Bond Economics

Who's The Spread Product Now?

2026-09-03T09:03:32.976-04:00

One of the points that is coming up in my news flow is the simultaneous rise in bond yields. This was of particular interest for the United Kingdom, where the chattering classes get deeply excited by rising gilt yields. I looked around at my data sources, and the above chart was somewhat interesting.

It shows American and Canadian 10-year government yields since 2020, which was when they were driven down in the pandemic panic. The yields were glued together until 2023 or so, when a divergence has opened up.

The interesting thing to long-time Canadian market watchers is that the U.S. yields are higher. This is not that novel — Canadian yields were lower in the 2010s as well (although I chopped that part of the history off so we can see recent moves). However, if we go back further, Canadian yields were persistently higher than American ones. This led to silliness like modelling the 10-year GCAN as a spread to U.S. Treasuries.

In a gold standard, senior countries will generally borrow at the lowest rates in the system. With fixed currency parities, the fair value spread between government bonds is zero, so the spread between them would reflect liquidity and credit/devaluation risk premia. (In the gold standard world, a devaluation is a form of default.) Modelling Canadian bonds as a spread to Treasurys would make sense in a gold standard system.(The historical experience might not align to this, as the Canadian dollar floated for most of the World War II era, and earlier, the British Pound would have been considered the “senior currency” for countries in the Commonwealth.)

That logic breaks down when currencies float. The fair value in each country is the expected value of the path of the overnight rate. So a country will have a positive spread over another if its expected path of nominal overnight interest rates is higher.

Up until a certain someone started swinging the tariff axe (and renaming bodies of water, threatening invasion, etc.), the Canadian and American economies were highly integrated and it made sense that the Bank of Canada would tend to shadow what the Fed is doing. What we started to see in the 2010s was that the Bank of Canada could keep its policy rate lower than the Fed while keeping the inflation rate on target.

This does not necessarily match market dogma, where people argue that Canada would need to keep interest rates higher because the U.S. dollar is a reserve currency, or whatever. The problem with international-led analyses like that is that international flows into fixed income are a pipsqueak facing off against the 200 kilogram gorilla that is the domestic housing market, which is far more interest-rate sensitive than international unhedged capital flows (which are dominated by equities).

The recent divergence is not really that surprising, and does not really reflect Canadian virtue or other pop psychological/political explanation one can come up with. (The British press loves such explanations.) Although the White House is hammering away on every “let’s raise prices!” button it can reach is obviously negative factor for U.S. bond prices, the tariff shocks on the Canadian economy darkens the growth outlook, which is supportive of bond prices. As such, the divergence is not surprising. It is unclear when the economic outlooks will re-converge.

Email subscription: Go to https://bondeconomics.substack.com/

I Try To Take A Break...

2026-08-26T15:02:08.379-04:00

I tried taking a break since I had some family commitments, but events have returned. Although I saw an interesting article from a theory perspective that I would like to write about, there have been two big market-oriented topics that have surfaced due to you know who.

Treasury Buybacks

There’s been a bit of excitement about buybacks of 30-year Treasurys at the behest of Treasury Secretary Bessent. There were some market fundamentalist responses, which in my view slightly miss the mark. The correct response to the situation is the following.

Although I am strong believer in “interest rate expectations determine risk-free yields,” there is an important qualification: the market in 30-year bonds is not a two-way market that is self-equilibrating. The financial authority (and the central bank if they go nuts with QE) along with regulatory stances are a major determinant of 30-year spreads versus the 10-year (which is closer to a self-equilibrating rate based on expectations).
Buying back 30-years that you just issued is an idiotic thing to do. If you do not like where the 30-year yield is, get your debt management team to stop issuing the damn things.

I did not bother digging into this story, but my guess based entirely on my prior beliefs is that Bessent did this solely because he wants to be seen doing big things that look impressive to ignorant people.

I have written about the “30-year disclaimer” in the past, so will repeat quickly. At shorter maturities (anywhere under the 5-year tenor for sure, the 10-year might be open to debate) there is an active two-way private sector market in relatively-low risk securities. For example, I believe the rule of thumb is that a brand new 30-year conventional mortgage lines up duration-supply wise with a 7-year Treasury under normal market conditions. Once you are at the 5-year point, there’s lots of credible issuers of 5-year bullet bonds. (A bullet bond has 100% principal repayment at maturity, conventional mortgages are amortising — principal payments occur during the lifetime of the bond — so they have much less weighting on the final payments.)

Private sector issuance of bonds (or borrowing at term rates) are economically equivalent to short-selling duration instruments. So “rate expectations: is not just what some punters at hedge funds and proprietary trading desks think, it is what everybody who borrows thinks as well. If an issuer thinks term rates are too high, they issue a shorter tenor (or float).

Although there has been some popular arguments about government bonds being irreplaceable, that is partly based on mysticism coming down from the 2008 Financial Crisis. In a financial crisis, central government bonds are irreplaceable. However, we are not continuously in the middle of major financial crises where the solvency of the banking system is being questioned. Instead, we have asset prices largely being determined by asset allocators whose primary decision can be (over-)simplified to “stocks versus bonds.” “Bonds” are supposed to be safe assets, but they typically have a smaller weight than equities. Instead of being confined to government bonds, you typically move some of your “private sector risk budget” to the bond portfolio and swap out central government bonds (“govvies”) for “spread product” (anything other than “govvies”). Although asset allocators typically hate bonds, everybody loves using some of their global risk budget to swap in spread product.

Problems arise in ultra-long tenors. Very few entities in the private sector can credibly issue ultra-long bullet bonds, so supply is almost entirely in the hands of governments. (Inflation-linked bonds have an even worse supply story.) Governments as near-monopoly issuers have no choice but to consider what the market will bear. And as the Brits proved in the 1990s, you can have too little supply. Pension fund liability-matching regulations forced funds into buying long-dated gilts beyond the ready supply, resulting in an extremely stupid yield curve.

American debt managers want to pretend that they are neutral and yields are purely market-determined, but that is just ideological silliness at the long end. Their ultra-long supply decisions are ultimately arbitrary. Which means, if the Treasury does not like the price, stop issuing so many of them.

Canada Trade War

The Canadian Trade War erupted again with Trump or other very senior officials allegedly sabotaging a nearly-concluded deal with demands that allegedly amounted to a vassalisation of Canada.

Although the “Conservative” Party of Canada might want to go along with Trump’s demands, they now largely represent a Western Canadian rural rump outside which the demands are complete non-starters. Dismantling French language protections is not going to happen. Even liquor “boycotts” which are recent policies are not easily negotiated away — the decision to pull American booze was provincial, and the provincial governments are powerful and almost inevitably more popular than the distant Federal Government. It would be necessary to bring the provinces into the negotiations. Furthermore, it might not even matter, many Canadian consumers are boycotting American products on their own.

Hight tariffs will be damaging for the Canadian economy, while the problems will be sectoral for the American. On paper, this would imply that the Americans have a stronger bargaining position. The problem is that this is an existential crisis for Canadians, while this is not a concern for most Americans. Pressing the issue would likely just result in an implosion in cross-border relations, not capitulation.

Email subscription: Go to https://bondeconomics.substack.com/

Participation Rate Drop Not That Big A Deal

2026-08-12T09:03:31.146-04:00

The 10-year U.S. Treasury yield (above) has been drifting higher, although it remains within a broad range that it settled into after the post-COVID Bond Apocalypse. Since I am not offering unsolicited investment advice to random strangers on the internet, I officially do not have a yield forecast. That said, the bond (price) weakness is somewhat surprising in retrospect given the economic uncertainty. However, the surprise factor is somewhat reduced by the reality that we now have a White House that is pushing almost every lever it can to raise prices. My bias is based on the post-1990 period where growth trumped inflation for bond yields — since inflation is a lagging variable. The difference is that we did not then have a lot of pro-inflation policymakers in power.

The (relatively) recent labour market data coming out of the United States has not been great. The above figure shows a measure that I like looking at — the employment-to-population ratio.

In order to give some background on the numbers, it is the (estimated) number of workers in paid employment in the economy divided by the population in the working age population. Since not everyone in that population is interesting in working/looking for work, the unemployment rate is not 100% minus the employment ratio. Instead, the unemployment rate is the percentage of the participating population (to be discussed next) that is employed.

The level of the employment ratio is not necessarily meaningful — as the demographics of the working population change, the employment ratio will reflect that. For example, the ratio was typically around 55%-58% (not shown) in the 1950s and 1960s (before when women (re-)entered the formal workforce in size). In the past couple of decades, the ageing population will reduce the ratio as the mix of population moves to a higher weighting in the cohorts near retirement age. Nevertheless, the weakness of the employment ratio is not a great sign — the ratio managed to rise in the 2010s during the expansion, and the demographic situation was not that radically different seven-eight years ago.

The American Labor [sic] Force Participation rate (above) generated a bit of interest. The overall participation rate (black line) plunged, generating some stories about potential workers fleeing the labour force.

Alexander Bick at the Saint Louis Federal Reserve wrote a short article dissecting the plunge in the participation rate. There were three factors found:

There was a statistical artefact that resulted from a benchmark changes of population estimates. Last year’s changes raised the participation rate, this year’s changes reversed that.
There is an effect due to the ageing population.
There remainder was a dip that represents less than half of the decline of the participation rate that might be a concern.

The ageing population issue is best illustrated by contrasting the whole working age population participation rate to the prime age (25-54 years) participation rate (red line). The prime age strips out most students, as well as oldsters (like myself) who are drifting towards qualifying for an old age pension. Although the prime age participation rate has dipped, the movement is less dramatic. The level gap in July was 22% (83.4% versus 61.4%) which explains the importance of demographic shifts out of the prime age population.

The non-excitement of the change in the participation rate is not perhaps itself of great interest, but I think it provides a justification to dig out the charts showing differences between the prime age and (total) working age populations. When the baby boom demographic was prime age, what was happening in the oldster cohorts did not have a lot of statistical weight. We now have a lot of weight on a population that is quite happy to drop out of paid employment if given the opportunity. This might change some intuitions about labour market behaviour. For example, it was a standard belief that a hot labour market will pull in people who stopped participating (thus giving the economy greater room to grow before hitting capacity constraints). Now, it may be that if things go well, some people will take advantage of the situation and retire earlier.

One could argue that the decline in the employment ratio (which I like to look at) is just downstream of a declining participation rate due to an ageing population. My response is that the participation rate is going to be a mushy variable, as it is now being influenced by near-retirees who are going to act in ways that do not conform to earlier eras where the prime age demographic was dominant. (Correspondingly, the unemployment rate loses some of its value.) The employment ratio is telling us how many people are working, which directly relates to the business cycle.

Although these effects should matter for determining what is happening to labour market constraints, I still feel that this is likely going to be a second order inflationary effect when compared to the flow of oil through the Strait of Hormuz being throttled down.

Email subscription: Go to https://bondeconomics.substack.com/