Friends, retail forex traders look at interest rates. Institutional desks look at the shape of the yield curve. There is a world of difference between the two, and that difference is where some of the most reliable FX signals live. A single rate tells you one thing: where the market prices a specific maturity today. A yield curve spread tells you something far more valuable: what the market believes is changing about the economic outlook, the policy path, or the risk environment. And those changes are what move currencies. Not the level. The delta. Two spreads dominate institutional FX analysis. The first is the 3-month to 2-year spread (3m2s). The second is the 2-year to 10-year spread (2s10s). They operate on different timescales, respond to different catalysts, and tell you different things about where currencies are headed. Most retail traders have never calculated either one. That gap is your opportunity.
The 3m2s Spread: Your 1-to-4-Week FX Signal
The 3-month Treasury bill rate is mechanically anchored to the current federal funds rate. Between FOMC meetings, it barely moves. It is, for practical purposes, a fixed point.
The 2-year yield, as we discussed in Part 3 of this series, is the market’s real-time estimate of the near-term policy path. It moves daily based on data releases, Fed speeches, and shifts in rate-cut or rate-hike expectations. When you subtract the 3-month rate from the 2-year yield, you isolate one thing: the market’s repricing of expected policy changes, stripped of the anchoring effect of current policy. This is the 3m2s spread, and it is one of the cleanest near-term FX signals available. When 3m2s widens (2-year yield rising relative to 3-month), the market is pricing in fewer rate cuts or later cuts. This is USD-positive because it means the rate differential between the US and other economies is expected to persist or widen. When 3m2s compresses or inverts (2-year yield falling relative to 3-month), the market is pricing in more cuts or sooner cuts. This is USD-negative because it means the rate differential is expected to narrow. The transmission lag is remarkably consistent. Academic research shows front-end rate changes transmit to FX within 1 to 4 weeks. An ECB conference paper by Krohn (2025), “Demand-Driven Risk Premia in Foreign Exchange and Bond Markets,” provides precise timing: Treasury demand shocks cause USD depreciation that peaks at approximately 10 days and reverts to insignificance after 14 days. The full paper is available from the ECB: www.ecb.europa.eu This 10-day peak maps to the 3m2s transmission window. When you see the 3m2s spread move sharply on a data release, the FX response will build over the following 1 to 2 weeks, not instantly. This is the lag that gives you a trading window if you are watching the right numbers. The BIS confirmed this timing in Working Paper 1195, documenting that monetary policy effects on longer-term yields “build up gradually over time, beyond a horizon of one month.” Short-end effects are faster: www.bis.org Here is a practical example. On March 6, 2026, US nonfarm payrolls printed at negative 92, 000 against a consensus expectation of positive 59, 000. A massive miss. The 2-year yield dropped immediately as the market repriced toward earlier rate cuts, compressing the 3m2s spread. Over the following two weeks, the dollar weakened against most G10 currencies. The 3m2s told you the direction before the FX pairs completed the move.
The 2s10s Spread: Your 1-to-4-Quarter Structural Signal
While the 3m2s captures near-term policy repricing, the spread between the 2-year and 10- year yields captures something entirely different: the structural outlook for growth, inflation, and term premium. The 2-year yield reflects expected Fed policy. The 10-year yield reflects expected Fed policy plus term premium, which compensates investors for inflation risk, fiscal risk, and duration risk over a longer horizon. The gap between them tells you how the market views the longer-term trajectory relative to the near-term path. When 2s10s steepens (10-year yield rising faster than 2-year), the market is either pricing in better growth prospects, rising inflation expectations, or increasing fiscal risk. For currencies, this typically supports USD when driven by growth optimism (capital flows into US assets) but can undermine USD when driven by fiscal concerns (investors demanding compensation for holding US duration). When 2s10s flattens or inverts (2-year yield rising faster than 10-year), the market is pricing in tighter near-term policy against a backdrop of slowing long-run growth. Classic late-cycle signal. For currencies, this has historically been USD-positive in the short run (rate differential widens) but USD-negative at longer horizons (recession approaching). The transmission lag for 2s10s is measured in quarters, not weeks. Chinn and Meredith (2004), in one of the most cited papers in the exchange rate literature, tested uncovered interest parity using G-7 long-maturity bonds. Their finding: the relationship between interest differentials and exchange rate changes fails at 1-to-12-month horizons but holds at 5-to-10-year horizons, with the coefficient approaching unity at the long end. The paper is available through the NBER: www.nber.org What this means in practice: when the 2s10s spread shifts, do not expect the FX market to respond this week. The signal operates over 1 to 4 quarters. It is a positioning signal for structural trades, not a day-trading trigger. This is a critical distinction that most retail analysis misses. The 3m2s is your tactical signal (weeks). The 2s10s is your strategic signal (quarters). Using one where the other belongs leads to timing errors that feel like the framework is broken when it is actually just being applied on the wrong timescale.
Term Premium: The Hidden Driver Most Traders Miss
Here is the institutional pearl in this article. Most retail traders think yield curve movements are entirely about Fed rate expectations. They are not. A significant and increasingly important component is term premium, the extra compensation investors demand for holding longer-duration bonds instead of rolling short-term bills.
Greenwood, Hanson, Stein, and Sunderam published “A Quantity-Driven Theory of Term Premia and Exchange Rates” in the Quarterly Journal of Economics in 2023. Their finding fundamentally reframes how the yield curve drives currencies: the component of long-term rate differentials that matters for exchange rates is a forecastable term premium differential, not expected future short rates. The paper is available through the NBER: www.nber.org And on SSRN: papers.ssrn.com Their estimated coefficients are striking. The short-rate differential coefficient is approximately 4. 72, and the term premium differential coefficient is approximately 2. 99. Both are highly significant. But here is the key insight: supply shocks to long-term bonds in one currency influence both the exchange rate and term premia in both currencies. When the US Treasury issues more long-term debt, it raises term premium on US Treasuries, which changes the term premium differential versus other countries, which moves exchange rates. This is why large fiscal announcements can move currencies even when the Fed has not changed anything. The mechanism is not through expected short rates. It is through term premium. The practical implication: when the 10-year yield rises, you need to ask why. If it is rising because the market expects higher short rates (hawkish Fed repricing), that is transmitted through the 2-year yield and the 3m2s spread. If it is rising because term premium is expanding (fiscal concerns, increased duration supply, or reduced foreign demand for Treasuries), that is a different signal with different FX implications. Rising term premium tends to strengthen USD when it reflects global demand for US safe assets. It tends to weaken USD when it reflects fiscal deterioration and loss of confidence in US debt sustainability. The same 10-year yield move can have opposite FX implications depending on the driver.
Bull Steepening, Bear Flattening, and Why the Names Matter
Institutional desks describe yield curve movements using four combinations, and each carries different FX implications. This taxonomy is worth learning because it tells you instantly what kind of macro environment you are in and what it means for currencies. Bull steepening (short rates falling, long rates stable or rising): The market is pricing in rate cuts while long-end inflation or growth expectations hold. This typically happens when the
Fed is expected to ease. USD-negative, because the front-end rate differential is compressing. Bear steepening (long rates rising, short rates stable): The market is pricing in higher term premium or rising inflation expectations while near-term policy stays put. USD-ambiguous: positive if driven by growth optimism, negative if driven by fiscal concerns. Bull flattening (long rates falling, short rates stable or falling faster): Classic risk-off move. Flight to safety into long-duration Treasuries compresses the 2s10s. USD behavior depends on whether the risk-off is global (USD-positive on haven demand) or US-specific (USD- negative on domestic stress). Bear flattening (short rates rising, long rates stable): The Fed is tightening or expected to tighten, pushing up the front end while the long end suggests the tightening will eventually slow the economy. This is typically the most USD-positive configuration because it maximizes the front-end rate differential that drives FX. Fleming and Remolona (1999) documented in their paper “The Term Structure of Announcement Effects” that bond market reactions are concentrated in specific tenors depending on the type of news: employment data moves the long end, monetary policy surprises move the short end. The paper is available from the Federal Reserve Bank of New York: www.newyorkfed.org Understanding which part of the curve is moving and why is the difference between knowing the yield curve is steepening and knowing what that steepening means for currencies.
UIP Failure at Short Horizons, Success at Long Horizons
There is a deep irony at the heart of currency markets that most retail traders never encounter. Uncovered interest parity, the textbook theory that higher interest rates should cause a currency to depreciate over time to offset the yield advantage, fails spectacularly at short horizons but works almost perfectly at long ones. Chaboud and Wright (2003) at the Federal Reserve published “Uncovered Interest Parity: It Works, But Not For Long,” confirming that UIP holds over very short windows (intraday) and very long windows (5+ years) but breaks down at intermediate horizons of 1 to 12 months. The paper is available from the Fed: www.federalreserve.gov Chinn and Quayyum extended this work at the University of Wisconsin, finding that long-
horizon UIP holds with increasing reliability as the maturity lengthens. Their paper is available here: users.ssc.wisc.edu This finding has profound practical implications. At the 1-to-4-week horizon where the 3m2s spread operates, higher US rates typically strengthen the dollar. The carry trade works. UIP fails, and that failure is your profit. At the 1-to-4-quarter horizon where the 2s10s spread operates, the picture is more nuanced. The carry advantage still exists, but the UIP correction starts to bite. Currencies with large rate advantages begin to give back gains. This is why structural FX positions require more than just a rate differential argument. They need a view on whether the differential is widening, narrowing, or stable. At the multi-year horizon, UIP roughly holds. The currency with the higher rate does eventually depreciate enough to offset the carry. This is why long-term fair value models (based on purchasing power parity or real effective exchange rates) tend to be mean- reverting, while short-term models (based on rate differentials) tend to be trend-following.
Two Speeds, One Framework
The yield curve gives you two clocks for FX. The 3m2s spread runs on a weekly timescale. The 2s10s spread runs on a quarterly timescale. Neither is better than the other. They answer different questions. The 3m2s tells you: is the rate channel supporting or undermining the dollar this month? The 2s10s tells you: is the structural environment shifting in a way that will reshape currency trends this quarter? When both spreads point in the same direction, conviction is high. When they diverge, it means the near-term signal and the structural signal are in conflict, which is itself a signal worth paying attention to. A widening 3m2s (hawkish repricing) combined with a flattening 2s10s (slowdown expected) creates a tug-of-war that often resolves with increased volatility rather than a clean directional move. This is exactly what the US Rate Structure panel on the 4xForecaster dashboard (www.ecb.europa.eu 4xforecaster. com/) is designed to show. The front curve spread and the term structure spread are displayed together, with directional indicators, precisely because both must be read in combination. One without the other gives you half the picture. Half the picture is worse than no picture at all, because it gives you false confidence. The yield curve is the market’s collective estimate of the future. The spreads are where that
estimate becomes actionable for currency traders. Learn to read them, and you will see FX moves forming before they show up on the price chart. This is Part 4 of the Macro-to-FX Transmission Series from 4xForecaster. Next: Rate Differentials: The Structural Floor Under Every Currency Move.