example—"If I told you, oh, this stock in Turkey is really hot... would you go to Turkey and open a brokerage account? No." Tokenization can directly connect emerging market assets to global capital.
Technology Analogy: Blockchain scalability follows an exponential growth pattern similar to Moore's Law—"the scalability keeps doubling... stuff like that often seems like a toy... with that kind of exponential improvement, it's not very long until it starts replacing legacy systems."
Relationship with Centralized Exchanges: It is not a zero-sum game. CrocSwap has designed a "sponsored pools" mechanism—"Binance can create a Binance pool subject to the approval of the CrocSwap token holders," allowing centralized exchanges to transparently access DeFi liquidity on the backend. This is analogous to the U.S. Reg NMS rule (ensuring orders are routed to the best available quote).
Missing Link: The greatest demand is for stablecoins—"people want dollar backed... especially middle income countries." This is currently met through over-collateralization (DAI) or centralized custody (USDC/USDT), but demand remains far from satisfied, leading to "people are going to keep doing riskier and riskier things to generate synthetic stable coins."
MEV (Miner Extractable Value): Analogous to HFT—"strategies where people who invest a lot in very, very high precision of how their trades... are executed, earn these outsized profits." The solution is not to eliminate MEV, but to "build better systems on the front end to stop users from making bad decisions that can be exploited."

Mentioned Positions

Position	Guest Attitude	Key Data
Uniswap	Acknowledges its innovation, but points out that V3's concentrated liquidity is not effectively utilized due to high gas costs	Growth of over 1,000% since 2020; V3 introduced concentrated liquidity but LPs cannot actively adjust quotes
Curve	Affirms its capital efficiency innovation	Designs special curves for stablecoin pairs, addressing the capital waste issue of classic AMMs
SushiSwap	Mentioned as a case of "vampire attack"	Captured nearly 50% market share from Uniswap overnight through liquidity mining rewards
CrocSwap	Founder's project, optimistic about its single-contract architecture	LP gas costs reduced by 50-70%, trader gas costs reduced by 10-20%
Binance	Viewed as a potential partner (sponsorship pool mechanism)	Can transparently access DeFi liquidity through CrocSwap's sponsorship pool
NASDAQ/NYSE	Seen as benchmarks for mature markets, but believes DeFi may integrate with them through a Reg NMS-like mechanism	Traditional HFT industry annual revenue of $20-20 billion

Judgments Worth Remembering

1. HFT's stable profitability stems from the law of large numbers, not winning every trade — "they're making money on 51% of their trades, but doing 10,000, 100,000 trades a day." Sharpe ratios can exceed 30, but capacity is extremely limited, and the difference between an annualized return of 500% and 2,000% matters less than absolute P&L.

2. The most profitable HFT strategies are often not market making, but cross-market statistical arbitrage — "they immediately cross the spread," exploiting the price differential between S&P futures and their constituent stocks. This refutes the common misconception that "HFT = market maker."

3. The capital efficiency problem of classic AMMs stems from "zero to infinity" price coverage — LPs are forced to allocate capital to price ranges that can never be reached. Curve optimizes for stablecoin pairs, and Uniswap V3 introduces concentrated liquidity, but the latter's high gas costs prevent LPs from actively managing positions.

4. Concentrated liquidity is the future of DEXs, but current gas costs reduce it to a passive "set and forget" tool — "people aren't moving these positions around... the positions go out of range and then they forget about it a week later." This directly contradicts the original intent of concentrated liquidity design.

5. Single-contract DEXs can reduce LP gas costs by 50-70%, enabling active market making — by placing all pools within a single contract, netting is allowed and token transfers back and forth are avoided. The engineering challenge lies in Ethereum's 24KB contract size limit.

6. The ultimate form of DeFi is the tokenization of real-world assets — "there's no reason a stock can't tokenize." Taking Turkey as an example, tokenization allows emerging market assets to directly access global capital without needing to open a local brokerage account.

7. DeFi and centralized exchanges are not a zero-sum game but can merge through a "sponsor pool" mechanism — CrocSwap has designed a mechanism where Binance can create a "Binance pool," allowing centralized exchanges to transparently access DeFi liquidity on the backend, similar to the U.S. Reg NMS rules.

8. MEV is essentially a mapping of HFT onto the blockchain — "strategies where people who invest a lot in very, very high precision of how their trades... are executed, earn these outsized profits." The solution is not to eliminate MEV, but to build better protection mechanisms on the front end.

9. Stablecoins are the largest unmet demand in DeFi — "people want dollar backed... especially middle income countries." Current demand far exceeds supply, leading to "people are going to keep doing riskier and riskier things to generate synthetic stable coins."

Doug Colkitt - The Evolution of Markets - [Invest Like the Best, EP. 255]

In plain words