The Bank of Canada vs the bond market

Carolyn Wilkins, Senior Deputy Governor at the Bank of Canada, gave a speech yesterday at noon. She said that the Bank of Canada had lowered its estimate of the (cyclically-adjusted) neutral rate of interest.

"All told, we think that the neutral rate of interest is lower than it was in the years leading up to the crisis because of these structural developments. We estimate that the real neutral policy rate is currently in the range of 1 to 2 per cent. This translates into a nominal neutral policy rate of 3 to 4 per cent, down from a range of 4 1/2 to 5 1/2 per cent in the period prior to the crisis."

Her speech was based on analysis in a Bank of Canada discussion paper by Rhys Mendes. (I can't tell whether that discussion paper was made publicly available at the same time.) [Update: Rhys emailed me to say it was released at the same time.]

I don't see much reaction from the bond market to the speech. Yields were down very very slightly (about 2 basis points) compared to Friday's close.

But what puzzles me is this: the 30-year bond is currently yielding 2.74%. That's below Carolyn's 3% to 4%. And Carolyn is talking about the Bank of Canada's target for the overnight rate (the "policy rate"), and that's normally below the 30-year bond yield.

On average over the business cycle, we would expect the Bank of Canada's policy rate to equal the (cyclically-adjusted) neutral rate. And we would expect the 30-year bond yield to be above that neutral rate on average. Now today isn't average, because the Canadian economy is still recovering, and the world economy is not looking good. But the bond market is supposed to anticipate the future path of interest rates, and 30 years is a long time, relative to most recessions.

Either the bond market thinks it will take a very long time for the Bank of Canada to get back to the neutral rate, or else the bond market thinks that the neutral rate is lower than the Bank of Canada thinks it is.

I don't know who's right.

The disagreement wasn't always this big. At the beginning of the year the 30-year bond was yielding 3.2%. It then fell steadily to 2.6% at the end of August, then started rising again.

Graph from the Bank's website:

Government of Canada Benchmark Bond Yields - Long-Term

GRAPH PERIOD: 20 September 2013 - 22 September 2014

Comments

[PelinoC]

By historical standards current long bond rates are crazily low. This seems to suggest that either: the capital markets are terrified of risk and are willing to sacrifice long-term returns in order to park their money in an asset class which will not suffer absolute nominal declines; or capital markets believe we have entered a period akin to the 19th century, where there will be essentially no inflation and nominal returns will be much closer to real returns; or capital markets believe that over the foreseeable future there is a reasonable chance that a deflationary period will cancel out any expected inflation (leading to the same overall result as the no inflation option). Personally, I think that the first option plays a bigger role in the truth...

[Lorenzo from Oz]

Plus the rate of savings by the expanding in size-and-wealth Chinese and Indian middle classes are constantly "pushing against" investment options and so pushing down the price of capital in a low inflation environment.

[=Peter]

Is the BoC's lowering of its "estimate of the (cyclically-adjusted) neutral rate of interest" the same topic as (or the Canadian version of) the issue of "low real interest rates" discussed in Vox EU's "Secular Stagnation: Facts, Causes, and Cures" e-book?

(http://www.voxeu.org/article/secular-stagnation-facts-causes-and-cures-new-vox-ebook)

E.g. "A bouquet of contributions [in the book] focuses on why *equilibrium real interest rates* have fallen steadily over the past few decades" (p10, emphasis added).

Is this evidence that Canada is also (possibly) facing Secular Stagnation? (Are there good estimates of Canada's Potential GDP out there? Has it been leveling off recently?)

=Peter

[Brian Romanchuk]

Expectations work reasonably well out to 10 years on the curve, but market conventions take over beyond that. That is, the 30-year is priced as a spread versus tne 10-year.

This is also affected by technicals. There is a massive need for long-dated duration to match actuarial liabilities, so long bond yields are surpressed.

But othe other hand, estimates of "neutral" rates have been totally wrong for the past few decades. Realised average short-term rates are far below what people thought "neutral" was. It can just be the case that the bond market has caught up to that reality, but the BoC hasn't.

[Andy Harless]

"And we would expect the 30-year bond yield to be above that neutral rate on average"

Would we, necessarily? The presumption seems to be that the marginal investor has a short horizon and therefore faces more capital risk than reinvestment risk. But can we always make that presumption? (Alternatively, one could argue that inflation risk should make long bond yields higher, but even if inflation risk is a significant issue, it's not obvious that it's a bigger one than reinvestment risk.)

[Phil Koop]

It is absurd to conflate risk-neutral and real-world probability measures this way. Yes, we are used to thinking of risk-neutral expectations of future rates as being higher than real-world expectations, but it is an empirical fact that this relationship often does not hold.

You must pick your poison: accept that the risk "premium" can be negative or accept that risk does not entirely account for the difference between the pricing and real-world measures. Remember that there is no truly risk-free asset to use as numeraire!

[Nick Rowe]

Brian, Andy, Phil: OK. Good point. And demographics might have made that risk premium go negative. Which makes me wonder whether it would be an even better idea now to change the law to make 30 year closed fixed rate mortgages possible, to increase the supply of long financial assets. My old post.

Peter: "Is the BoC's lowering of its "estimate of the (cyclically-adjusted) neutral rate of interest" the same topic as (or the Canadian version of) the issue of "low real interest rates" discussed in Vox EU's "Secular Stagnation: Facts, Causes, and Cures" e-book?"

Basically yes. The Canadian (real) interest rate, in the longer term, is mostly made in world markets, and not made in Canada much..

[SvN]

I'm confused. I think Wilkins is talking about the current value of the neutral rate. I assume that the 30-yr bond rate will depend (among other things such as those noted above) on how the neutral rate is expected to evolve over the next 30 years.

[Jeremy]

Your 30 year reference rate isn't exactly the best rate to use for the purpose of determining the market's estimate of the longer term neutral rate. It seems likely given current economic conditions that the overnight rate will stay low for some additional period of time. To remove the influence of current economic conditions and our current point in the cycle you should use a forward rate, using the 30yr rate 2 years forward shifts your chart up about 20 basis points, adding more time moves that rate up. It might even be better to look at the 1 year rate 29 years out to gauge the bond market's estimate of the rate of interest in the 30th year but that's fraught with other problems.

Wilkins' estimated range isn't so different from current pricing in the forward market.

[Frank Restly]

I have never bought that long term interest rates on government debt are totally a reflection of anticipated changes in short term interest rates set by the central bank.

First, even without a central bank, a term premium exists between lending short term versus lending long term. Part of that term premium is default risk (essentially zero for government debt) the other part of the term premium is lost opportunity risk - by lending now for a long time I may give up better opportunities to lend later on better terms. The ability to lend later on better terms depends on how liquid the long term bond is.

If long term bonds are highly illiquid, then the term premium increases. If they are very liquid, then the term premium decreases.

Second, setting aside term / liquidity premium, there is still the relative supplies of long term versus short term government debt to contend with.

30 year Treasury bonds are a relatively new invention in the U. S. They were first sold in 1978 I believe, sales were suspended in 2002, and then renewed again in 2006.

[Nick Rowe]

Simon: OK, that would be another way to reconcile the two, if the bond market thinks the neutral rate will drift down over time.

Jeremy: yep. But the 10 year rate is at 2.20%. If it takes 10 years to get back to normal, that puts that normal rate at around 3.00% (back of the envelope). It's still at the low end of Carolyn's range.

[rab]

In olden times the curve would lead the economy. Now the economy (slower) will lead the curve (lower).

[Frank Restly]

Rab,

I wonder if that is a function of how indebted a government is relative to size of the economy - low debt to GDP = curve leads economy, high debt to GDP = economy leads curve.

[Jason Smith]

Another possibility hinted at above by PelinoC and Lorenzo from Oz is that the Bank of Canada will be expected to undershoot its 2% inflation target ...

http://informationtransfereconomics.blogspot.com/2014/09/does-canada-know-about-information.html

[Brian Romanchuk]

I looked at it, and relative value considerations do a very good job of explaining the level of the 30-year. I put the analysis up on my site.

[Market Fiscalist]

"We estimate that the real neutral policy rate is currently in the range of 1 to 2 per cent. This translates into a nominal neutral policy rate of 3 to 4 per cent"

Question: If the Deputy Governor is correct why are nominal rates not currently at 3-4% in Canada (I assume they are not) ?

[Majromax]

> Question: If the Deputy Governor is correct why are nominal rates not currently at 3-4% in Canada (I assume they are not) ?

Because the current rate is not supposed to be neutral, but instead expansionary.

[The Market Fiscalist]

Majromax,

You say: "Because the current rate is not supposed to be neutral, but instead expansionary."

If they know what the neutral rate is , and that the actual rate is currently being set way below it (because its meant to be expansionary) - why aren't they additionally creating new money like crazy to allow them to raise the actual rate to match the neutral rate much sooner than on current trajectory ?

[Majromax]

Because that's a contradictory use of instruments?

The Bank of Canada operates by controlling interest rates, and it creates or destroys money so maintain effective rates close to its target rate. If the BoC says that the target rate is 1% (as it is now), then it's restricted in its ability to "create new money like crazy."

Doing the latter would require a short-term rate target below 1%, and the BoC is deliberately choosing not to change its target rate. Presumably, it feels that slower adjustments are inherently better than an interest-rate whipsaw.

[TallDave]

It will be interesting to see what happens over the next year.

Also, curious about something from your earlier post:I basically agree with David. Monetary policy *can* offset fiscal policy, and US conservatives need to get that, and support a better monetary policy.

Do you see less of this foolishness from Canadian conservatives generally? Harper at least seems more enlightened than any major conservative figure in the United States, I thought it was interesting he was being criticized for a monetary policy that was too loose:

http://www.thestar.com/news/canada/2014/01/27/former_conservative_mp_critical_of_harper_government_over_monetary_policy.html

Unfortunately libertarians in the United States tend to be somewhere between "the Fed is printing way too much money" and "back to the gold standard!" (headsplode).

[rp1]

Long term rates are often dreadfully wrong at the end of a secular move. Are we there yet? If so, whew. If not, deja-vu.