What Is Steepening and Flattening of the Yield Curve?
Two things happen to the curve on any given day. Yields go up or down, and the gap between the long end and the short end widens or narrows. Cross the two and there are four ways the day can land. Which one you are looking at tells you what moved the market, and it changes the trade.
1. What does steepening and flattening actually mean?
The yield curve plots yields across maturities, from the 2 year out to the 30 year. Steepening means the gap between long and short yields widens. Flattening means it narrows.
That is only half of it. Widening and narrowing tell you nothing about which way yields moved, so you have to read direction at the same time. Yields falling is a bull move, because bond prices rise. Yields rising is a bear move. Every session gives you one of each, a bull or a bear and a steepener or a flattener. That is where the four regimes come from.
The most quoted single measure of the shape is the 2s10s spread, the 10 year yield minus the 2 year yield, in basis points. Our yield curve guide goes further on what the shape signals.
2. What is a bull flattener?
Yields fall, the long end falls more than the short end, and the gap narrows.
That is demand for duration. Buyers reach further out the curve to lock in yield before it goes, usually because growth worries them or because inflation expectations are easing.
The same shape turns up on risk off days in equities, and a tech led selloff is the cleanest version of it. Money leaves stocks for long duration Treasuries, and that pulls the long end down fastest. You still get bull flattening, but the equity tape drove it rather than a growth scare, and the policy story has not changed at all.
None of that is a call on the next meeting. It is about the medium term, the market saying growth and inflation look weaker further out even though the central bank has not moved yet. Late in a cycle this is common, and a sustained run of it is what eventually produces inversion.
3. What is a bear flattener?
Now yields rise, and it is the short end rising more than the long end, so the gap narrows again. This is the classic hawkish repricing, and the easiest of the four to spot live.
An upside inflation surprise lands, the market prices more tightening or fewer cuts, and the front end sells off hardest because it prices the policy path most directly. The long end rises less, because tighter policy now argues for lower growth and inflation later.
A hot CPI print that produces a violent move in the 2 year and a muted one in the 30 year is a bear flattener.
4. What is a bull steepener?
Yields fall again, but this time the short end falls more than the long end, so the gap widens. This is the dovish mirror of the bear flattener.
Weak data lands, the market prices cuts sooner or deeper, and the front end rallies hardest for the same reason it sells off hardest on hawkish news: it tracks the policy path most directly. A soft payrolls print is the textbook trigger.
People watch a sustained bull steepening closely, because it is often the first curve move of an easing cycle. It is the market saying out loud that it has stopped believing in higher for longer.
5. What is a bear steepener?
Yields rise here too, but the long end rises more than the short end, so the gap widens.
This one is not really about the central bank. The long end is demanding more compensation, whether that is rising term premium, heavy issuance, fiscal concern or worries about long run inflation. A badly received 30 year auction can produce one on its own, which is why the tail on a long dated auction is worth watching even if you never trade bonds.
Watch which end is doing the work. A steepening led by the front end rallying is the market repricing the Fed. A steepening led by the long end selling off is the market repricing risk or supply. They draw the same direction on a 2s10s chart, but the cause is different and so is the trade.
6. How do you trade these with Treasury futures?
Futures prices move inversely to yields, so buying a contract is being long duration and it makes money when yields fall.
Putting on a steepener
You make money when long yields rise relative to short yields. That means the long maturity contract falls relative to the short maturity one, so you buy the short maturity contract and sell the long maturity contract. Buy ZT, sell ZN, for example.
Putting on a flattener
Exactly the reverse. Sell the short maturity contract and buy the long maturity contract. Sell ZT, buy ZN.
Both are trades on the shape of the curve, not its direction, and that only holds if the legs are DV01 weighted. Trade them one lot for one lot and you have taken a directional position without meaning to. It is the most common way a curve trade goes wrong for reasons that have nothing to do with the curve.
7. What is DV01, and why does it set the ratio?
DV01 is the dollar value of a one basis point move in yield, for one contract.
Every Treasury futures contract has a different one, because each tracks a different cheapest to deliver bond with its own maturity and conversion factor. A one basis point move is worth far more in the bond contract than in the two year, because there is more duration in it.
So you cannot trade curve spreads one for one. Size each leg so that both carry the same total DV01. A parallel shift in yields then nets out between the legs and only the change in the spread reaches your profit and loss, which is the whole point of the trade. The ratio you want is the DV01 of the long maturity leg divided by the DV01 of the short maturity leg.
The catch is that DV01 is not fixed. The cheapest to deliver bond changes, and every contract's DV01 drifts with it, so the correct ratio changes too. Question ten covers where to get the current numbers.
8. Why does ZT need different treatment when charting?
A lot of people get caught out here, and the chart ends up disagreeing with the position they are holding.
ZT carries double the notional
ZT, the 2 year note future, has a face value of $200,000. ZF, ZN, TN, ZB and UB all have $100,000, so one full point of price movement is worth twice as much in ZT as in any of the others.
Execution ratio and charting ratio are not the same number
Take the TUT spread. To execute it DV01 neutral the ratio is roughly 2 x ZT - ZN. To chart the same position in price terms the formula is 4 x ZT - ZN.
The doubling is the notional, not the position. You hold two ZT per ZN, and then you double again in the chart formula because each ZT point is worth twice each ZN point.
Chart 2 x ZT - ZN and you will get a line that does not match the profit and loss of the position you actually hold. The same doubling applies to every spread with a ZT leg, so it affects TUF and TUL as well as TUT.
9. What are TUF, TUT, FOB, NOB, BOB and TUL?
They are the standard nicknames for the listed inter-commodity spreads, and each one pairs two points on the curve.
TUF, TUT and TUL all start at the two year. TUF runs it against the five year, ZT versus ZF, TUT against the ten year, ZT versus ZN, and TUL all the way out to the ultra long bond, ZT versus UB. Further up the curve, FOB is five year over bond, ZF versus ZB, and NOB is notes over bonds, ZN versus ZB. BOB is bond over bond, the classic bond against the ultra bond, ZB versus UB.
In each case the shorter maturity is the front leg. Buying the listed spread generally means buying that front leg and selling the back one, which is a steepener. Check that convention on the contract specification rather than assuming it.
Any spread with a ZT leg inherits the notional doubling from question eight. And the further apart the two legs sit on the curve, the more the spread is a bet on term premium and supply rather than on the Fed, which is why TUL and NOB behave quite differently from TUF.
10. Where do I get the correct weights and margin credits each quarter?
From the CME Group QuikStrike Treasury Analytics tool, under the inter-commodity spread section. It publishes the current weightings for each listed spread along with the margin credits the exchange applies.
The ratios are not fixed. The cheapest to deliver bond changes and every contract's DV01 moves with it, so a weighting that was right last quarter can be wrong after the roll. Check it every quarter at rollover instead of carrying an old number forward, because a stale ratio turns a curve trade into a small directional one without telling you.
Margin is worth knowing about too. Trading the listed spread, rather than legging in separately, is what earns the recognised margin offset, so the same economic position can require very different margin depending on how you put it on.
Where the terminal fits
Helious sorts the curve into these four regimes automatically. Instead of reading one spread it compares the average of the short leg against the average of the long leg, which catches moves led by the belly that a single 2s10s reading can hide, and it names the regime live on the rates board. Beside the board sit the calendar, every release scored against its own history the second it prints, the auction results that drive bear steepeners, and a live squawk. It is $39.99 a month with a free tier, and the methodology page shows the workings.
Where to go next
You can watch the curve live on the rates page, and the 2s10s spread has a page of its own. For more on shape there is the yield curve guide, plus term premium and inversion.
CPI and the FOMC hub cover what moves the front end, and Treasury auctions cover the back. If the units are new to you, start with basis points and duration. There is a longer piece on how to read Treasury auctions as well, and one on what a basis point is.
Helious names which of the four curve regimes is running, live, with the calendar, the auction results and every release scored the second it prints on the same screen. $39.99 a month with a free tier. Built by traders, for traders.
Contract specifications, spread weightings and margin credits are set by CME Group and change over time, so always confirm current values on the exchange's own tools before trading. This post is general information and not financial advice, and trading futures carries substantial risk.
Launch the terminal →