Inflation Compass (Varadi’s growth and inflation sector model)
David Varadi's Inflation Compass sorts markets into four regimes by growth (the S&P 500 against its 200-day average) and inflation (5-year breakeven inflation above 2% and rising). It holds energy, technology, utilities, or consumer staples with Treasuries, the sector that has done best in each regime.
Designed by David Varadi, 2026. Implemented and tracked by Tactfolio.
| Apr 2003 – Sep 2026 | Strategy | SPY |
|---|---|---|
| Annual return (CAGR) | 20.4% | 11.8% |
| Worst drawdown | -23.8% | -55.2% |
| Sharpe ratio | 1.10 | 0.70 |
| Volatility | 18.4% | 18.5% |
| Annual return since publication (Aug 2026) | — | — |
The Inflation Compass is a sector rotation model from David Varadi, published on his CSSA blog in July 2026. It reads two things from markets: growth, from the S&P 500's trend, and inflation, from the bond market's inflation expectations. Each of the four combinations points to one sector: energy, technology, utilities, or consumer staples with Treasuries.
The idea#
Varadi's premise is that sector leadership follows the economic climate. Energy tends to lead when the economy is growing and inflation is rising, and technology when growth comes with falling inflation. In a downturn, utilities have held up better when inflation is high, and consumer staples and Treasuries when it is falling.
Varadi reads the climate from market prices instead of economic reports, which arrive late. The S&P 500's 200-day average stands for growth. Five-year breakeven inflation, the inflation rate the Treasury market expects, stands for inflation.
How it works#
At the close of the last trading day of each month:
- Growth is up when the S&P 500 (SPY) is above its 200-day average.
- Inflation is rising when 5-year breakeven inflation (FRED series T5YIE) is above 2% and higher than 60 days earlier.
- Hold energy (XLE) when growth is up and inflation rising, and technology (XLK) when growth is up and inflation falling.
- Hold utilities (XLU) when growth is down and inflation rising, and 50% consumer staples (XLP) with 50% intermediate Treasuries (IEF) when both are down.
This is a partial version. Varadi also counts inflation as rising when a ratio of cyclical to defensive sectors has an upward 60-day slope, which the builder cannot compute. Without it, this version calls inflation rising less often.
What the backtest shows#
The test starts in April 2003, when breakeven inflation data begin. Since then it returned about 20% a year, against 12% for the S&P 500 and 8.6% for a 60/40 stock and bond mix. Its worst drawdown was about 24%, against 55% and 33%, and its risk-adjusted return was far above both.
It lost 10% in 2008, when the S&P 500 lost 37%. It gained 51% in 2021 and 17% in 2022, when energy led, while the S&P 500 lost 18% in 2022. Its worst drawdown is recent, from May to July 2026.
Almost all of this history is the period Varadi designed the model on, so treat the results as in-sample. The model was published in July 2026, too recently for a track record.
When it struggles#
- Concentration. It holds one sector at a time, apart from the staples and Treasuries mix, so a bad month in that sector hits the whole portfolio.
- Regime changes. A 200-day average and a 60-day change react slowly, so it can hold the wrong sector for a while when the climate turns.
- Choppy inflation expectations. When breakevens hover near 2%, the signal can flip between energy and technology.
Using it on Tactfolio#
The live strategy above runs the model on SPY, the four sector funds, and IEF, with breakeven inflation from FRED. Copy it to try other sectors or thresholds, or compare it with a growth-only switch such as the 200-day moving average.
Year by year
| Year | Strategy | SPY |
|---|---|---|
| 2026* | 13.5% | 14.0% |
| 2025 | 21.9% | 17.7% |
| 2024 | 21.0% | 24.9% |
| 2023 | 28.0% | 26.2% |
| 2022 | 16.9% | -18.2% |
| 2021 | 51.2% | 28.7% |
| 2020 | 33.4% | 18.3% |
| 2019 | 28.2% | 31.2% |
| 2018 | 2.6% | -4.6% |
| 2017 | 34.3% | 21.7% |
| 2016 | 16.2% | 12.0% |
| 2015 | -0.1% | 1.2% |
| 2014 | 11.6% | 13.5% |
| 2013 | 21.7% | 32.3% |
| 2012 | 12.2% | 16.0% |
| 2011 | 1.0% | 1.9% |
| 2010 | 11.5% | 15.1% |
| 2009 | 25.7% | 26.4% |
| 2008 | -10.3% | -36.8% |
| 2007 | 38.0% | 5.1% |
| 2006 | 33.7% | 15.8% |
| 2005 | 12.8% | 4.8% |
| 2004 | 30.0% | 10.7% |
| 2003* | 43.0% | 30.9% |
* Partial year.
The rules as implemented
This is the exact tree Tactfolio runs, rebalanced monthly with signals and trades at the close. Open it to inspect or copy it.
- StrategyInflation Compass
- WeightEqual
- Ifcurrent price of SPY is above 200d moving average of SPYThen
- WeightEqual
- IfAll of 2 conditions
- 5-year breakeven inflation is above 2
- 60-day change in 5-year breakeven inflation is above 0
Then- WeightEqual
- TickerXLE
Otherwise- WeightEqual
- TickerXLK
- IfAll of 2 conditions
Otherwise- WeightEqual
- IfAll of 2 conditions
- 5-year breakeven inflation is above 2
- 60-day change in 5-year breakeven inflation is above 0
Then- WeightEqual
- TickerXLU
Otherwise- WeightSpecified
- TickerXLP50%
- TickerIEF50%
- IfAll of 2 conditions
- WeightEqual
- Ifcurrent price of SPY is above 200d moving average of SPY
- WeightEqual
Sources and caveats
- Growth is SPY above its 200-day average; inflation is rising when 5-year breakeven inflation (FRED T5YIE) is above 2% and higher than 60 days earlier, and XLE, XLK, XLU, or 50% XLP with 50% IEF are held by regime, as in the source.
- The source also counts inflation as rising when a 60-day regression slope of a cyclical-to-defensive sector basket is positive. The builder cannot compute that slope, so this version leaves it out and calls inflation rising less often.
- Breakeven inflation begins in 2003, so the test starts then.
- Each economic figure counts only from the day it was first published, and as first reported: at a month end the latest figure is the previous month's, the one-month lag the source describes.
- Signals and trades use the close of the last trading day of each month, as in the source.
Common questions#
What is the Inflation Compass?#
It is David Varadi's sector rotation model that sorts markets into four regimes by the S&P 500's trend and the direction of expected inflation, and holds energy, technology, utilities, or consumer staples with Treasuries.
How does it measure inflation?#
It uses 5-year breakeven inflation, the gap between regular and inflation-protected Treasury yields. Inflation counts as rising when it is above 2% and higher than 60 days earlier.
What ETFs does the Inflation Compass use?#
XLE, XLK, XLU, and XLP with IEF, one choice per regime, plus SPY for the growth signal.
Does the Inflation Compass still work?#
It was published in July 2026, so there is no track record yet; the backtest covers the period it was designed on. The live record on Tactfolio shows how it is doing from here.