Short-Term Statistical Arbitrage

Short-term statistical arbitrage is the source’s label for the Medallion Fund-style move toward many small, repeated, short-horizon trades. In vol.103.文艺复兴科技西蒙斯的封神之路：是量化之王，更是洞察人性的大师, Elwyn Berlekamp argues that holding positions for hours or days can make weak signals tradable if costs, market impact, margin, and sizing are modeled correctly.

The source distinguishes this from both classic long-term fundamental investing and modern millisecond high-frequency trading. Its edge comes from probability repetition, short exposure windows, many markets, and disciplined execution.

Key Claims

• A tiny edge can matter if it is repeated often, kept low-correlation, and not erased by trading friction.
• Kelly Criterion and fractional Position Sizing matter because overbetting can destroy a positive-expectation strategy.
• Shorter holding periods can reduce exposure to single long-term thesis failure, but they create execution and cost sensitivity.
• The strategy depends on Quantitative Data Moat because weak effects need a lot of clean observations.
• Alpha Decay and Market Efficiency mean short-term signals need constant replacement and testing.

Connections

• Elwyn Berlekamp — source figure most associated with the shift.
• Medallion Fund and Renaissance Technologies — fund and firm context.
• Kelly Criterion, Position Sizing, and Investment Risk Management — sizing and survival framework.
• Quantitative Investing and Quantitative Data Moat — broader method and required data substrate.
• Alpha Decay and Market Efficiency — why the edge is small and temporary.