Daily research update
we backtested it — Sharpe 0.64 · quant relevance 0.97
The study establishes convergence and dimension-dependence bounds for SGLD applied to regularized, softmax-parameterized portfolio CVaR minimization, with numerical validation on simulated Gaussian losses. The optimizer is implementable using historical ETF returns, but its simulation results do not establish realized investment performance, and its i.i.d. assumptions need not hold for market returns.
quant relevance 0.99
The study directly evaluates daily cryptocurrency volatility forecasting and risk-aware portfolio allocation, including transaction costs. It reports that a historical baseline forecasts next-day absolute returns better than more complex models, while inverse-volatility allocation performs best after 10 bps costs. The framework is implementable with daily crypto prices, but the five-coin coverage must be verified and the reported sample through September 2026 exceeds the documented historical backtest window.
quant relevance 0.99
The strategy ranks three Indian sector indices quarterly using earnings growth, historical valuation percentiles and momentum, then equally weights the top two. It reports higher test-period returns than the Nifty 100, but no statistically significant Sharpe-ratio improvement after multiple-testing adjustment and no clear advantage over momentum alone. The strategy is directly relevant to systematic investing, but its Indian-market prices and sector-level valuation histories are unavailable here.
The rest of the batch
The study identifies election-night prediction-market prices that lag vote-count projections and estimates model-dependent expected profits from buying the projected winner.
The study measures how alternative arbitrage-free repairs of the same crypto option quotes change Greeks, hedge notionals, and risk estimates.
Monthly margin purchases relative to trading value predict subsequent Chinese A-share returns, with positive high-minus-low quintile spreads over one-, three-, and six-month horizons.
The paper develops theoretical conditions for arbitrage-free non-convex cost-function market makers, fee-induced no-trade bands, and optimal routing across parallel makers.
The paper develops regime-switching derivative valuation methods with fast-switching corrections and numerical fallback, directly relevant to options pricing and hedging.
FinTime-Fuse forecasts financial prices by combining temporally aligned financial text and numerical price patches through FinBERT representations and a frozen LLaMA backbone.
and 14 more · see the full radar
Backtests are hypothetical research results; forward-portfolio figures are a live record. Neither is a guarantee of future performance. Research and education only: not individualised investment advice, not an offer to sell or a solicitation of an offer to buy any security.