Aave reserve state
Borrow/supply rates, exact debt indexes, liquidity, collateral thresholds and reserve restrictions.
Stack Research · early access
Test the idea. Inspect the evidence.
Would consistent investing and active loan protection help your treasury? Start with evidence you can reproduce, then compare Stack against unlevered investing and a competent direct-Aave implementation.
Evidence checkpoint: October 7, 2026. This page is not a live collector-status monitor.
What you can try today
Read eight reconstructed daily Base native-USDC observations and reproduce the interest on a hypothetical fixed 1,000 USDC variable-rate loan. From September 30 to October 7, 2026, the debt-index calculation gives 0.914804389 USDC in interest, before other costs. Independent direct-Aave reads matched the calculation.
This measures loan interest only. It does not measure portfolio returns, DCA purchases, liquidation prevention or a Stack advantage. The same loan has the same protocol interest through Stack or direct Aave; service and execution costs must be added separately.
curl -fSs https://www.stackit.ai/research/sample.json -o sample.json curl -fSs https://www.stackit.ai/research/reproduce.py -o reproduce.py # Inspect the script, then run locally (Python standard library only): python3 reproduce.py sample.json
What we are collecting
Base, Arbitrum and Polygon. Native USDC and bridged stablecoins have separate identities. Collection began October 7, 2026; selected older reserve states were reconstructed later and retain their ingestion times.
Oracle and sequencer state
Aave oracle prices, available oracle observations and sequencer/grace-period state. Missing reads remain gaps.
Execution context
DEX quote observations, modeled gas/action costs and bounded finalized transaction fee samples. These are not verified Stack execution costs.
Protocol changes
Finalized event observations, freeze/pause signals and deployment checks. Complete borrower coverage is not established.
Configured collection targets: chain snapshots every 3 minutes; Base V4 state every 5 minutes; DEX quotes and fee samples every 15 minutes; deployment watch daily. These are schedules, not a guarantee of successful or complete collection. V4 observations remain provisional pending deployment/state qualification.
Hypothetical stress calculation
How much room does your health factor leave?
Aave positions below health factor 1 can become eligible for liquidation, subject to protocol conditions. Explore a simple price shock; this does not predict a real liquidation or simulate Stack protection.
Resulting health factor: 0.960
Below the HF 1 threshold
Price decline reaching HF 1: 37.50%.
Assumptions: all collateral prices fall equally; debt asset price and liquidation thresholds stay fixed; no deposits, repayments or collateral yield. HF after shock = starting HF × (1 − price decline) ÷ (1 + debt increase). Real protection outcomes need timed oracle observations, executable quotes, costs and actual action receipts.
Why investigate consistent deposits?
A fixed contribution buys more units when prices are lower. A fair comparison gives every strategy the same cash on the same dates, counts unspent cash, and separates deposits from profit. Regular investing can reduce timing decisions; it does not always outperform holding cash or investing earlier.
The planned strategy lab will compare contribution patterns, a 180-day DCA policy, loan repayment and protection under observed data or explicitly hypothetical price paths. That combined engine is not available here yet.
What would justify using Stack?
Measure the complete result: net equity after debt and fees, drawdown, minimum health factor, liquidation eligibility, missed actions and operator maintenance. Compare against a DIY implementation using the same rules and information.
Reliable execution, monitoring and recovery are hypotheses to verify with timed replay and operational evidence. The dataset alone does not prove that Stack earns its fees or that protection will always execute.
Access and limitations
- The public sample, coverage manifest and calculator are free and require no wallet or API key.
- The larger research API is private and requires operator-approved access. It is separate from the public treasury sandbox and MCP tools.
- No anonymous database export or public full-strategy backtest endpoint is available. Research x402 payments are not activated.
- No continuous three-year dataset, complete intraday price/position history, full borrower census or verified historical liquidation/protection outcome is claimed.
- Historical reconstruction records late ingestion. Do not treat those observations as information known to Stack at the original block time.
To request bounded research access, email support@stackit.ai with the question, chains, assets and date range you want to evaluate. Access and commercial terms require operator review; sending a request does not grant access.