Perpetual Futures for Stocks: The SpaceX Pre-IPO Market

Battle Count: 52

Highly relevant for quantitative trading in several dimensions: (1) The unified no-arbitrage pricing framework provides a rigorous basis for perpetual futures pricing and basis trading strategies. (2) The filtering/martingale reading of price discovery offers a principled approach to pre-IPO and illiquid asset valuation. (3) The segmentation equilibrium model explains persistent basis premiums structurally, informing carry and convergence trade strategies. (4) The stochastic volatility analysis (volatility enters only through carry) is directly actionable for volatility risk premium harvesting. (5) The funding rate as observer gain/learning rate provides a microstructure-informed signal for algorithmic execution. (6) The SpaceX case study demonstrates real-world applicability with quantifiable forecast accuracy. However, the paper is primarily theoretical and the empirical component is a single case study, limiting immediate backtestability.

The theoretical framework involves advanced stochastic calculus (martingale representation, subordination, Doob martingales, stochastic approximation), Kalman filtering, and Bayesian inference. The no-arbitrage pricing (Proposition 1) and random time change (Proposition 2) are analytically tractable. The stochastic volatility analysis (Proposition 5) requires careful handling of the funding clock and spot dependence. The filtering section (Section 8) is standard linear Gaussian filtering but the nonlinear extension via generative Bayesian computation (Section 11) is computationally intensive and requires simulation-based inference. The segmented market equilibrium (Proposition 6) is relatively simple mean-variance optimization. The SpaceX empirical application requires access to perpetual futures data, funding rate series, and IPO pricing data. Overall, implementing the full pipeline from theory to estimation is moderately to highly complex.

The theoretical framework is fully specified with proofs. The SpaceX empirical application uses publicly available market data (cnbc.com, investing.com, tradingview.com, cryptobriefing.com, barchart.com). However, the generative Bayesian computation estimation procedure (Section 11) lacks specific implementation details, prior specifications, and code. The paper is primarily theoretical with a single empirical case study, limiting full reproducibility of the estimation pipeline.