perf: avoid smart money OBV allocation

[stmr]

Summary

Avoid materializing a full OBV series inside the smart-money strategy. The strategy only needs to know whether OBV increased over the trailing obv_window, so this computes that trailing delta directly.

Type

• Bug fix
• New strategy
• New feature
• Refactor
• Documentation
• Tests
• Performance

Related Issue

Related to #2.

Changes

• Replace _obv_series(...) with _obv_trending_up(...) in app/strategies/quant_algos/smart_money.py.
• Avoid allocating one cumulative OBV entry per candle on every smart-money signal check.
• Preserve the old decision rule: obv[-1] > obv[-(obv_window + 1)] is equivalent to trailing OBV delta > 0.

Testing

• python3 -m py_compile app/strategies/quant_algos/smart_money.py
• git diff --check
• Local equivalence + benchmark script on a synthetic 20,000-candle cache:
  • OBV trend equality ok for lookbacks 1, 5, 20, 250, 1000
  • 5,000 helper calls: old 14.5807s, new 0.0043s, 3388.77x faster
  • 5,000 full smart_money(...) calls after patch: 0.0133s

Checklist

• No secrets, API keys, or private keys committed
• Type hints on all new public functions
• docs/connections/service-map.md updated (if adding a new call path) - not applicable
• docs/architecture/data-model.md updated (if changing DB schema) - not applicable
• New strategy registered in app/strategies/signals.py (if adding a strategy) - not applicable
• New strategy added to strategy index in app/onchain_logger.py (if adding a strategy) - not applicable
• Breaking changes documented in PR description - not applicable

If selected for payout: USDC 0x10DFE6771131BEc830A7a44D7Be45Ced0741b2b3.

[vercel]

@stmr is attempting to deploy a commit to the dragoon4002's projects Team on Vercel.

A member of the Team first needs to authorize it.