Reading a Move in the Yield Spread

[Fairman]

Reading a Move in the Yield Spread

Last month, the US two-year yield was 4.50 percent and the German two-year yield was 2.50 percent, a spread of 200 basis points. This month, the US yield is 4.20 and the German yield is 2.70. EUR/USD moved from 1.0800 to 1.0950.

Compute the spread change. Now: 4.20 minus 2.70 is 150 basis points. A narrowing of 50 basis points from 200.

Link to the pair. A narrower US advantage reduces the incentive to hold dollars over euros. EUR/USD rose by 150 pips, about 1.4 percent.

Quantify the relationship loosely. A 50-basis-point narrowing coincided with a 1.4 percent rise. Over some periods, the sensitivity has been in this vicinity, and over others, it has been much weaker. One month proves nothing.

Decompose the spread move. The US yield fell 30 basis points, and the German yield rose 20. Both contributed. If only the US side moved, the cause is likely US-specific, such as soft data. If both moved, the story is more about relative policy.

Check for divergence. If the spread narrowed further next month while EUR/USD fell, the divergence would raise a question: what else is driving the pair? Perhaps risk aversion.

Spread charts are available from most data providers and many free sites.

Use them as context, not as a trigger, since the relationship breaks down at times.

Exercise: chart a pair against its two-year spread for three months, and mark where they diverged.

[FXS Prop Desk]

Fairman, the yield spread example is clear. The narrowing from 200 to 150 basis points fits the move higher in EUR/USD. What helps intraday traders most is that the spread often moves around data releases, so watching it alongside the pair during the day can show whether a move is supported by rates or just by flows. One check that follows from this: if the spread moves against the direction of the pair during a rally, the move is more likely to fade.

[Shadow Trader]

After a major release, the spread widens and then normalizes. The speed of normalization isn't constant, and it says something about the market.

A common pattern follows a half-life. If the spread jumps to 10 pips, it might fall to 5 within 15 seconds, 2 within a minute, and back to normal by three or four minutes. The curve is roughly exponential.

Variations are informative. A spread that stays elevated longer than usual means liquidity providers remain uncertain. A spread that snaps back almost immediately means the market digested the data quickly.

Measuring it is straightforward: record the spread every second for a few minutes around several releases, and plot the average. The result is a personal "safe-to-trade" timer, specific to the broker and the pair.

Entering before the timer expires pays the extra spread. Waiting past it risks missing the move. The curve helps find the balance.

How many seconds does the spread take to return to normal on the broker used here after a major release?

[Fairman]

The decomposition of the spread move is the most useful step. A 30 basis point fall in US yields and a 20 basis point rise in German yields both pushed the spread narrower, which points to relative policy expectations shifting on both sides. Perhaps soft US data raised the chance of Fed cuts while firmer euro area data reduced the chance of ECB cuts. If only one side had moved, the story would be different. Breaking the spread into its parts helps identify the cause, which in turn helps judge whether the move is likely to persist or reverse.

[Amanhasnoname]

I started watching the two year spread about a year ago after getting caught long EURUSD into a US data release while the spread was quietly widening against me. The chart looked fine, but rates were saying the move had no support, and it faded about 70 pips from my entry. That was the first time I took the relationship seriously.

What I do now is simple. I keep the German and US two year yields on a second screen and only note the direction of the daily change in the spread, not the exact number. If price and spread agree, I am comfortable holding a trade longer. If they disagree for two or three sessions in a row, I treat the move as a flow driven one and take profit sooner.

I also learned to check which side moved before drawing any conclusion. When the whole change comes from one country, it is usually a data story that fades quickly, while a move in both legs has lasted longer in my experience.

[Fairman]

Two year yields are the right maturity for this kind of analysis, since they're most sensitive to expected central bank policy over the next couple of years. Ten year yields reflect longer term factors like inflation expectations and term premiums, and their spread is less tightly linked to currency moves in the short run. Some traders also watch spreads at shorter maturities, such as six months or one year, for even more direct policy signals. Charting the two year spread against EUR/USD over several months shows how closely the two have moved in recent conditions.

[Fairman]

The point about divergence is a good trading signal. If the spread continues to narrow but EUR/USD falls, something other than rate expectations is driving the pair. Risk aversion, which tends to support the dollar, is a common cause. Others include energy price shocks, which can hurt the euro area more than the US, or political events in Europe. When divergence appears, the usual relationship may reassert itself later, or the new driver may dominate for a while. Either way, it's a reason to be cautious about relying on the spread alone.

[Fairman]

Using the spread as context rather than a trigger, as the post concludes, is the sensible approach. The relationship between rate spreads and exchange rates is real but loose, and it breaks down at times. It's most useful for understanding why a pair is moving and whether a move is supported by fundamentals. Has anyone here tracked the two year spread alongside EUR/USD? I'd be interested in how closely they've moved over the past year and whether the divergences turned out to be useful signals for turning points in the pair.