giacosdila re-weights each dollar-cost averaging tranche against real-time volatility signals, so capital is deployed when conditions are statistically favourable rather than on a fixed date.
Read-only API connection. No withdrawal permissions requested.
Standard DCA buys on a fixed schedule regardless of conditions. Smart Entry keeps the same disciplined cadence but adjusts tranche timing and size using a volatility-weighted entry score, so the same capital is allocated with more context.
Realised volatility is sampled continuously across each configured asset, rather than at the moment of a scheduled purchase.
The current volatility regime is compared against its recent historical range to produce a weighting score for the upcoming tranche.
A portion of the scheduled DCA allocation is sized according to the weighting score, within the risk thresholds you define.
The order is placed through your connected exchange API, with the weighting rationale recorded for later review.
Each capability below exists to keep the system auditable and your exposure bounded, rather than to maximise trading frequency.
Exchange connections use read-only or trade-only API scopes where supported. Withdrawal permissions are never requested, and credentials are stored encrypted at rest.
Entry weighting draws on stochastic volatility models recalibrated against rolling market windows, rather than a single static training set.
Set maximum allocation per tranche, drawdown-triggered pauses, and an asymmetric risk-reward ratio for how aggressively weighting can deviate from baseline.
Pricing and order-book depth are aggregated from licensed exchange feeds to reduce single-source data gaps.
giacosdila was built around a simple position: traders evaluating an automated system deserve to see how it reasons, not just what it promises. Every weighting decision is logged with the volatility inputs that produced it, so allocation history can be reviewed line by line rather than taken on trust.
The platform is intended as infrastructure that sits alongside your existing exchange relationships, not a replacement for your own risk judgement.
Historical testing is run on a walk-forward basis: each model version is validated only on data occurring after its training window, so results cannot be fitted to the past with foresight the live system would never have had.
Historical price series are checked for gaps, duplicate timestamps, and exchange outage artefacts before being used in any recalibration run. Periods with unresolved data quality issues are excluded rather than interpolated.
Recalibration follows a fixed cycle rather than continuous tuning, which limits how tightly the model can adapt to short-term noise in any single market cycle.
Collect new volatility data from the completed window.
Re-train weighting parameters on the extended dataset.
Validate against the unseen out-of-sample period.
Deploy only if validation metrics hold within tolerance.
The same weighting engine supports different accumulation goals, depending on how risk thresholds are configured.
Investors building a long-term position who want disciplined entries without manually checking volatility conditions before every scheduled purchase.
Traders running intraday strategies who use giacosdila separately to manage a core holding, so short-term positioning and long-term accumulation do not interfere with each other.
Allocators who want exposure capped during periods of unusually high realised volatility, using the asymmetric risk-reward setting to reduce tranche size rather than skip it entirely.
Account setup and risk-parameter configuration are handled in a single session. No trading activity begins until thresholds are confirmed.
Typical setup time: under 15 minutes