Track record

Every year on the record. Including the one that hurt.

Out-of-sample results for the Kairos regime model, 2020 to mid-2026. The rules were locked before the window and never changed to fit the data. Hypothetical model performance: not audited, not client accounts.

Annualized return. S&P 500: 15.5%.
40.6%
Sharpe ratio. S&P 500: 0.81.
1.39
Worst peak-to-trough drop. S&P 500: −33.7%.
−24.5%
What $100,000 became. S&P 500: $252K.
$896K

Kairos regime model vs SPY, out-of-sample, 2020 through mid-2026. Source: Kairos strategy deck.

Growth of $100,000

Same start date, different paths. The flat stretches are safety mode: the model holding T-bills while conditions were fragile.

Growth of $100,000 out-of-sample 2020–2026: Kairos versus S&P 500 (SPY). Kairos ends at $896K; SPY at $252K. Pink band marks the 2022 bear market.
Year by year

Six years ahead, one behind. Tap a year to see both numbers, or switch to the table.

Calendar-year returns

Kairos regime model vs S&P 500, out-of-sample

  • Kairos
  • S&P 500 (SPY)
  • 2020
    +72%
    +18%
  • 2021
    +59%
    +29%
  • 2022
    +8%
    −18%
  • 2023
    +95%
    +26%
  • 2024
    +26%
    +25%
  • 2025
    +3%
    +18%
  • 2026 YTD
    +20%
    +9%

Source: Kairos strategy deck. S&P 500 annual figures are read from the deck chart and rounded. Out-of-sample model results, not audited or client-account performance.

2022

The year it was built for.

Stocks and bonds fell together. The S&P 500 finished down about 18%. The model finished up 8%, not because it predicted anything, but because the five checks read fragile and the written rule said step aside.

2025

The year it cost us.

The S&P 500 made about 18%; the model made 3%. The same defenses that earned 2022 kept it cautious through what the source deck calls a steady no-pullback grind. We’d rather show you that year than hide it.

The year it lagged
The risk side of the ledger

What it would have felt like to hold through it.

−24.5% worst drop

On $100,000, a fall to about $75,500 before recovering. The S&P 500’s worst in the same window was −33.7%.

59% of days in the market

The other 41% were spent in T-bills. Being out is a position, not an absence.

About 11 trades a year

One decision a day, but most days the decision is to change nothing.

72% of trades won

Which means 28% did not. The rule is built to keep those losers small, not to avoid them.
Read this before the numbers

What this record is, and what it isn’t.

Out-of-sampleIt is
The rules were finalized before the 2020–2026 window and not tuned on it.
CostedIt is
Near-close execution, commissions, and slippage are modeled.
CompleteIt is
Every calendar year is shown, including 2025, when it lagged the S&P 500 by about 15 points.
Measured in real T-billsIt is
Safety mode is measured in short-term T-bills (SHV as the proxy), not an assumed zero-risk return.
Not auditedIt isn’t
No third party has verified this series.
Not client accountsIt isn’t
Your own results will differ with timing, fees, taxes, and sizing.
Not a promiseIt isn’t
A defensive rule that worked in 2022 can be late or wrong in the next drawdown.
Backtest, out-of-sample, live: the difference

How the record was built.

Strategy
Kairos regime model. Growth mode: 83% TQQQ (about 2.5× effective Nasdaq exposure). Safety mode: 100% short-term U.S. Treasury bills.
Window
January 2020 through mid-2026. Out-of-sample: rules finalized before the window and never refit on it.
Decision
Five-factor regime score computed from public market data each trading day at 3:50 PM ET.
Execution
Near-close, with commissions and slippage modeled. SHV used as the T-bill proxy.
Benchmark
SPY (S&P 500 ETF), same start date.
Not included
Kairos fees, taxes, and differences in individual account timing or sizing.
Newer systems
Kairos Select and Kairos Enhanced have been tracked live as models since 30 September 2025: about a year, gross of fees and slippage, not audited. Too short to publish as a record; we walk through them on a call with that caveat said out loud.

Keep reading

Walk through these numbers with the person who built them.

Bring the questions a skeptic should ask: why 2025 lagged, what the worst drop felt like, what's modeled and what isn't. Thirty minutes, books open.

Book a callSee the full record

12 months to change your mind · no lock-up · no obligation

Important information

Hypothetical, out-of-sample model performance. Results do not reflect actual trading in client accounts, are not audited, and do not include Kairos fees or taxes. Model results have inherent limitations, including the benefit of hindsight in their design. Leveraged ETFs reset daily and can lose money in flat or volatile markets. Past performance does not guarantee future results. See disclosures.