Count every cost in a round trip before you need the move

The short answer

A round trip is buy plus sell, and each side carries costs that are charged on different bases. Adding percentages together is usually wrong. Model each fee where it actually applies, then compute the gross move required to break even.

The cost inventory

• Venue or platform fee, charged on entry, exit, or both.
• Liquidity-provider fee, usually already inside the quoted output.
• Network base fee, paid in the chain's native asset.
• Priority fee or tip, paid to improve inclusion during congestion.
• Token transfer fee, charged by the token contract itself where present.
• Failed attempts, which can cost network fees without producing a position.
• On-ramp and withdrawal costs, if fiat is involved.
• Price impact, which is not a fee but changes the economics in the same direction.

A simplified model

Let C be the principal that buys token exposure, f_b the entry fee rate applied on C, f_s the exit fee rate applied to gross sale value, F total fixed costs, and g the gross appreciation:

net        = C × (1 + g) × (1 − f_s) − C × (1 + f_b) − F
breakEvenG = ((1 + f_b) + F / C) / (1 − f_s) − 1

Use decimal fractions inside the formula. This model assumes the entry fee is added on top of C. A platform that deducts its fee from your input requires a different calculation, so always state which model you are using.

Worked example

C = $100, f_b = 1%, f_s = 1%, F = $2.

breakEvenG = (1.01 + 0.02) / 0.99 − 1 = 1.040404… − 1 ≈ 4.04%

At a 10% gross appreciation:

net = 100 × 1.10 × 0.99 − 100 × 1.01 − 2 = 108.90 − 101 − 2 = $5.90

So a 10% move produces $5.90, not $10. And a 4% move — which feels like a win — is roughly break-even. On small principals the fixed component dominates: the same $2 of fixed cost against a $20 position requires about a 13% move before anything is earned.

Mistakes that distort the arithmetic

• Double counting. If a fee is already inside the quoted output, do not add it again.
• Treating price impact as an extra fee. It changes the fill, it is not a separate line added to it.
• Ignoring the native-asset balance. Running out of it strands a position you cannot sell.
• Forgetting failed attempts. Three failures before a successful entry are part of the trade's cost.
• Mixing currencies. If you record SOL costs and USD outcomes, preserve the conversion rate, its source and its timestamp.

What this cannot establish

This model estimates costs you know about. Unknown costs stay unknown — record them as unknown rather than zero. It also assumes you can exit at all; an unsellable position has a break-even of infinity, which is the real reason liquidity research precedes cost arithmetic.

Checklist

• Which fee model does this platform use: added to input, or deducted from it?
• Which costs are already inside the quote?
• What are the fixed costs, in total, for both sides?
• What gross move is break-even at my actual size?
• Do I hold enough native asset to exit as well as enter?