Conservative strategy: Core vault explained
The Conservative agent strategy routes USDC to the Core vault — Spark Savings, Morpho Steakhouse Prime, and Aave V3 on Base mainnet.
Last updated: May 25, 2026 · Published 2026-05-25
Forge Treasury smart contracts are unaudited. Yield is variable and not guaranteed. Read Risks & Disclosures before depositing USDC or integrating MCP tools.
In Forge product canon, Conservative is a strategy, not a vault name. For MVP, the Conservative strategy means 100% of allocated USDC goes to the Core vault on Base mainnet at `0x2C9a3922b426005B979FDD1A8F43Eb61B309193B`. There is no separate "Conservative vault" contract — only the Core vault with a stable yield adapter book spanning Spark Savings, Morpho Steakhouse Prime, and Aave V3 on Base.
What the Core vault holds
The Core vault is an ERC-4626 USDC vault. Depositors receive fUSDC shares representing pro-rata claims on vault assets. On deposit, USDC routes across a flat adapter list with governance-set allocation weights (basis points summing to 10,000). Adapters wrap external yield venues; the vault owner can schedule changes on a 48-hour timelock. See live allocations on Stats or via MCP vault info tools.
| Adapter | Target bps | Role |
|---|---|---|
| Spark Savings USDC | 6000 (60%) | Savings-rate leg on Base |
| Morpho Steakhouse Prime USDC | 2500 (25%) | Curated MetaMorpho exposure |
| Aave V3 USDC | 1500 (15%) | Supply-side lending — withdraw last |
P2-C0 expansion (timelock ~May 27, 2026)
Track P2-C0 expands the Core adapter book toward a five-leg allocation including Gauntlet Prime, Steakhouse HY, expanded Steakhouse Prime, Spark, and Aave. Scheduling uses Forge's standard 48-hour timelock — pending adapters are visible on-chain before execution. Until timelock executes, live weights may differ from targets. Read the full Morpho context in Morpho on Base.
| Index | Adapter | Target bps |
|---|---|---|
| 0 | Gauntlet Prime | 1500 |
| 1 | Steakhouse HY | 1250 |
| 2 | Steakhouse Prime | 2250 |
| 3 | Spark Savings USDC | 3500 |
| 4 | Aave V3 USDC | 1500 |
Who should use Conservative routing
Agents that prioritize stablecoin-denominated yield over ETH beta or experimental infra exposure should default to Conservative routing. The strategy avoids Middle vault adapters (LST swaps, Avantis avUSDC, Moonwell) and planned Infra venues (Virtuals, Venice). It is the simplest single-vault deposit path today: one USDC approval, one `deposit` call. For cross-vault exposure, see Balanced strategy.
- DAO treasuries holding idle USDC on Base that want diversified protocol exposure without LST beta.
- MCP agents using legacy `profile: "conservative"` — maps to Core until strategy refactor.
- Integrators testing Forge with minimal surface area before Middle timelock completes.
Yield expectations (honest framing)
There is no fixed APY. Core legs earn from variable underlying protocol rates — typically ~3.5–4% at the USDC adapter level in May 2026 research snapshots (Aave Base USDC rates, Morpho Steakhouse Prime), but rates move daily with utilization and governance. FORGE emissions, if claimed, are a separate inflationary usage subsidy — not vault yield. Do not treat MCP illustrative APY strings as realised returns; read FORGE emissions vs vault yield.
Deposit and withdraw mechanics
- Approve USDC (`0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913`) for the Core vault address.
- Call `ForgeVault.deposit(assets, receiver)` — vault routes USDC across active adapters.
- Receive fUSDC shares; NAV accrues as adapters earn yield.
- On withdraw, Forge charges a 15% performance fee on profits above high-water mark — not on principal.
- Post P2-C0 first deposit may require elevated gas (~4M) due to five-leg routing.
Fees and token economics
Fee routing intent includes protocol buyback via `FeeDistributor`; MVP components may be stubbed. FORGE rewards flow through a separate Merkle `RewardDistributor` — claim is not automatic on deposit. See Tokenomics for the emission and fee model, and Risks & Disclosures for fee accounting risk.
Share token and NAV mechanics
Core vault shares trade under the legacy symbol fUSDC — treat this as the Core share token in new integrations. Share price equals total vault assets divided by total shares, standard ERC-4626 math. As Spark, Morpho, and Aave adapters accrue yield, share price rises; depositor principal is not guaranteed and can fall on smart-contract failure or USDC depeg. Agents reporting "APY" should derive it from share price delta over time, not from MCP default constants.
| Parameter | Value | Notes |
|---|---|---|
| Chain ID | 8453 | Base mainnet |
| USDC | 0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913 | Circle native USDC |
| Core vault | 0x2C9a3922b426005B979FDD1A8F43Eb61B309193B | Basescan verifiable |
| Share token | fUSDC | ERC-4626 shares |
| Strategy weight | 100% Core | No Middle/Infra in MVP Conservative |
Timelock governance and adapter changes
Forge vault owners schedule adapter changes on a 48-hour timelock. Pending adapters appear on-chain before execution — integrators must not assume static books across weeks. P2-C0 is the next major Core change (~May 27, 2026), shifting from three legs to five. After execution, first deposit gas requirements increase because the vault routes across more wrappers. Withdraw order also changes: Aave remains last in the withdraw sequence to reduce pool stress.
Conservative vs Balanced — decision guide
| Factor | Conservative (Core only) | Balanced (Core + Middle) |
|---|---|---|
| Asset exposure | USDC stable yield only | USDC + ETH beta + avUSDC + Moonwell |
| Deposit txs | One | Two (today) |
| Gas (first deposit) | ~4M post P2-C0 | ~4M Core + ~6M Middle |
| Timelock dependency | Core live | Middle adapters must execute |
| Agent complexity | Lowest | Medium — split amounts |
If your agent policy requires stablecoin-denominated exposure without LST price beta, Conservative is the correct default. If you accept Middle book risks for diversified return sources, read Balanced strategy after verifying Middle `adapterCount` on Stats.
Operational notes for integrators
- TVL on mainnet is early-stage (~$10 smoke testing as of May 2026) — not a production AUM figure.
- Share price and adapter allocations update on-chain; subgraph indexing may lag minutes.
- Use elevated gas limits (~4M) after P2-C0 multi-adapter expansion on first deposit.
- Legacy MCP `conservative` profile maps here until multi-vault strategy tools ship.
- Spark mechanics detailed in Spark Savings USDC; Aave in Aave V3 on Base.
Conservative strategy remains the recommended default for agent policy templates until Middle and Infra vaults complete smoke testing and external review milestones. Even after Balanced goes live, many treasuries will keep a Core-only sleeve for stablecoin liquidity and a Middle sleeve for risk budget — mirroring traditional "cash vs investments" splits, implemented as separate ERC-4626 share tokens on Base.
Guided learning path for new integrators
Start with MCP-native treasury to understand calldata-only integration, then read this Conservative guide for Core-specific parameters. Drill into protocol legs as needed: Spark Savings USDC for savings-rate mechanics, Morpho on Base for curated lending (note P2-C0 timelock), and Aave V3 on Base for supply-side diversification. Finish with FORGE emissions vs vault yield before writing any user-facing APY copy.
Deploy test transactions with minimal USDC on Base mainnet only after reading Risks & Disclosures and verifying live adapter state on Stats. Conservative routing is the lowest surface-area path — one vault, one share token, three protocol adapters (expanding to five post P2-C0). Document chain ID 8453 and native USDC address in your agent system prompt to prevent cross-chain asset mistakes.
Appendix: reading on-chain Core state
Integrators should treat the Core vault contract on Basescan as source of truth. Key view functions include `totalAssets()`, `totalSupply()`, and adapter registry getters exposed by the ForgeVault ABI. Subgraph indexing on Stats lags blocks — for dispute resolution or large deposits, read contracts directly via viem, ethers, or cast. Share price equals totalAssets/totalSupply adjusted for ERC-4626 decimals.
When P2-C0 executes, watch for `adapterCount` increasing from three to five. First deposit after expansion is the gas stress test — recommend ~4M gas limit in agent UX. Withdraw paths may touch multiple adapters in governance-defined order; Aave remains last. If any adapter reverts on withdraw, entire vault withdraw may fail — this is composability risk across Spark, Morpho, and Aave legs.
| Task | Via MCP | Manual (no MCP) |
|---|---|---|
| Read vault metadata | forge_get_vault_info | Basescan + ABI |
| Build deposit tx | forge_deposit calldata | ERC-4626 deposit encode |
| Check timelock | Tool warnings + Stats | Read pending adapter events |
| Claim FORGE | Separate tool / docs | RewardDistributor contract |
Frequently asked questions
Is there a "Conservative vault" contract?
No. Conservative is allocation guidance pointing at the Core vault only. Searching Basescan for "Conservative" will not find a match — use Core address `0x2C9a3922b426005B979FDD1A8F43Eb61B309193B`.
Can I deposit directly to Spark instead of Core?
Yes, but you lose Forge adapter diversification and FORGE emission eligibility tied to fUSDC shares. Core vault bundles Spark with Morpho and Aave in governance weights. Direct Spark deposit is a different risk/reward path — see Spark Savings USDC.
What is smoke-test TVL and why does it matter?
Mainnet TVL (~$10 as of May 2026) reflects intentional private beta sizing, not production AUM. Low TVL means less battle-testing of withdraw paths under stress. Agents should disclose early-stage status to end users and link Risks & Disclosures.
How do I verify adapters before a large deposit?
Call MCP vault info, read `adapterCount` and per-adapter bps on Stats, and cross-check Basescan contract state. After P2-C0 timelock, re-verify — books change without frontend deploys.
Summary for agents
Conservative = 100% Core vault on Base mainnet (`0x2C9a3922…193B`). Verify live adapters and bps on Stats before every deposit — P2-C0 timelock (~May 27, 2026) expands the book from three legs to five. Use MCP calldata from MCP guide; sign from agent EOA only. Separate FORGE Merkle claims from USDC share price yield per FORGE emissions guide. Read Risks & Disclosures: unaudited contracts, ~$10 smoke TVL, variable rates, no principal protection.
External resources
Primary protocol documentation and data sources. Forge is not affiliated with these projects; links are for education only.
- Core vault — Basescan ↗Live ForgeVault contract on Base mainnet.
- Spark documentation ↗Spark protocol docs — savings rate mechanics.
- Morpho documentation ↗Morpho Blue and MetaMorpho vault architecture.
- Aave protocol ↗Multi-chain liquidity protocol — Base V3 deployment.
- Aave Base USDC rates ↗Live supply APY observation — variable, not promised.
- Morpho Steakhouse Prime rates ↗Curated vault rate snapshot on Base.
Related reading
- Spark Savings USDC — what's under the hoodHow Spark Savings USDC works on Base, why Forge routes Core vault USDC through it, and what risks agents should understand.
- Morpho on Base: curated vaults and Forge's Core bookHow Morpho Blue and Steakhouse Prime USDC work on Base, and how Forge's Core vault routes through them after P2-C0 expansion.
- Aave V3 USDC on BaseSupply-side Aave V3 USDC lending on Base and its role in the Forge Core vault adapter book.
Share this article
Draft copy for social posts — review before publishing. URL: https://forgetreasury.com/learn/conservative-strategy
Open Graph / preview card
Conservative strategy = Core vault on Base 100% USDC to Core — Spark, Morpho, Aave. Strategy not vault. Unaudited, variable yield. Verify adapters on /stats. https://forgetreasury.com/learn/conservative-strategy
Twitter / X
Conservative = a strategy, not a vault. MVP routes 100% USDC to Core on Base — Spark, Morpho, Aave. Variable yield, unaudited contracts: https://forgetreasury.com/learn/conservative-strategy
Forge Treasury separates vault buckets from agent strategies. Conservative routing sends USDC entirely to the Core vault — diversified stable yield via Spark, Morpho, and Aave on Base. Contracts are unaudited; rates are variable. Strategy explainer: https://forgetreasury.com/learn/conservative-strategy