Recent Posts

Friday, May 22, 2020

The Incoherence Of Yield Curve Control

Chart: Long-Term Treasury Yields And Yield Cap

Yield curve control -- setting bond yields by the central bank -- has returned to discussion. Mainstream economics over-emphasises the role of interest rates in guiding the economy, and based on previous experience, it seems entirely likely that some form of new policy will be attempted to counter-act economic weakness. The main options appear to be negative interest rates and yield curve control, and it seems clear that Fed officials prefer yield curve control to negative rates, Yield curve control will either accomplish very little, or be viewed as a mistake in retrospect. That said, it is still a more sensible policy that large scale purchases of Treasurys by the Fed (Quantitative Easing).

From the perspective of Modern Monetary Theory (MMT), this topic is interesting, since fixing term interest rates has a strong resemblance to a proposed policy of locking interest rates at zero. However, the implementation that would be used by the Federal Reserve would be the wrong way to do it, and missing the point of the policy.

Setting a trading range for Treasury yields is not unprecedented policy for the Fed. Even if we put aside recent experiments elsewhere, Treasury yields were capped at 2.5% during World War II and the immediate post-war era. I discussed this very quickly in one of my very early posts, which also linked references on the topic. Keynes also referred to such a policy in the General Theory.


Setting an upper bound for Treasury yields is a straightforward operation for the Federal Reserve: it just announces the cap level, and purchases Treasurys without quantity limitations whenever yields push near that level.

So long as the long-term yield is higher than the overnight rate (in particular, the repo rate) and nobody expected the cap to be lifted, investors will do most of the heavy lifting. They will buy the bonds at yields below the cap level, and they can finance the positions with positive carry. There is very little risk for the investors, so long as they can sell the bonds to the Fed at the capped rate. This is like any other peg set by a central bank -- as long as the peg is credible, it will not need to intervene too aggressively for the peg to hold.

However, if investors expect the peg to disappear and yields to rise significantly above the cap, they will not be willing to hold the bonds. In that case, the Federal Reserve could be forced to buy up the entire stock of issued Treasury bonds.

By providing a price signal, this policy is more effective than purchasing volumes of Treasurys (Quantitive Easing), since market participants had zero guidance on what level of Treasury yields the Fed wanted to achieve.

A Potential Nothingburger

Chart: Slope of 2-Year Treasury over Fed Funds Target
One potential implementation of yield curve control is to only cap yields up to a certain point of the curve, for example, the 2-year. Such a policy would be easily rolled off. For example, imagine that the policy is put into place until 2023. At which point, the cap would apply to bonds maturing in 2025. All that needs to be done is then not roll that maturity date forward, and so the policy would have been exited by 2025.

However, such a policy accomplishes very little in practice. The figure above shows the slope between the Fed Funds target rate (midpoint of a band when the Fed switched to using a band) and the 2-year Treasury. If we look at the mid-2010s, we see that the slope was negligible. This is what one would expect in an economy that is stagnant. The only thing such a policy would prevent was the knee-jerk selloffs that resulted from bond investors over-reacting to Fed rate cuts (as in 2010). Given the magnitude of the economic contraction, I doubt that a similar sell-off could be engineered any time soon.

But even if such a sell-off materialised, pushing rising forward rates more than two years into the future will not accomplish much. Term borrowing rates would rise, which is what the Fed presumably wants to avoid.

Capping Long-Dated Bonds Trickier to Exit

To be more interesting, Treasurys of all maturities would need to have a yield cap. The problem is straightforward: if you cap a 10-year note yield, it will not be possible for the policy to naturally roll off for about a decade (since a cap on a one-year is not too significant).

Realistically, the Fed would have to abandon the policy in a way that has bonds jumping from capped status to floating yields. Unless the Fed is willing to buy every bond holder out, widows and orphans will show up and burn the effigies of Fed officials on the steps of the Federal Reserve headquarters.

How Will it Affect the Economy?

I am in the camp that lowering rates by small amounts will do very little to stimulate growth, and we can look at recent cycles for evidence for that assertion. Furthermore, capping Treasury yields may matter little if the Fed is forced to buy a large portion of the float, pushing everyone else into riskier securities. Credit spreads would just peel off from the Treasury curve, and thus there would be no benefit to private sector borrowing costs.

The effect may be more psychological, as was the case for Quantitative Easing. Hard money advocates will scream about "price discovery," and insist that hyper-inflation is around the corner. This might generate some frantic investment activity, which may or may not add to aggregate demand. However, that is what happened in response to "money printing" worries in the last cycle, and the recovery was still notoriously slow.

Major Contrast with MMT Policy

 One might argue that capping Treasury yields would open policy space for fiscal policy. This is not really the case. "Fiscal space" is a psychological construct of policymakers, and the elites in charge of fiscal policy would much prefer that monetary policy do the heavy lifting. People would very reasonably argue that bond yields would spike when the yield curve cap is lifted, and so the Treasury cannot extrapolate the current low level of yields forever.

By contrast, the MMT proposal is based upon the premise that monetary policy is weak, and is to be taken out of the picture permanently. With rates locked at 0% by statute, there is no "yields will spike, leading to a fiscal death spiral" story to sell.

Yield curve control is in an unhappy middle ground. It is an admission that lowering the policy rate does very little, but by leaving the alleged threat of an "independent" central bank driving the Treasury into bankruptcy in place, fiscal policy remains knee-capped. (I don't believe the stories that the Fed has the right to drive the Treasury into bankruptcy, but many fiscal conservatives do not let facts get in the way of their campfire scare stories. What matters is what the fiscal policymakers believe.)

Concluding Remarks

Barring a miracle cure for COVID-19, the United States is drifting into a multi-year period of extremely depressed activity. There does not appear to be capacity to eradicate the virus, nor are older consumers or office workers willing to take meaningful health risks to benefit capitalism. Unless there is a magical transformation in the attitudes of the ruling elites, the fiscal policy response will remain reactive, and ineffectual. The Fed is the only entity in the United States that takes any responsibility for the effectiveness of policy, and so we should expect to see greater leaps in its policy framework.

Although yield curve control is the most likely next step, negative interest rates cannot be ruled out. Health worries might strengthen the hand of those advocating the abolishing of paper money, removing one institutional barrier to negative interest rates.

(c) Brian Romanchuk 2020


  1. Great article. Thanks for the analysis of the current situation. In my opinion, I think it’s important to separate the emotions from investing. How the market reacts to news, I can’t control. However, I can control how and when I invest, so I tend to focus on that.

  2. I am not quite making sense of all of this. I don't think the logic is correct.

    Let's consider a quiescent market with cash all in the hands of the most recent borrower. Next we add two disturbances. The borrower spends all his cash AND a new potential lender has all that cash.

    Here are the new emerging questions: What shape of yield curve would the next very large borrower need AND what shape of yield curve would the next very large lender need?

    It seems to me that the shape would be a mutual agreement between the two parties.

    Now shift the problem by placing cash into a spectrum of smaller owners. Now the problem is to accumulate cash into the quantities needed by the big borrower, in preparation for the next big borrow event. I can't see that this effort should have any direct impact on the yield curve.

    I am just trying to understand the basic theory behind interest rates and the time curve.

  3. I continue reading and trying to understand the concepts behind your words. Thanks for presenting a tantalizing post.

    As I read for the umpteenth time, I am wondering about using a standardized bond document, for the purpose of presenting a common loan document for all loans. For example, a law that forced every loan document to have a zero interest rate specified no matter what the term. Thus, the initial yield curve would always be flat, set at zero.

    If the underlying law said nothing about interest rates on the resale market, resales presumably could be found to generate a yield curve. Resale purchasers would be more or less reluctant to buy depending upon the term to payout. They would react by forcing discounts or charging fees to sellers.

    Along comes the Central Bank and decides to reset the RESALE yield curve. Presumably this would only be done if the CB thought the discounts offered ware too onerous. If I have the patterns correctly visualized, CB action would shift the yield curve into a flatter (closer to original) shape.

  4. What do you think will happen to the 20 year us bond?

    1. I don’t do market calls, but if the last cycle is the guide, all we’ll get is some range-trading for an extended period. Hard to see why the Fed hikes, and I am skeptical that the Fed will resort to deeply negative rates.

    2. How is it possible governments can borrow unlimited amounts of money for decades for free? Who is lending this ? Are they lenders going bankrupt slowly ? Where is all this money to lend for no return coming from and how can it go on for decades or an entire human lifetime? What's going on?

  5. This comment has been removed by a blog administrator.

  6. This comment has been removed by a blog administrator.

  7. Sir,
    what will happen to the Yield Curve when MMT would be implemented?Would there be a flat 0% curve along the whole curve? The whole financial system ist relying on it.

    1. Saying “when MMT is implemented” doesn’t really work. It’s like saying “when neoclassical economic theory is implemented.”

      MMT proponents call for government yields to be pinned at zero - stop issuing bonds. My view is that nothing much exciting would happen to the economy, other than some govvie traders needing to find new jobs.


Note: Posts may be moderated, and there may be a considerable delay before they appear.

Although I welcome people who disagree with me, please be civil.

Please note that my spam comment filter appears to dislike long "anonymous" posts.

Note: if you want to post comments from Apple devices (iPhone, iPad), you apparently need to turn off "prevent cross-site tracking" in Safari privacy settings. (The reason presumably is that another URL handles comments, and so the user session needs to be preserved when redirected to that site. I don't like this, but this is not enough to make me switch my hosting service.)