Infolayer · a lead, not a finding

Exit on a target, or on opportunity cost?

A crypto trade dies. When should you have closed it — at a fixed profit target, or the moment it stopped being your best available bet? Foraging ecology answers the second question. Move the shape and see what it says.

What this is. A conjecture, not a result. It transfers a proven rule from behavioural ecology onto trade exits — and it has not been vigorously pursued: it is theoretically scoped and empirically untested at the current sample (n≈8 closed trades). It is here because it is worth pursuing and worth holding while it waits. Read the honest edges at the bottom before you trust any of it.

The rule, borrowed whole

A bird works a patch of berries. The patch depletes as it eats, so its intake rate falls. When should it leave for the next patch? Charnov's Marginal Value Theorem (1976) gives the optimum, and it's been confirmed in real animals: leave the moment this patch's intake rate drops to the average rate of the whole habitat — net of the travel time to the next one.

A trade is a patch. You enter, you extract value while the edge lasts, and the whole game is when to leave and redeploy. So the transfer writes itself:

Foraging	Trading
A patch	An open position
Intake rate, falling as the patch depletes	The position's edge, decaying as the thesis plays out
Travel time to the next patch	Cost + slippage to find and enter the next setup
Average rate across the whole habitat	Your best available alternative right now (the dashed line)
Leave when intake drops to the habitat average	Exit when the edge drops to your opportunity rate

That's the crossing you were just moving. And two predictions fall straight out of the geometry — both a little counterintuitive, which is the interesting part:

Barren regime → hold winners longer. When nothing else is worth doing, a fading position still beats the alternatives, so the line sits low and the crossing slides right. The instinct to "take profits" in a dead tape is, on this reading, exactly backwards.

Illiquid asset → hold longer too. Higher cost to re-enter is a longer "travel time," which the theorem says should push you to stay in each patch longer. Exit-timing should vary by asset class — a lever most fixed-target rules don't have.

Where it breaks — and the shape shows you

The theorem needs a decaying patch. Toggle the trade to accelerating and watch: a breakout's edge rises, so "leave when the rate falls to your alternative" fires at the very start — it tells you to cut the winner at its worst possible moment, right before it pays. So the honest scope is archetype-conditional: an opportunity-cost exit fits mean-reversion and carry (edges that deplete); a trailing exit still fits trend and breakout (edges that accelerate). The trader panel already labels which archetype a trade is — so the switch is knowable at entry.

The honest edges (kept in view, not smoothed)

This is a lead precisely because these aren't closed:

Untested. The bench is ~8 closed trades — far too few to backtest an exit rule. This is a hypothesis with a designed test, not a validated method.
The habitat rate is estimated, not measured. A forager feels the habitat average from experience; a trader has to guess it (from how many setups the panel is flagging). The rule is only as clean as that proxy.
Carry is a hybrid. The funding payment is roughly constant income; only the convergence leg depletes. The theorem fits the spread leg cleanly, the income leg less so.
"Opportunity-cost exits" isn't itself new — cost-of-capital is old. What's specific here is the leaving condition and the two counterintuitive regime predictions above.

Two more seeds from the same run

The same afternoon threw off two more cross-domain leads, un-chased, parked here to hold:

TCP congestion control → position sizing. The internet solved "probe for capacity without causing collapse" with additive-increase / multiplicative-decrease. Ported to exposure: grow the position additively while the regime cooperates, cut it multiplicatively on an adverse shock — a provably cascade-avoiding, anti-martingale sizing rule.
Ecological "undermatching" → why funding extremes persist. Animals systematically under-respond to payoff gradients, leaving a durable edge for whoever does respond. That's a first-principles reason funding extremes should persist longer than efficient-market intuition predicts — which is exactly what the carry lens harvests.