Strategies · Macro and prediction markets
Fed-cut rotation
Kalshi’s traders put a price on what the Fed does at its next meeting. When they make a cut likely, long Treasuries and growth stocks stand to gain most from lower rates; when they make a hike likely, short Treasuries and banks hold up better. The market for the meeting often moves before the rest of the market has finished adjusting.
Pre-registered Sep 26, 2026, before it was tested. Results computed Sep 26, 2026.
Hypothetical backtest; not investment advice; paper trading only.Past results, real or simulated, do not predict future ones. ARCANE places no live orders from this page.
The rule
Trades every trading dayIf the market gives a cut at the next meeting more than 60%, hold long Treasuries (TLT) and the Nasdaq-100 (QQQ); if it gives a hike more than 60%, hold short Treasuries (SHY) and banks (XLF); otherwise the S&P 500. Checked daily.
Where the numbers came from. 60% means a clear majority, chosen before looking. The pairs are the textbook rate-sensitive and rate-insensitive holdings.
- Hold what is below
- If the Fed cut at the next meeting (Kalshi) is greater than 60
- then: Split equally across 2 blocks
- Hold TLT
- Hold QQQ
- Otherwise
- If the Fed hike at the next meeting (Kalshi) is greater than 60
- then: Split equally across 2 blocks
- Hold SHY
- Hold XLF
- Otherwise
- Hold SPY
As text
symphony "Fed-cut rotation"
description "Follow the prediction market's odds for the next Fed decision."
rebalance daily
benchmark SPY
weight equal
if ARCANE:kalshi/fed-cut-next > 60
weight equal
asset TLT
asset QQQ
else
if ARCANE:kalshi/fed-hike-next > 60
weight equal
asset SHY
asset XLF
else
asset SPY
How it would have done
Apr 11, 2023 to Sep 25, 2026, against SPYThe whole history
235 tradesA year
+15.5%
Annualized return
In all
+64%
Apr 11, 2023 to Sep 25, 2026
Sharpe
1.08
Return per unit of risk
Worst fall
−19%
Max drawdown
Swings
14%
Volatility, a year
Beta
0.92
To the benchmark
Each decision reads closing prices up to that day and trades at the next day’s close. It trades toward its targets every day. Costs: 5 bp of slippage on every trade and 0.3 bp of fees on sales. Uninvested cash earns nothing. Hypothetical: no real orders were placed.
- Growth of $1
- SPY, held
- Held-out period
What it held: SPY 73% of the time; QQQ + TLT 21% of the time; SHY + XLF 6% of the time.
In sample and the sealed holdout
The holdout was set before the preset was tested, and opened once| Measure | In sample Apr 11, 2023 – Sep 24, 2025 | Sealed holdout Sep 25, 2025 – Sep 25, 2026 |
|---|---|---|
| Annualized return | +16.3% | +13.4% |
| Sharpe ratio | 1.08 | 1.08 |
| Sortino ratio | 1.63 | 1.60 |
| Max drawdown | 19.0% | 10.2% |
| Calmar ratio | 0.86 | 1.31 |
| Volatility | 15.0% | 12.4% |
| Alpha against the benchmark | −4.2% | −2.7% |
| Beta | 0.92 | 0.93 |
| Correlation | 0.96 | 0.97 |
| Winning days | 56% | 52% |
| Turnover a year | 5.3× | 8.1× |
| Chance the Sharpe is above zero | 96% | 86% |
Is it real?
Checked on the in-sample window only- Overfitting: Likely overfit
- Trails its benchmark
The result depends on the exact numbers chosen, or does not beat what luck would produce.
- After counting 5 backtests of this idea, the out-of-sample Sharpe does not clear the bar luck would set (deflated Sharpe 0.80; 0.95 is the house bar).
- The best version in one half of history usually lands in the bottom half of the other (probability of overfitting 83%).
- Against SPY it earned −5.9% a year, clearly worse than just holding it.
| Backtests counted | 5 |
| Deflated Sharpe ratio (0.95 is the house bar) | 0.80 |
| Probability of backtest overfitting (20% is the house ceiling) | 83% |
| Walk-forward Sharpe (choosing, then testing) | 0.71 |
| Walk-forward Sharpe with costs doubled | 0.69 |
| Extra return a year against SPY | −5.9% |
The nudged versions (4)
| Change | Sharpe |
|---|---|
| threshold 60 on kalshi/fed-cut-next → 54 | 1.09 |
| threshold 60 on kalshi/fed-cut-next → 66 | 1.07 |
| threshold 60 on kalshi/fed-hike-next → 54 | 1.08 |
| threshold 60 on kalshi/fed-hike-next → 66 | 1.10 |
The house backtest gate
It does not pass. A strategy needs every one of these before ARCANE would paper-trade it.
| Check | Value | Needs | Result |
|---|---|---|---|
| Evaluation ran without problems | clean | no problems | Passed |
| Report belongs to this pre-registration | fed-cut-rotation | fed-cut-rotation | Passed |
| Evaluated only inside the pre-registered data window (holdout untouched) | 2023-04-11..2025-09-24 | within 2023-04-10..2025-09-24 | Passed |
| Walk-forward folds | 6.000 | >= 3 | Passed |
| Out-of-sample observations | 366.000 | >= 252 | Passed |
| Deflated Sharpe ratio (5 trials counted) | 0.804 | >= 0.95 | Failed |
| Probability of backtest overfitting | 0.829 | <= 0.2 | Failed |
| Out-of-sample Sharpe (annual, after costs) | 0.710 | >= 0.5 | Passed |
| Out-of-sample maximum drawdown | 0.190 | <= 0.25 | Passed |
| Sharpe with doubled costs | 0.688 | > 0 (needs a turnover series) | Passed |
| Sealed holdout: positive and not clearly worse than the backtest | mean 0.00053, consistent | mean > 0 and consistent | Passed |
- On 38 sessions between 2023-06-12 and 2025-03-18 its data had a gap (ARCANE:kalshi/fed-cut-next, ARCANE:kalshi/fed-hike-next), so it could not decide and kept what it held.
The pre-registration
Written before the backtest; changing any of it makes a new variant- Claim
- Rotating to TLT and QQQ when Kalshi prices a next-meeting cut above 60%, to SHY and XLF when it prices a hike above 60%, and holding SPY otherwise, beats SPY buy-and-hold on Sharpe after 5 bp costs over the sealed holdout.
- Why the edge should exist
- Event-contract prices aggregate views on the next Fed decision quickly; broad asset prices finish adjusting to a newly likely decision over days, not minutes.
- Where it comes from
- event market
- Who is on the other side
- Investors who rebalance on the Fed’s own communications rather than on the market-implied odds.
- Why it is not arbitraged away
- Fed-funds futures carry the same information, so the edge, if any, is small and depends on Kalshi leading; it is too small for large funds and may not exist.
- Universe
- Fed cut at the next meeting (Kalshi), Fed hike at the next meeting (Kalshi), QQQ, SHY, SPY, TLT, XLF
- In-sample window
- Apr 10, 2023 to Sep 24, 2025
- Sealed holdout
- Sep 25, 2025 to Sep 25, 2026
- Evaluation
- walk-forward, on sharpe, against SPY buy and hold; 13 variants planned
- Costs
- 5 bp slippage a trade, no commission
- Capacity
- $1,000,000: All four funds are among the most liquid in the US; the signal trades a handful of times a year.
- What stops it
- Paper Sharpe significantly below the backtest after 40 sessions. Kalshi delists the Fed decision contracts.
The data
Each decision reads closing prices up to that day and trades at the next day’s close. It trades toward its targets every day. Costs: 5 bp of slippage on every trade and 0.3 bp of fees on sales. Uninvested cash earns nothing. Hypothetical: no real orders were placed.
- Prices
- Yahoo Finance (unofficial chart API): adjusted daily closes for QQQ, SHY, SPY, TLT, XLF. Unofficial and not licensed for redistribution. A licensed vendor (Alpaca, Tiingo or Polygon) is a founder decision before the builder is sold.
- Fed cut at the next meeting (Kalshi)
- Kalshi, from Apr 8, 2023. History: point-in-time. Daily closing prices of Kalshi’s Fed decision contracts (FEDDECISION, then KXFEDDECISION) from April 2023. Each day reads the meeting still ahead and the price at the end of the day before, so a meeting’s result is never read on its own day.
- Fed hike at the next meeting (Kalshi)
- Kalshi, from Apr 8, 2023. History: point-in-time. As the cut series: Kalshi’s hike contracts from April 2023.