peer-rate-optimizer

name	peer-rate-optimizer
description	Optimize ZKP2P vault and LP deposit rates using market intelligence and PnL feedback. Analyzes spreads, volume, and performance data to recommend or execute rate adjustments. Use when the user wants to optimize rates, improve LP returns, or automate rate management strategy.

ZKP2P Rate Optimizer

Closed-loop rate optimization for ZKP2P vaults and LP deposits. Uses Peerlytics market intelligence to inform rate decisions, executes adjustments on-chain, and monitors PnL outcomes via the indexer.

Overview

The optimization loop has three stages:

1. DATA COLLECTION
   Peerlytics API → market spreads, competitor rates, volume by pair
   Indexer GraphQL → ManagerAggregateStats (vault PnL), ManagerStats (per-intent)

2. RATE COMPUTATION
   For each (paymentMethod, currency) pair:
     analyze PnL, volume, market share, competitor spread
     compute target rate using algorithm below

3. EXECUTION
   setMinRate() or setMinRatesBatch() on DepositRateManagerRegistryV1
   Monitor outcomes in next iteration

Data Collection

Before adjusting rates, gather these data points:

Market Data (Peerlytics)

Query the Peerlytics orderbook for current market state:

// Install: npm install @peerlytics/sdk
// Access: API key or x402 pay-per-request (USDC on Base)

// Orderbook data available at: orderbook.peerlytics.xyz
// Key metrics to collect:
// - Best available rates per (platform, currency) pair
// - Total liquidity depth per pair
// - 24h/7d volume per pair
// - Number of active LPs per pair
// - Spread distribution (min, median, max rates)

Vault Performance (Indexer)

Query the staging indexer at https://indexer.hyperindex.xyz/00be13d/v1/graphql:

# Vault aggregate stats — overall PnL and volume
query VaultPerformance($rateManagerId: String!) {
  ManagerAggregateStats(where: { rateManagerId: { _eq: $rateManagerId } }) {
    totalFilledVolume
    totalFeeAmount
    totalPnlUsdCents
    fulfilledIntents
    currentDelegatedBalance
    currentDelegatedDeposits
    updatedAt
  }
}

# Per-intent stats — granular performance data
query IntentPerformance($rateManagerId: String!, $since: BigInt!) {
  ManagerStats(
    where: {
      rateManagerId: { _eq: $rateManagerId }
      createdAt: { _gte: $since }
    }
    order_by: { createdAt: desc }
  ) {
    intentId
    depositId
    amount
    quoteConversionRate
    marketRate
    spreadBps
    pnlUsdCents
    managerFee
    managerFeeAmount
    createdAt
  }
}

# Current vault rates
query VaultRates($rateManagerId: String!) {
  RateManagerRate(where: { rateManagerId: { _eq: $rateManagerId } }) {
    paymentMethodHash
    currencyCode
    managerRate
    updatedAt
  }
}

Competitor Rates (Indexer)

# All active deposits with their rates (competitor analysis)
query CompetitorRates($paymentMethodHash: String!, $currencyCode: String!) {
  MethodCurrency(
    where: {
      paymentMethodHash: { _eq: $paymentMethodHash }
      currencyCode: { _eq: $currencyCode }
      conversionRate: { _gt: "0" }
    }
    order_by: { conversionRate: asc }
  ) {
    depositId
    conversionRate
    managerRate
    rateManagerId
  }
}

Optimization Algorithm

Core Logic (Pseudocode)

For each (payment_method, currency) pair managed by this vault:

  1. Compute current metrics:
     - current_rate = vault's managerRate for this pair
     - pnl_7d = sum of pnlUsdCents for intents in last 7 days
     - volume_7d = sum of amounts for intents in last 7 days
     - fill_count_7d = count of fulfilled intents in last 7 days
     - market_best_rate = lowest competitor rate for this pair
     - market_median_rate = median competitor rate for this pair

  2. Apply rules:
     IF pnl_7d < 0:
       → WIDEN spread by 20 bps (increase rate by 0.002 * 1e18)
       → Rationale: losing money, need better margin

     ELSE IF fill_count_7d == 0 AND current_rate > 0:
       IF days_since_last_fill > 7:
         → DISABLE pair (set rate to 0)
         → Rationale: no demand, stop quoting
       ELSE:
         → TIGHTEN spread by 15 bps (decrease rate by 0.0015 * 1e18)
         → Rationale: may be priced too high

     ELSE IF volume_7d > 0 AND current_rate > market_median_rate * 1.05:
       → TIGHTEN spread by 10 bps
       → Rationale: overpriced vs market, can capture more volume

     ELSE IF volume_7d > 0 AND current_rate < market_best_rate:
       → WIDEN spread by 5 bps
       → Rationale: underpriced, leaving money on the table

     ELSE:
       → HOLD steady
       → Rationale: performing within acceptable range

  3. Apply safety constraints:
     - new_rate >= MINIMUM_SPREAD_FLOOR (default: 1.001 * 1e18 = 10 bps)
     - new_rate <= MAXIMUM_SPREAD_CAP (default: 1.10 * 1e18 = 10%)
     - abs(new_rate - current_rate) <= MAX_CHANGE_PER_ITERATION (default: 50 bps)
     - If change > 30 bps, log WARNING for review

TypeScript Implementation

import { createWalletClient, http, parseAbi } from 'viem';
import { base } from 'viem/chains';
import { privateKeyToAccount } from 'viem/accounts';

// --- Constants ---
const PRECISE_UNIT = 1_000_000_000_000_000_000n; // 1e18
const BPS_UNIT = PRECISE_UNIT / 10_000n;         // 1 bps in 1e18
const MIN_SPREAD_FLOOR = PRECISE_UNIT + (10n * BPS_UNIT);   // 1.001 (10 bps)
const MAX_SPREAD_CAP = PRECISE_UNIT + (1000n * BPS_UNIT);   // 1.10  (10%)
const MAX_CHANGE_PER_ITER = 50n * BPS_UNIT;                 // 50 bps

const REGISTRY_ADDRESS = '0x3125F621482887d158cb51cE9b54D9D25b145877';
const INDEXER_URL = 'https://indexer.hyperindex.xyz/00be13d/v1/graphql';

// --- Types ---
interface PairMetrics {
  paymentMethodHash: string;
  currencyCode: ;
  : ;
  : ;
  : ;
  : ;
  : ;
  : ;
  : ;
}

  {
  : ;
  : ;
  : ;
  : ;
  : ;
  : ;
}


 ():  {
   newRate = metrics.;
   reason = ;

   (metrics. < ) {
    
    newRate = metrics. + ( * );
    reason = ;

  }   (metrics. ===  && metrics. > ) {
     (metrics. > ) {
      newRate = ;
      reason = ;
    }  {
      newRate = metrics. - ( * );
      reason = ;
    }

  }   (
    metrics. >  &&
    metrics. > (metrics. *  / )
  ) {
    newRate = metrics. - ( * );
    reason = ;

  }   (
    metrics. >  &&
    metrics. < metrics.
  ) {
    newRate = metrics. + ( * );
    reason = ;
  }

  
   (newRate !== ) {
     (newRate < ) newRate = ;
     (newRate > ) newRate = ;

     delta = newRate > metrics.
      ? newRate - metrics.
      : metrics. - newRate;
     (delta > ) {
      newRate = newRate > metrics.
        ? metrics. + 
        : metrics. - ;
      reason += ;
    }
  }

   changeBps = (
    ((newRate - metrics.) * ) / 
  );

   {
    : metrics.,
    : metrics.,
    : metrics.,
    newRate,
    changeBps,
    reason,
  };
}


  () {
   changed = adjustments.( a. !== a.);
   (changed. === ) {
    .();
    ;
  }

   account = (privateKey);
   client = ({
    account,
    : base,
    : (),
  });

   registryAbi = ([
    ,
    ,
  ]);

   (changed. === ) {
    
     adj = changed[];
    .();
     hash =  client.({
      : ,
      : registryAbi,
      : ,
      : [
        rateManagerId  ,
        adj.  ,
        adj.  ,
        adj.,
      ],
    });
    .();
  }  {
    
    .();
     paymentMethods = [changed.( a.  )];
     currencies = [changed.( a.  )];
     rates = [changed.( a.)];

     hash =  client.({
      : ,
      : registryAbi,
      : ,
      : [
        rateManagerId  ,
        paymentMethods,
        currencies,
        rates,
      ],
    });
    .();
  }
}

Implementation — Full Optimization Loop

async function runOptimizationLoop(
  rateManagerId: string,
  privateKey: `0x${string}`
) {
  console.log(`=== Rate Optimization: ${new Date().toISOString()} ===`);

  // Step 1: Fetch current vault rates from indexer
  const currentRates = await queryIndexer(`{
    RateManagerRate(where: { rateManagerId: { _eq: "${rateManagerId}" } }) {
      paymentMethodHash, currencyCode, managerRate, updatedAt
    }
  }`);

  // Step 2: Fetch vault performance (last 7 days)
  const sevenDaysAgo = BigInt(Math.floor(Date.now() / 1000) - 7 * 86400);
  const performance = await queryIndexer(`{
    ManagerStats(
      where: {
        rateManagerId: { _eq: "${rateManagerId}" }
        createdAt: { _gte: "${sevenDaysAgo}" }
      }
    ) {
      amount, spreadBps, pnlUsdCents, createdAt,
      quoteConversionRate, marketRate
    }
  }`);

  // Step 3: Fetch aggregate stats
  const aggregateStats = await queryIndexer(`{
    ManagerAggregateStats(where: { rateManagerId: { _eq: "${rateManagerId}" } }) {
      totalFilledVolume, totalFeeAmount, totalPnlUsdCents,
      fulfilledIntents, currentDelegatedBalance
    }
  }`);

  
   : [] = [];
   ( rate  currentRates.) {
     competitors =  ();

     competitorRates = competitors..(
       (c.)
    );
     marketBestRate = competitorRates. > 
      ? competitorRates[]
      : (rate.);
     marketMedianRate = competitorRates. > 
      ? competitorRates[.(competitorRates. / )]
      : (rate.);

    
     pairIntents = performance..(
        
    );
     pnl7d = pairIntents.(
       sum + (s.), 
    );
     volume7d = pairIntents.(
       sum + (s.), 
    );

     :  = {
      : rate.,
      : rate.,
      : (rate.),
      pnl7d,
      volume7d,
      : pairIntents.,
      : (pairIntents),
      marketBestRate,
      marketMedianRate,
    };

    adjustments.((metrics));
  }

  
  .();
   ( adj  adjustments) {
     status = adj. === adj. ?  : ;
    .();
    .();
    .();
  }

  
   (rateManagerId, adjustments, privateKey);
}


  (): <> {
   resp =  (, {
    : ,
    : { :  },
    : .({ query }),
  });
   json =  resp.();
   json.;
}

 ():  {
   (intents. === )  ;
   latest = .(...intents.( (i.)));
   .((.() /  - latest) / );
}

Monitoring

After each optimization run, track these metrics over time:

Metric	Source	Purpose
totalPnlUsdCents	ManagerAggregateStats	Is the vault profitable?
totalFeeAmount	ManagerAggregateStats	Revenue from management fees
fulfilledIntents	ManagerAggregateStats	Volume throughput
spreadBps per intent	ManagerStats	Are spreads converging to optimal?
currentDelegatedBalance	ManagerAggregateStats	Is liquidity growing or shrinking?

Compare these week-over-week to assess optimization effectiveness.

Safety Guardrails

1. Maximum change per iteration: 50 bps. Prevents catastrophic mispricing from a single bad data point.
2. Minimum spread floor: 10 bps (1.001x). Never set rates below cost (LP must always earn something).
3. Maximum spread cap: 10% (1.10x). Prevents absurd rates that would never fill.
4. Disable threshold: Only disable a pair (rate = 0) after 7+ days of zero fills. Prevents premature shutdown.
5. Warning threshold: Log a WARNING for any change exceeding 30 bps. Enables human review of large adjustments.
6. Dry run mode: Always log recommendations before executing. Add a --dry-run flag to scripts.

Scheduling

Run the optimization loop on a regular cadence:

# Run every 15 minutes (recommended for active vaults)
*/15 * * * * cd /path/to/skills && python3 scripts/optimize.py VAULT_ID

# Run every hour (for low-volume vaults)
0 * * * * cd /path/to/skills && python3 scripts/optimize.py VAULT_ID

# Run every 6 hours (for stable, set-and-forget vaults)
0 */6 * * * cd /path/to/skills && python3 scripts/optimize.py VAULT_ID

Adjust frequency based on volume:

• High volume (>$10K/day): every 15 minutes
• Medium volume ($1K-$10K/day): every hour
• Low volume (<$1K/day): every 6 hours

Contract Reference

Contract	Address	Chain
DepositRateManagerRegistryV1	0x3125F621482887d158cb51cE9b54D9D25b145877	Base (staging)
DepositRateManagerController	0x2CF2FA7F21be0F920E1D8f4bb9C08E2c07F0E5d7	Base (staging)
Indexer	https://indexer.hyperindex.xyz/00be13d/v1/graphql	-
Peerlytics Orderbook	https://orderbook.peerlytics.xyz	-

Rate Encoding Reference

Rates use 18-decimal fixed-point (PRECISE_UNIT = 1e18):

Scenario	Encoded Value	BPS Markup
1:1 (no markup)	1000000000000000000	0 bps
10 bps markup	1001000000000000000	10 bps
50 bps markup	1005000000000000000	50 bps
2% markup	1020000000000000000	200 bps
5% markup	1050000000000000000	500 bps
Disabled	0	N/A