What the two lines actually measure
The lower panel of the Spread History chart shows two stepped lines — a green one and a red one. Each line represents one of the two directional trades you can take on this pair, expressed as a percentage. The green line is your LongA spread: how much you would lock in by buying the instrument on exchange A at the ask and selling the same instrument on exchange B at the bid, at the moment you click Open. The canonical formula is (bid_B − ask_A) / ask_A × 100%.
The red line is the ShortA spread, the mirror operation — sell the instrument on exchange A at the bid and buy the same instrument on exchange B at the ask. Canonical formula: (bid_A − ask_B) / ask_B × 100%. But this is where the chart's convention surprises new users. The red line is plotted with the sign flipped, so what you see on screen is −ShortA rather than ShortA itself. The reasons for this deliberate inversion are spelled out in the next section.
Each tick is rendered as a flat horizontal segment (step line) because the underlying spread engine evaluates the order books at discrete moments, not as a smooth continuous curve. The visible vertical distance between the two lines is the round-trip bid-ask cost — what you pay just for crossing both stacks with market orders. Wide gap = illiquid pair with deep slippage built in; narrow gap = tight market where you give back almost nothing on entry and exit.
Why the red line is inverted
Flipping the sign of a line looks like a quirk at first. It is in fact a deliberate design choice — four independent benefits stack on top of each other, and once you trust them, reverting to the raw convention feels like reading a map upside down. One card per benefit below.
See it: inverted versus native
The widget below replays the same 100-tick scenario in both conventions. A signal forms, the spread widens to roughly 1%, a LongA trade opens at the Open threshold, the spread holds wide for a while, then reverts and briefly inverts. The trade closes profitably at the Close threshold. Toggle between Inverted and Native to see how the same physical events render under each convention.
In Inverted mode the Open marker sits in the upper half and the Close marker sits in the lower half — the trade reads top-to-bottom as one coherent path. In Native mode both threshold lines are forced above zero and the red line mirrors the green like scissor blades. The close event still happens, but it visually competes with the open event. After a few minutes of pattern-recognition with the inverted view, the native view feels actively worse — every chart in Arbitron uses the inverted convention precisely so this becomes muscle memory.
Backtest thresholds: the trading channel
When you run a backtest on the chart, two horizontal dashed lines appear: a green one and a red one. They are not arbitrary visual decorations — each is a price level the engine has decided is the optimal trigger for entering or exiting a trade over the loaded history window. The numbers next to them in the backtest panel (Open, Close, Cycles, Profit) reflect what the engine simulated using those exact thresholds.
The green dashed line is the Open threshold: the level at which the green LongA spread, climbing from below, would trigger a new position. Wider threshold = more conservative entry, fewer cycles but each one locks in more spread. Tighter threshold = more aggressive, more cycles but each pays less. The backtest searches the grid of plausible (open, close) pairs and reports the combination with the highest total profit after fees over the window. The mode pills (LongA, ShortA, Both) re-run the grid search for whichever direction you pick.
The red dashed line is the Close threshold, and it sits below zero in the inverted convention. The red −ShortA line falling through it is the same physical event as the canonical ShortA spread rising through +closeTh: the price gap has compressed (or reversed) enough that closing the position now locks in the cycle's profit. The fact that the two thresholds sit on opposite sides of zero gives the chart its characteristic horizontal channel, and it lets you eyeball whether the current spread is pushing toward Open, drifting in the no-trade middle, or approaching Close — without reading numbers.
Trade markers: how Open and Close events appear
On top of the spread lines, completed cycles drop two markers per trade — an Open marker anchored exactly on the green dashed threshold line, and a Close marker anchored exactly on the red dashed threshold line. Each marker is pinned to its trigger threshold, not to the live spread value at the moment of trigger. This is intentional: real fills happen at the threshold or slightly worse due to network latency, and pinning markers to the threshold prevents visual drift when you compare backtest runs.
Markers are rendered as outlined letters — B for Buy and S for Sell — with a contrast border so they remain readable against either the green or the red lines, in either light or dark theme. Hover the chart's crosshair to read exact timestamps and spread values for any tick. Use the Trades checkbox below the chart to toggle marker visibility when you want to compare backtest results across multiple threshold settings without the visual clutter of execution markers.
When backtest mode is Both, the same chart shows two kinds of cycles interleaved — LongA cycles open in the upper half and close in the lower half, ShortA cycles do the opposite. Inversion is what makes this readable: each cycle still flows visually from one extreme to the other regardless of direction, instead of crowding into one half-plane. If you ever find yourself confused by which markers belong to which cycle, switch the mode pill to a single direction; the chart re-renders with only that direction's cycles.
Cheat sheet
Green above zero means LongA is profitable right now — the wider above zero, the bigger the spread you would lock in by opening a LongA position at the current quotes.
Red below zero means ShortA is profitable right now (the line is plotted as minus-ShortA, so deep below zero corresponds to a large positive ShortA spread). Closing a LongA position when red dips deep below zero is the most profitable exit moment, because price has flipped in your favor.
Both lines together up high (and staying there) means strong one-sided dominance — only LongA is tradable on this pair, ShortA would lose roughly the same amount per cycle. This pattern is common for tokens that consistently trade at a premium on one venue.
Both lines hugging zero means no edge — the spread has already been arbitraged away by faster bots and there is nothing left to harvest on this pair right now. Move on.
Lines crossing each other repeatedly means the canonical spread is changing sign — each crossing is a potential turning point and often marks where the bot crowd either piles in or steps back. A pair with many crossings tends to produce many trade cycles; a pair with no crossings produces at most one trend-following cycle and then stalls.