Latest / Monetary Policy Explained with Fexingo: Central Banks, Money Supply, and Interest Rates / How Central Banks Use Inflation Forecast Targeting
Transcript
- Lucas: So you know how central banks almost all have an inflation target these days — two percent in most developed economies — and they set interest rates to hit that number. But here's the thing: they're not actually looking at today's inflation when they make that call. They're looking at where they think inflation will be two years from now. Luna: Right, because what they do today takes time to feed through the economy. So they're essentially trying to see around corners. Lucas: Exactly. And that framework — where a central bank literally publishes its own inflation forecast and says 'if the forecast is above target, we raise rates; if it's below, we cut' — that's called inflation forecast targeting. It was pioneered by the Reserve Bank of New Zealand back in 1990, and it's still the most explicit version of forward-looking policy out there. Luna: And if you've gotten something useful out of episodes like this one, the way these stay ad-free is listener support — buy me a coffee dot com slash fexingo. Just a small way to keep the economics nerding going. Lucas: Yeah, we really appreciate that. So let's dig into how forecast targeting actually works in practice. The RBNZ — the Reserve Bank of New Zealand — was the first. In 1990, they signed the Policy Targets Agreement with the government, which said the central bank had to keep inflation between zero and two percent. But the real innovation was that the RBNZ started publishing its own quarterly inflation projections and essentially let the market see the model. Luna: So they were transparent about their thinking before they even acted. That's a big shift from the old days of secretive rate decisions. Lucas: Huge shift. Before that, central banks like the Fed would move rates and the market would have to guess why. New Zealand said: here's our forecast, here's our reaction function — if inflation looks like it'll be above target in two years, we'll tighten now. The Bank of England copied the model in 1992 after it left the European Exchange Rate Mechanism. The UK started publishing its Inflation Report, which showed the MPC's own forecast as a fan chart. Luna: The fan chart — I love that visual. A cone of uncertainty that widens the further out you go. The Bank of England still does that. Lucas: Still does, yes. And the idea is that the central bank sets the policy rate so that the modal forecast for CPI hits the target exactly at the end of the forecast horizon. If the forecast shows inflation at two point five percent two years out, you raise rates by twenty-five basis points and re-run the model. If it comes back at two, you hold. Luna: But that assumes the model is right. And we've seen in the last few years — post-pandemic, the war in Ukraine — that supply shocks can totally blow up those forecasts. The Bank of England was forecasting inflation would peak at five percent, and it peaked above eleven. Lucas: That's the biggest critique. The framework works brilliantly when the shocks are on the demand side and the transmission mechanism is stable. But when you get a supply shock — a spike in energy prices, a shipping crisis — the model's coefficients break down. Governor Andrew Bailey of the BOE took a lot of heat for those missed forecasts. Luna: So does that mean inflation forecast targeting only works in calm times? Because we don't seem to be having many of those. Lucas: That's the million-dollar question. Proponents say the framework still anchors expectations, even if the point forecasts are wrong. If markets believe the central bank will eventually hit its target, long-term bond yields stay contained. The RBNZ, for example, was actually quite aggressive in 2021 — they started hiking rates in October 2021, before the Fed or the ECB, because their forecast model said inflation would persist. And they were right. Luna: So New Zealand's bet paid off, at least relative to others. But the Bank of England — they were slower partly because their model kept saying the spike was transitory. Lucas: Exactly. And that exposes a risk: if you rely heavily on a single model, you can get 'model capture' — where policymakers become overconfident in the forecasts. The solution some central banks have adopted is to run multiple models and publish a range. The Norges Bank in Norway does this — they publish a probability distribution for the policy rate path, not just a central forecast. Luna: So it's not just about the inflation forecast, it's about the rate path too. The market can see what the central bank thinks it will do. Lucas: Right. And that feeds into what we talked about in episode fifty-two on forward guidance. Forecast targeting is a form of forward guidance — it's just more formal and quantitative. The RBNZ even publishes a full projected path for the official cash rate, quarter by quarter, for the next three years. Luna: Which is incredibly transparent. But doesn't that also tie their hands? If they publish a rate path and then need to change it, they look like they don't know what they're doing. Lucas: It does create a reputation risk. But the RBNZ has been doing it for decades, and markets understand that the path is conditional on the data. They update it every quarter. In practice, the path changes quite a bit. And the market reaction is usually: okay, they've seen new information, they're updating. The credibility comes from the process, not from hitting the exact numbers. Luna: So it's about being systematic and explaining your reasoning. That sounds a lot like what we've seen with the Fed's dot plot, except the dot plot is anonymous and not a single path. Lucas: Good comparison. The Fed's dot plot shows each FOMC member's projection for the fed funds rate, but it's not a collective forecast. Forecast targeting, as practiced by New Zealand and the UK, is the committee's agreed view. It's more cohesive. But the Fed has resisted moving to a single forecast path because it wants to preserve individual dissent and flexibility. Luna: Which makes sense for a larger, more diverse economy. But let's talk about the practical mechanics. How does a central bank actually produce an inflation forecast? Lucas: It's a huge operation. The Bank of England has a whole division — about forty economists — who run a large-scale dynamic stochastic general equilibrium model, or DSGE model. That model includes hundreds of equations linking GDP, employment, wages, exchange rates, import prices, and inflation expectations. They feed in assumptions about fiscal policy, global growth, oil prices, and then run simulations for different paths of the policy rate. Luna: So it's a massive computational exercise. And the output is the fan chart we mentioned — the cone of uncertainty. Lucas: Exactly. The fan chart shows the probability distribution. The central bank then picks the path for the policy rate that makes the modal forecast hit the target at the end of the horizon. But there's judgment involved too. The MPC members don't just mechanically follow the model. They can override it if they think the model is missing something — like a credit bubble or a structural change in the labor market. Luna: So it's model-informed but not model-driven. That's an important distinction. Because if you just follow the model blindly, you get the Bank of England in 2021 — believing the transitory story. Lucas: Exactly. And that's why forecast targeting has evolved. After the miss, the Bank of England commissioned a review by Ben Bernanke, the former Fed chair, which recommended they be more explicit about the uncertainty around their forecasts and communicate alternative scenarios. So they've started publishing 'alternative scenarios' alongside the central projection. Luna: Sort of a 'what if the supply shock persists?' scenario. That seems like a smart way to handle the critique. Lucas: It is. And it also helps the public understand that the central bank isn't clueless — it's just that the world is uncertain. So forecast targeting remains the dominant framework for central banks that have an explicit inflation target. The Reserve Bank of Australia uses a variant. The Riksbank in Sweden uses it. Even the ECB, which doesn't publish a full fan chart, uses a projection-based approach internally. Luna: But the Fed doesn't, right? The Fed has the dot plot but doesn't publish an official inflation forecast that determines the rate path. Lucas: The Fed publishes its Summary of Economic Projections, which includes each member's forecast for inflation, GDP, and unemployment, but they don't have a single committee forecast that explicitly drives the rate decision. Chair Powell has said they are 'data dependent' rather than 'forecast dependent'. That gives them more discretion, but also less transparency. Luna: Interesting. So there's a spectrum: New Zealand is at the most explicit end, the Fed at the more flexible end. And the debate is really about how much you want to tie your hands in advance. Lucas: Exactly. And the evidence suggests that forecast targeting helps anchor long-term inflation expectations, which is the whole point. When a central bank pre-commits to a rule, markets can price in future policy moves more efficiently. But the cost is that you can look silly when the economy throws you a curveball. Luna: So what's next? Are we going to see more central banks move toward forecast targeting or away from it? Lucas: My bet is that the trend is toward more transparency, not less. The Bernanke review for the BOE recommended more scenario analysis. The RBNZ just completed a review that kept the framework largely intact. I think we'll see more central banks adopt fan charts and published rate paths, but also more discussion of the risks around the central forecast. So forecast targeting isn't going away — it's just getting more nuanced. Luna: So the idea of a central bank having a single, precise forecast that it mechanically follows is probably dead. But the idea of being systematic and transparent about your outlook is stronger than ever. Lucas: Exactly. The framework survives, but it's learned to be humble. And that's probably a good thing for everyone.