New Risk Parity optimizer demo
SynthEngine — a new Risk Parity optimizer built on the Synth API — launched a demo showing forward-looking volatility percentiles for weighting, rebalancing alerts, Monte Carlo sims and VaR/CVaR outputs with a 9-month backtest vs. SPY demo posted. It’s another example of productizing forward-looking volatility into allocation signals for tactical risk-weighting.
The demo appears to source the forward-looking distribution directly from Synth’s prediction-percentiles API — the docs show GET /insights/prediction-percentiles returns full forecast percentiles (0.005–0.995) across the 24h horizon with 289 time steps for the 24h horizon. docs.synthdata.co Synth exposes a dedicated volatility insights endpoint that returns both forecasted and realized volatility metrics and explicitly lists SPY among covered equities, making those series available to feed portfolio-level risk modules. docs.synthdata.co The platform’s forecasts are ensemble-based: the docs state each miner/model outputs an ensemble of simulated price paths (the glossary and intro document reference 1,000 simulated paths per model per horizon), which provides the raw scenario set used for Monte‑Carlo-style aggregation, VaR/CVaR estimation and percentile-driven reweighting. docs.synthdata.co Synth’s documentation includes a concrete portfolio-VaR example built from prediction-percentiles, and the endpoints overview lists option-pricing and v2 prediction-path endpoints for advanced backtests; Synth also supports short-lived demo API keys for notebook/quick-prototype use. docs.synthdata.co
