Week Ahead: Funding Rate Predictions
The new model run is in.
507 pairs cleared the prefilter.
Today's export surfaced only the top 15 rows, and that sample needs cleaning before it is useful.
There are duplicate rows.
There are also several degenerate fits with persistence above 1 and half-lives of 99,999 hours.
Those are not forecasts.
They are ceiling-rate artifacts.
Strip those out and the week ahead becomes clearer.
The cleaner signal still lives in a small cluster of Hyperliquid pairs with finite half-lives and strong R².
The Usable Model Set
The most credible rows in today's sample are the ones with finite half-lives and persistence below 1.
| Exchange | Symbol | Persistence (β) | Equilibrium (%) | Positive % | R² | Half-Life (h) |
|---|---|---|---|---|---|---|
| Hyperliquid | BABYUSDT | 0.7049 | -0.0030 | 50.0 | 0.962 | 15.8 |
| Hyperliquid | DOTUSDT | 0.9564 | -0.0035 | 33.9 | 0.911 | 124.4 |
| Hyperliquid | VIRTUALUSDT | 0.9259 | -0.0002 | 65.6 | 0.898 | 72.1 |
| Hyperliquid | KPEPEUSDT | 0.9203 | -0.0013 | 32.8 | 0.884 | 66.8 |
These are the rows worth carrying into the new week.
They have sensible persistence values.
They have finite half-lives.
And they fit the same structural story that has been building in Settled's recent data.
The aggregate market is still negative, but not collapsing.
That makes mean-reversion quality more important than raw extremity.
BABYUSDT Is Still The Cleanest Setup
BABYUSDT remains the standout.
Its persistence is 0.7049.
Its equilibrium sits at -0.003%.
Its R² is 0.962.
And its half-life is just 15.8 hours.
That is still the sharpest combination in the file.
A half-life under 16 hours means deviations should resolve quickly if the structure holds.
The 50% positive settlement rate is also useful.
This is not a one-way carry trade.
It oscillates around a mildly negative center.
That is exactly the profile to watch when the broader regime is stabilizing rather than breaking.
For Settled users browsing prediction markets, BABYUSDT is the cleanest candidate for short-horizon rate normalization.
DOTUSDT And KPEPEUSDT Keep The Negative Drift Theme Alive
DOTUSDT is much slower.
Half-life is 124.4 hours.
Positive settlement share is only 33.9%.
Equilibrium is -0.0035%.
That says the negative bias is persistent, but the path back to equilibrium is measured in days, not sessions.
KPEPEUSDT tells a similar story with a shorter half-life of 66.8 hours and an even weaker positive share at 32.8%.
Neither pair looks like a fast snapback trade.
They look like week-long regime expressions.
If the market keeps healing, these are the kinds of names where the shift will show up late.
If the market slips back into stress, they should confirm it quickly.
VIRTUALUSDT Sits Near Flat
VIRTUALUSDT is the most balanced row in the group.
Equilibrium is almost zero at -0.0002%.
Positive settlement share is 65.6%.
Half-life is 72.1 hours.
That is a different read from DOT or KPEPE.
This is not a structurally bearish book.
It is a mostly positive tape that still mean-reverts toward flat.
In a week where the aggregate regime has already become less negative, this kind of near-equilibrium pair matters.
If rates drift lower and recover quickly, the model holds.
If they begin to sustain negative prints below equilibrium, the regime is worsening again.
What To Ignore In This Export
A few rows should not be traded off at all.
Binance DMCUSDT printed persistence of 78.1255, R² of 1.000, and a 99,999-hour half-life.
42USDT printed persistence of 1.3525 with a 99,999-hour half-life.
AVLUSDT also came through above 1.0 persistence with the same half-life cap.
Those are not investable forecasts.
They are usually capped or pathologically short histories that make the model look more certain than it is.
The same caution applies to YALAUSDT.
Its equilibrium is listed at 1.1583% with an 87.1% positive rate and a 419.1-hour half-life.
That may be real, but it is extreme enough to treat as a watchlist item, not a baseline expectation.
Read Through For The Week Ahead
The broad market regime improved from Apr 08 through Apr 10.
That context matters.
When the aggregate tape is still negative but moving closer to zero, the best setups usually come from pairs with stable reversion parameters, not from the loudest headline rates.
So the week ahead screen is simple.
Watch BABYUSDT for fast reversion.
Watch VIRTUALUSDT for confirmation that near-flat books stay near flat.
Watch DOTUSDT and KPEPEUSDT to see whether persistent negative-drift names keep dragging or start to normalize.
If those four pairs hold their structure while the aggregate market continues to stabilize, the current negative regime is maturing toward exhaustion.
If they all lean more negative at once, last week's improvement was only a pause.
That is the real forecast.
Not a single token call.
A dispersion test.
And funding markets are usually most tradeable when that dispersion starts to compress.
Related reading:
The research arm of Zirodelta. Data-driven analysis of crypto sentiment markets, model development, and market microstructure research. Data-driven. Real-time. Across 6 exchanges and 3,700+ perpetual futures.
Learn more about Zirodelta →