Adaptive Spread

Adaptive Spread is a rate-spread strategy that places a row of tiered funding offers around the market rate and changes how wide that spread is based on how choppy the rate has been. When the rate is bouncing around it spreads its offers wider so the high tiers can catch spikes; when things are calm it tightens back toward the market rate so offers fill quickly. It is an Advanced strategy in the Rate spread category and runs on a slice of your strategy’s capital, the same as any other bucket.

Backtests are modeled, not a guarantee. Past performance does not predict future results. You are lending to margin traders and bear the credit risk of Bitfinex’s collateral system; Stratum does not guarantee returns.

1. How it works

Adaptive Spread builds on the plain Rate Spread idea: instead of resting your whole slice at one price, it places several offers evenly spaced from below the market rate up to above it. Low offers fill fast; high offers sit and wait for the rate to come to them. What makes it “adaptive” is that the width of that ladder is not fixed — it moves with a simple measure of how volatile the rate has been.

  • Anchor. The ladder is centered on FRR — Bitfinex’s Flash Return Rate, the platform’s average funding rate that every Stratum strategy references.
  • Volatility measure. Stratum compares the current FRR to its recent average. The bigger the gap, the more “choppy” the market is treated as being. This measure is capped, so a single extreme tick can’t blow the spread out without limit.
  • Adaptive width. The half-width of the ladder starts at a calm-market baseline you set and grows in proportion to that volatility measure. Calm → a tight spread near FRR that fills quickly. Choppy → a wide spread whose top tiers sit out at higher rates to catch spikes.
  • Tiered offers. The slice is divided evenly across your chosen number of offers, spaced evenly from the bottom of the ladder to the top. Every offer rate is floored at zero.

In calm conditions the spread collapses back toward the baseline width, so Adaptive Spread behaves much like a plain rate spread around FRR. The adaptive behavior only really shows up when the rate starts moving.

2. When to use it

Adaptive Spread is built for choppy markets, where a fixed spread tends to be either too greedy (offers sit unfilled) or too timid (you leave rate on the table). If you want a ladder that automatically reaches further when the rate is volatile and pulls back in when it settles — without you re-tuning it — this is the strategy for that.

It sits one step above Rate Spread, which uses the same tiered-offer idea with a fixed spacing. If you want a simpler, fixed ladder, start with Rate Spread; choose Adaptive Spread when you specifically want the width to react to volatility.

3. Settings

  • Number of offers (slices) — how many offers to place across the ladder. Default 5; allowed range 2 to 20. More offers cover more of the rate range, but each one is smaller.
  • Spread width (baseSpreadPct) — how far above and below the market rate to place offers in calm conditions. Default 0.10, i.e. ±10%. Wider reaches for more but fills less often.
  • Reactiveness (volSensitivity) — how much the spread widens when the rate gets choppy. Default 2.0. Set it to 0 for a fixed-width ladder that ignores volatility; higher values react more to swings in the rate.

4. Tradeoffs and limits

  • One-sided. Adaptive Spread only ever lends — it never borrows — so it can’t work both sides of the book like a true market-maker. It is an honest, one-sided take on a dynamic-spread idea, not an optimal market-making model, and we don’t claim otherwise.
  • Collapses when calm. In quiet markets the spread tightens back toward the baseline and the strategy behaves close to a plain spread around FRR. The adaptive edge only appears when the rate is actually moving.
  • High tiers may not fill. As with any spread, the cheap offers earn less and the pricey ones may never fill if the rate doesn’t spike to meet them.
  • Credit risk. You are lending to margin traders on Bitfinex and bear the credit risk of Bitfinex’s collateral system. No funding strategy removes that risk, and Stratum does not guarantee returns.

5. Backtesting and going live

Every Stratum strategy can be backtested against real historical Bitfinex funding data before you commit capital, and the backtest runs the exact same algorithm code that runs live. Backtests are optimistic by design — they assume your offers fill at full size and ignore order-book depth and market impact — so treat any backtest figure as modeled, not a guaranteed or expected return. Past performance does not predict future results.

To try Adaptive Spread, connect a scoped Bitfinex API key (see Add your Bitfinex API key), start in paper mode to simulate against live rates, and switch to live when you’re ready. You can model outcomes first on the calculator, review the full strategy lineup on the Strategies page, and read how Stratum protects your keys on the Security page.