Skip to content

2023 Q2 Quarterly Research Archive

Records scoring at least 40 within the primary scope pass rule review and are published without additional manual review. This page does not validate author claims or provide investment advice.

  • Coverage: 2023-04-01 to 2023-06-30
  • Passed rule review: 5
  • Sources: 4

Topic distribution

Domains

  • volatility: 2
  • microstructure: 2
  • execution costs: 2
  • hedging exposure risk: 1

Methods

  • financial ml: 2
  • research methods: 1

Facets

  • instrument index options: 1
  • structure straddle: 1

Passed rule review

Volatility (LIVE IN VEGAS) panel discussion

  • Published: 2023-06-29
  • Source: The Derivative by RCM Alternatives
  • Publication status: unknown
  • Original source: Open original source

Why it matters

This episode is a panel narrative on volatility, macro links, and platform dynamics. It is useful as risk narrative context but should not be treated as empirical evidence by itself. (description:S1, description:S3, description:S4, description:S5, description:S9)

Main author claims

  • The authors report: The episode preview frames discussion of volatility dynamics, risk management, and evolving market structure, including liquidity and options topics. (description:S3, description:S4, description:S5, description:S9)
  • The authors report: The description indicates discussion of volatility’s portfolio role, OTDE options, hedging, tail protection, Fed constraints, and large-fund influence. (description:S7, description:S8, description:S9, description:S1)
  • The authors report: The episode includes disclaimers that it is informational only and not legal, business, or tax advice, with risk warnings. (description:S11, description:S12, description:S13, description:S14)

Data, method, or discussion scope

Evidence is limited to promotional panel topics and does not include quantified model validation, sample periods, or inferential testing. (description:S1, description:S6, description:S9, description:S10)

Main limitations

Narrative format lacks data links, auditable metrics, and explicit falsification criteria. (description:S1, description:S7, description:S10, description:S3)

Constructing Time-Series Momentum Portfolios with Deep Multi-Task Learning

  • Published: 2023-06-08
  • Source: arXiv Quantitative Finance History
  • Publication status: preprint
  • Original source: Open original source

Why it matters

The paper claims a multi-task framework jointly learning TSMOM construction and volatility tasks, which could reduce modeling mismatch; stated outperformance and tail protection must be interpreted through execution and design constraints. (abstract:S2, abstract:S4, abstract:S5, abstract:S6)

Main author claims

  • The authors state that TSMOM performance depends on both momentum signal quality and volatility estimation efficacy. (abstract:S2, abstract:S3)
  • They propose deep MTL jointly learning portfolio construction and auxiliary volatility tasks, tested on diversified continuous futures portfolios. (abstract:S4, abstract:S6)
  • They claim backtests from Jan 2000 to Dec 2020 show outperformance versus existing TSMOM after up to 3 bp transaction costs, and that auxiliary tasks improve performance. (abstract:S5, abstract:S6)

Data, method, or discussion scope

Evidence scope is confined to abstract-level backtest claims without asset-level composition, architecture specifications, cost model definitions, or confidence intervals. (abstract:S4, abstract:S5, abstract:S6, abstract:S1)

Main limitations

Backtest details and benchmarks are omitted, and statements about tail-risk behavior are not stratified by stress scenarios. (abstract:S1, abstract:S5, abstract:S7)

A Leland model for delta hedging in central risk books

  • Published: 2023-05-11
  • Source: Mathematical Finance
  • Publication status: peer_reviewed
  • Original source: Open original source

Why it matters

The paper places market orders, limit orders, and liquidity provision in one delta-hedging framework for joint trading and hedging decisions in a central risk book. (abstract:S1, abstract:S4)

Main author claims

  • In the model's continuous-time limit, optimal limit-order exposure is obtained pointwise and depends on adverse selection, bid–ask spreads, and volatility. (abstract:S2, abstract:S3)
  • The authors characterize the corresponding option price through a nonlinear PDE and report that parameter estimation and strategy simulation can be connected to real high-frequency data. (abstract:S4, abstract:S7)

Data, method, or discussion scope

The abstract identifies the Leland extension, centralized automated desk, continuous-time solution, and high-frequency parameterization, but not the market, sample, latency, fill assumptions, or numerical benchmark details. (abstract:S1, abstract:S2, abstract:S6, abstract:S7)

Main limitations

A reduced-form model being reconciled with high-frequency data is not independent empirical validation; the abstract is insufficient to assess discrete execution, non-fills, or stress-market limits. (abstract:S6, abstract:S7)

Term spreads of implied volatility smirk and variance risk premium

  • Published: 2023-05-08
  • Source: Journal of Futures Markets
  • Publication status: peer_reviewed
  • Original source: Open original source

Why it matters

The paper links S&P 500 implied-volatility curve term factors to variance risk premium predictability and notes differential predictability across return proxies, which matters for implementation specificity. (abstract:S1, abstract:S3, abstract:S4)

Main author claims

  • The authors study implied-volatility curves and their predictive power for the variance risk premium. (abstract:S1, abstract:S2)
  • They report that the level-factor term spread predicts VRP proxies (straddle and variance swap returns) in-sample and out-of-sample. (abstract:S3)
  • The authors report: Predictability is stronger for straddle returns than for variance-swap returns. (abstract:S4)

Data, method, or discussion scope

Evidence is limited to abstract-level statements on curve structure and predictive relations, without sample-window, significance threshold, or market-structure sensitivity details. (abstract:S1, abstract:S3, abstract:S4)

Main limitations

The abstract omits term-structure definitions, friction modeling, and noise controls, all of which matter for operational robustness. (abstract:S3, abstract:S4, abstract:S1)

Recent advances in reinforcement learning in finance

  • Published: 2023-04-07
  • Source: Mathematical Finance
  • Publication status: peer_reviewed
  • Original source: Open original source

Why it matters

This peer-reviewed RL survey, while similar in scope to the preprint, is stronger as a methodological reference point but still does not replace deployment-level validation. (abstract:S2, abstract:S3, abstract:S4, abstract:S7)

Main author claims

  • The authors claim RL can make better use of large financial data with fewer model assumptions than traditional analytical approaches. (abstract:S2)
  • They present an MDP overview, value- and policy-based methods, and the extension to deep RL via neural networks. (abstract:S4, abstract:S5, abstract:S6)
  • The authors report: Applications listed include optimal execution, portfolio optimization, option pricing and hedging, market making, smart order routing, and robo-advising. (abstract:S7)

Data, method, or discussion scope

Evidence is review-level and directory-like, with no per-algorithm objective formulations, training costs, or stability boundaries in the abstract. (abstract:S1, abstract:S4, abstract:S8)

Main limitations

No migration criteria across application scenarios, deployment latency, compute budget, or delivery constraints are detailed here. (abstract:S3, abstract:S7, abstract:S8)