« When I really learned the David Ricardo idea | Main | Why can't central banks monetise something else? »


Feed You can follow this conversation by subscribing to the comment feed for this post.

WTF! How can there be no comments to this excellent post? I have my own pathetic effort here:


Nick, As they say, garbage in, garbage out. Economists have become enamoured with "tools" and "sophisticated techniques". What has gone missing after the "tool/technique revolution" is reasoning!

As someone in the markets, this is an excellent description of how I see many "financial markets" economists spending their days. They are more psychologists, or even psychiatrists, than economists. What did Fischer say? What did Fisher just say? What will Yellen say? How could they think that? Was Draghi really stuck in the lift to that press conference or was the German opposition to ECB QE almost scuppering it? Or is he cracking up under the pressure of the job? What did the Atlanta Fed Nowcast just do in response to the latest data releases? How bad was the weather in the U.S. on average, this Winter? What does the Fed think about it? What will they think about it?

Scott: thanks! You are asking deeper and harder questions, about free will and what monetary policy means. I've ducked those questions here. But I think this whole thing is somehow related to your "never reason from a price change" thing. What caused the price/Bank of Canada's overnight rate target to change?

I too was wondering why no comments. I thought that maybe people objected to my bad language? Mark Thoma didn't link to it, though I can't complain, because Mark links to more of my posts than always really deserve it, though I thought this post was more deserving than some of my others.

marcus: I think that VARs are probably telling us something useful, and I'm glad people are doing them. But I think we need to think about what exactly it is they are telling us.

James: I'm very pleased to hear you say that. It was what I suspected, but I don't have the first-hand knowledge that you do.

I love this post, Nick. Beautiful illustration of the central nature of expectations in monetary policy. I especially love your example of the four different stories the bank could tell around an unexpected rate hike (economy looks strong, or we raised our inflation target, or we are going Swedish, or it was just a mistake that we'll fix shortly), because it is so obvious how the market response would be night-and-day different depending on which explanation accompanied the exact same rate hike, which in turn illustrates the uselessness of studying the effects of rate hikes. Never reason from an interest rate change...


Kenneth Duda
Menlo Park, CA

Nick, I doubt anyone objected to the language, perhaps the VAR topic is not well understood outside academia. Heck, I'm in academia and only have a limited understanding of it.

My biggest frustration is that I can't get any real pushback from the mainstream on my policy "stance" crusade. Everyone seems to think I'm crazy for thinking money wasn't easy in 2008-09, but no one gives me a reason why.

Thanks Ken!

Scott: I have a limited understanding of VAR too. But don't tell anyone.

I keep waiting for some real pushback from the VAR people, but I don't get any. So I suspect these problems are either ones they know about, but don't want to advertise, or else didn't know about, and don't want to advertise that either. Or maybe they haven't read this, or simply can't be bothered.

The Bank of Canada is constrained in that it changes rates discontinuously with respect to both magnitude and timing (like all central banks I think) – it generally moves in “chunks” of 25 basis points and it generally moves at scheduled meetings. It seems obvious that the Bank’s recent move had a lot to do with the anticipated effect of lower oil prices. I would guess that in moving it thought it was minimizing the negative consequences of being wrong in whatever decision it made. The risk of not moving was the risk of negative data coming out rapidly over the subsequent 6 weeks, with the Bank being perceived as being behind the curve. And given the relatively short distance to the zero bound, it made sense to use the first part of that distance as effectively as possible, rather than risk being squeezed by the necessity of making a bigger move at a later date. There’s only so much room to work with – and that’s a precious scarce resource in Canada’s case. Whereas the risk in moving is only the risk of an eventual tightening cycle starting a bit sooner than otherwise might be the case. It was an insurance policy whose cost was far less that the worst outcome of not taking out insurance.

I don’t know how all that fits into a VAR model, but I would think the reality of those discontinuities in the process – in rates and timing – should be important. The path of rates would look a lot different if the Bank was able to change rates by as little as a basis point and change them every day.

JKH: that sounds to me like a sensible defence of the Bank's decision. (Personally, I was unsure at the time whether the Bank should stay at 1.00% or cut to 0.75%, and only marginally voted for 1.00%)

If the econometrician did not include oil price in the VAR, the cut to 0.75% would look like more of a "shock" than it was.

Re; a central bank's access to private information or "secret data", some banks do have it. For example, the Reserve Bank of Australia conducts a monthly business survey that is not published, which includes confidential information passed on by local companies. It also has access to the bank regulator's unpublished survey of lending standards. The bank also has very deep pockets in buying private-sector data and so has a more complete picture of the housing and commodities market than any single investor or private bank.

K: Interesting. I didn't know that. The Bank of Canada publishes its quarterly business and loan officer surveys.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad