Trend Follower

Directional strategy for confirmed momentum.

Best Used When

trend is strong and persistent
structure supports continuation
liquidity is normal
Hermes wants to follow momentum instead of fading it

Avoid When

price is range-bound
trend strength is weak or noisy
event risk is high
the market is too chaotic for clean continuation

How It Works

The strategy acts on the configured side, not on an internal trend model.
trade_side controls whether it can buy, sell, or accept either side. both is symmetric and depends on Hermes or the operator for direction.
balance_target is a wallet-allocation target, not a position-size target.
For buy, sub-1.0 values stop the strategy once the base share of wallet value reaches the target.
For sell, sub-1.0 values stop the strategy once the quote share reaches the target.
When an order would overshoot a sub-1.0 target, the last order is clamped to the remaining gap.
entry_offset_pct sets the limit price offset for new entries.
cooldown_ticks pauses the next few ticks after a successful order.

Parameters

trade_side: Allowed side selection. buy and sell make it one-sided; both is symmetric and relies on Hermes for direction.
balance_target: Allocation target between 0.0 and 1.0. 1.0 means keep trading until the funding side is exhausted. Lower values stop once the wallet reaches the requested mix.
entry_offset_pct: Limit-price offset used for new entries. Buy orders are placed above the last price, sell orders below it.
exit_offset_pct: Reserved for exit or reversal pricing in policy and supervision output. It is part of the contract even though the current entry path only uses entry_offset_pct.
order_notional_quote: Target quote notional per order.
max_order_notional_quote: Optional hard cap on quote notional per order.
dust_collect: Lets the shared sizing helper consume leftover size more aggressively when a venue minimum would otherwise strand a small remainder.
cooldown_ticks: Number of ticks to wait after a successful order before the strategy can act again.
debug_orders: Enables order-placement debug logging.

Hermes Policy Mapping

Hermes controls whether the side is active.
Hermes may set balance_target to define how far the wallet should rotate.
Hermes may set the quote notional and offsets.
apply_policy() currently records the derived values but does not rewrite the config the way the grid and exposure strategies do.

Notes

Hermes decides when trend following is allowed.
Trader maps policy to concrete order behavior.
The strategy reports side, quote notional, and policy-derived settings.
trade_side lets Hermes or the operator run a long-only, short-only, or symmetric directional instance.
balance_target=1.0 means keep trading until no more usable size remains on the funding side.
In buy mode, values below 1.0 target the base share of wallet value. Example: 0.5 stops near a 50/50 base-quote split.
In sell mode, values below 1.0 target the quote share of wallet value. Example: 0.5 stops near a 50/50 quote-base split.
When an order would overshoot a sub-1.0 balance_target, the final order is clamped to the remaining target gap instead of sending the full configured notional.
live fee rates are used directly, so the strategy does not need a configured fee fallback.

Useful Example

{
  "trade_side": "buy",
  "balance_target": 0.8,
  "entry_offset_pct": 0.0035,
  "exit_offset_pct": 0.002,
  "order_notional_quote": 25,
  "max_order_notional_quote": 40,
  "dust_collect": true,
  "cooldown_ticks": 3,
  "debug_orders": false
}

This is a practical long-only accumulation setup. It keeps buying until the wallet is roughly 80 percent base by value, uses a modest entry offset, and avoids oversized last orders.

trend_follower.md 3.8 KB 文件歷史 原始文件