I hold the Czech National Bank in high regard relative to other central banks in the world, but I do think they have fallen a little bit prey to the baseline mentality and models with linear approximations. I think the analytical period over the past few quarters illustrates this when reviewing their current analysis, scenario construction, and forward guidance communication. Of course, I don’t have skin in the game so who knows maybe I wouldn’t be so critical if I were in the trenches.
Priors up front, I don’t think the Czech economy is anchored to 2% inflation, too many variables suggest its a real possibility. And reading the CNB’s past couple of Monetary Policy Reports (Spring and Summer) and current analysis around incoming data feels way too complacent with this risk lurking.
It’s something I’ve been concerned about in the post-COVID-19 period and a scenario that hit home when the pub next door to my apartment raised their beer prices 30% (from 48czk to 62) in December 2023. It was a bit weird that they waited so long to raise their prices but I’ve long suspected that beer is the unofficial numeraire of the Czech economy. Then almost every month since then there has been something new that I’ve noticed got a major price jump whether it was a recreational activity, school supplies for kids, public transport, housing costs, insurance, hotels, etc. Mind you we are 2 years removed from when inflation peaked, and I dont know about other people but I still see it. Before COVID-19, I’m not old enough to have ever worried about inflation because price stability was the status quo and the whole point of price stability is that you’re not meant to recognize inflation. For the first time in my life I can finally understand the kooks that lived through the 1970s and describe how high inflation infects the economy gradually for years without too much fanfare if not sufficiently managed. It’s never all at once, it’s a slow boil, and these little reminders here and there lend me to believe that price stability still isn’t here and that the CNB and most other advanced economy CBs, still have more work to do and should be resisting lower rates getting priced in financial markets before the job is complete and maybe even embrace a little undershooting of the target if that’s what it will take.
Complacency is a general critique. A more specific criticism of the CNB’s analytical framework is the construction and use of scenarios. The main purpose of scenarios is to be able to tell different stories and these different narratives provide a structure for interpreting incoming data to answer the question: what state of economy are we living in? Going into 2024 the different states in question were whether the economy was anchored to 2% inflation or not? And to the credit of the CNB they produced these 2 scenarios in their Winter 2024 MPR: A baseline where inflation was assumed to be anchored and an alternative scenario with elevated inflation expectations. The problem is that using these scenarios and comparing them with the data that have come in so far this year, it’s difficult to discern which state of the economy the data supports.
Here are some charts showing the evolution of data and different scenarios prepared by the CNB with some near-term assumptions for illustration purposes:
Figure 1. Modest upside assumptions for inflation for the rest of the year would flip the price stability/monetary policy is restrictive narrative on its head and be more in line with a scenario where inflation is not well anchored. The Baseline projection hinges on seeing historical seasonal inflation patterns returning. For example, September usually sees deflation of about -0.4% MoM and inflation tends to remain low at the end of the year before the annual repricing in January.
Figure 2. However, why might we not be in normal times and why would inflation continue to surprise to the upside? Culprit number one is a tight labor market which could easily be much tighter than currently assumed by the CNB given the uncertainty around NAIRU. Has the CNB done enough to kill excess labor demand when looking at vacancies? Maybe. Could large net migration flows play a role? Probably wont be able to come to a definitive answer but the risks around a 4% assumption definitely seem skewed to the higher end.
Figure 3: Wage growth decelerated in 24Q2 but still higher than the Winter 2024 elevated inflation expectations scenario, but this doesn’t seem to have triggered the concern it deserves, perhaps in part because elevated wage inflation has been factored into the baseline projection. But this just calls into question what is the point of alternative scenarios if there is no way to differentiate between the two scenarios when new critical data are presented that are inconsistent with the central bank mandate?
So maybe you can still defend the baseline but you can also definitely build a convincing argument that the elevated inflation expectations scenario is materializing which makes the complacency critique more of a headscratcher since the CNB should be prepared for this but the main communication I am getting is that the program continues as planned.
I think this gets to the heart of what is the purpose of doing multiple scenarios in terms of monetary policy analysis and communication? My sense from the CNB is that they are produced as afterthoughts and don’t carry much weight from projection round to projection round to structure the current analysis. Could be wrong but if they were more important than you would expect that their beautiful Winter 2024 elevated inflation expectations scenario would have continued to feature as a framing device to discuss current analysis around elevated wage growth, low unemployment and upward inflation surprises.
Meanwhile, the baseline projection lacks a purposeful macro narrative in the sense that for the first year of the scenario, there are several variables that could easily be consistent with an economy where inflation is unanchored. A low unemployment rate rising only to 4% (3.8% now), wage inflation remaining elevated (over 7% for the first year) and real GDP growth expected to go to 3% and above. Meanwhile, non-traded inflation remains high (>4%), credit growth is high (>6%), and equity prices are on the moon (up 18% over the past year). The only thing holding the baseline scenario together is headline inflation, but we can also start seeing the cracks form there as well now. All of it is pretty suggestive that something is fundamentally not right here but somehow can still be defended as being broadly in line with the baseline scenario which really hinges on the 2nd year of the projection to see noticeable signs that the economy is on a stable path back to equilibrium.
Instead, under the Forecasting and Policy Analysis System (FPAS) Mark 2 as practiced by the Central Bank of Armenia, the staff abandon the baseline projection and consider important non-linearities such as imperfect policy credibility. They are expected to produce at least 2 scenarios (Case A and Case B) that challenge the prevailing narrative that is currently priced in financial markets from opposite perspectives. The pricing in financial markets in terms of consensus forecasts for real GDP, inflation, interest rates serve as the baseline scenario for the central bank. Under FPAS Mark 2, the staff no longer need to engage in the analytical monstrosity of a baseline scenario that is always going to be a conglomerate of different views and opinions that serve no real purpose for current analysis other than to cloud an effective strategy.
If we apply FPAS Mark 2 to the CNB in 2024, the Case A scenario where interest rates need to be higher than what is priced in financial markets would be the Winter 24 scenario with elevated inflation expectations and this scenario would have been followed up on throughout the year to frame the incoming data. The Case B scenario would be a negative external demand scenario via Germany or China and suggestive the rates need to go lower.
Figure 4: Now what are the policy implications under FPAS Mark 2? We are closing in on 3 quarters removed from the Winter 2024 elevated inflation expectations scenario yet monetary policy has been re-affirming a lower expected path of the policy rate priced in financial markets where inflation is assumed to be anchored.
Elevated wage inflation, upward surprises to CPI inflation, and low unemployment, it’s clear the messaging should be that we are moving in a Case A direction. Trying to fit the latest data to be broadly consistent with the baseline scenario is a recipe for losing the plot on how the CNB intends to achieve its objectives under a highly uncertain macro environment.
What should the strategy be today, yesterday? The current view I can get from the communication of various board members is they believe monetary policy is restrictive and if inflation progress slows that they can simply stop rate cuts and still be confident that monetary policy is restrictive. This is a dangerous confidence in my opinion and where the linear mindedness kicks in. Once higher underlying inflation becomes embedded in the economy it is a different beast and likely requires a non-linear response from monetary policy. I think its past due to start taking out some insurance on higher underlying inflation and start nudging market interest rates higher until we are more confident that inflation is anchored (credit to Board member Eva Zamrazilová who has been consistently expressing these types of fears in the minutes this year). But this of course is a trade-off decided by the CNB Board but I know I am tired of above target inflation where price increases are still top of mind and so postponing true price stability probably isn’t the end of the world but it would be disappointing and leaves the country vulnerable to future inflation shocks where anything can happen.
In my recent analysis on the US I ended it with how the existence of Bark Air, the airline exclusively for dogs was a sign that financial conditions were not tight enough and an indicator of a equity bubble and the Fed may want to lean against the wind a little bit. My Czech version of this is not as fun but I do have stories of the economy handing out auto loans like confetti for brand new BMW’s to people with highly speculative income. I just think more can be done.
Kommentare