The Economist has an article discussing the issue of forward guidance in monetary policy. Here’s an excerpt:
To guide expectations credibly, officials must eventually follow through with the changes they indicate. The quandary is deciding what to do when conditions change, as they have since the Powell pivot, with inflationary pressure stronger than expected, which has rendered rate cuts less suitable. Staying the course might no longer be appropriate; changing it risks harming central bankers’ ability to jawbone investors in the future. . . .
Ben Bernanke, a former Fed chairman, once warned that such considerations can quickly degenerate into a “hall of mirrors”. If policymakers mimic market expectations, which then shift as a result, endless distortions are possible. Suitably enough, Mr Bernanke’s more recent work reviewing the Bank of England’s approach to forecasting offers a way out, suggests Michael Woodford of Columbia University. One crucial recommendation was that the bank ought to start publishing its projected policy rate under a range of different economic scenarios, rather than just its central forecast. Doing so would help investors understand how policymakers would react to different conditions, allowing them to change course in response to new data without losing face.
In my view, making interest rate forecasts conditional on macroeconomic conditions is an improvement over unconditional forward guidance. Unfortunately, it is difficult to predict how changing macroeconomic conditions might impact future movements in the natural rate of interest.
An alternative would be to offer more specific guidance as to the policy goals of the Fed. For instance, one could imagine a nominal GDP target that calls for 4% annual growth, with make-up policies to correct any short run deviations from this trend line. This sort of policy regime is called “NGDP level targeting”, because it targets the level of NGDP, not the growth rate.
Even more precise guidance could be provided by specifying the exact nature of the make-up policy. For instance, the Fed could indicate that the make-up would occur at a rate of 1%/year, until back on the trend line. Thus if a mistake pushed NGDP 1% above the target path, the Fed would aim for 3% growth over the next 12 months. If a mistake pushed NGDP 2% above target, the Fed would aim for 3% NGDP growth over the next two years. If NGDP fell 1.5% below target, the Fed would aim for 5% NGDP growth over the next 18 months.
One advantage of this sort of policy regime is that it would make it easier to interpret the information in interest rate futures markets. Today, policymakers don’t know whether an anomalous movement in fed funds futures reflects expectations of what sort of future interest rate would be required to achieve 4% NGDP growth, or a lack of confidence that the Fed is actually trying to achieve 4% NGDP growth. To make the point more concrete, if the fed funds futures show rates falling to 3.5% over the next year, is that because markets expect a weaker economy, or is it because markets expect an easy money policy that will trigger a stronger economy?
I’ve advocated a “guardrails” approach, where the Fed would take unlimited short positions on 5% NGDP growth futures contracts and unlimited long positions on 3% NGDP growth futures contracts. But even if this market-guided policy regime is politically infeasible, a clearer statement of the Fed’s desired path for NGDP growth would lead to an environment where existing financial markets could provide a rich source of information to policymakers struggling with the question of where to set their interest rate target.
I believe that setting a clear NGDP level target would lead to much less volatility in NGDP growth over time. Indeed, an NGDP level targeting regime with a clearly specified make-up rule would likely have allowed us to avoid a severe recession in 2008-09, and a severe inflation overshoot in 2021-22.
READER COMMENTS
Thomas L Hutcheson
Jun 12 2024 at 9:45am
Forward guidance should be about the outcome the central bank is pursuing, NOT the instrument settings that it expects to use in achieving those outcomes.
At any decision point, the expected probability of adjusting some policy variable “up” or “down” should be the same. The driver should create the expectation that they will follow the road however it twists and turns, not that they will probably turn lest or right at the next curve.
A corollary of this kind of forward guidance is that a central bank must be willing to reverse the setting of any instrument from up to down more down to up from one decision point to another. There should be no information about future instrument setting from past instrument settings.
As far as I can see, it would help if there were nor decision points, weekly if not daily resetting (or not) of policy instruments.
Rajat
Jun 12 2024 at 4:38pm
I completely agree with your comment that while making interest rate moves conditional on macroeconomic conditions would be an improvement, it’s difficult to predict how changing macroeconomic conditions might impact future movements in the natural rate of interest. In place of a ‘hall of mirrors’ metaphor, I would say that such an approach involves the Fed ‘chasing its own tail’. The approach takes the markets as mugs – by treating the market as too dumb to understand what is really going on, which is that the Fed is continually updating its own interest rate reaction function instead of being transparent about what it is trying to achieve and in what timeframe. I’ve never been a fan of forward guidance for that reason.
The trick with a level target, as you indicate, is defining how it addresses the same issues – the timeframe for achieving the Fed’s ultimate goal. In Australia, the RBA (an inflation-targeter, not a level-targeter) is giving itself until 2026 for inflation to get back within its target range. That is the better part of 5 years of above-target headline inflation and 4 years of above-target core inflation. That seems far too long. Yet if Covid had been much more severe, we may consider that conformance to a growth target within 4 years (and say, a level target within 5 years) was just fine or even too fast. I think this is the issue that stops serious central bankers – ie not those just trying to avoid accountability – from wholeheartedly adopting level targets. It’s the risk of a major shock like Covid, or in their minds, the Great Financial Crisis, even though that doesn’t fall in the same (real) category. Loosening or changing the timeframe for hitting a level target is probably less as more radical and credibility-destroying than loosening or changing forward guidance. Or at least, that’s an explanation that seems plausible – but not persuasive – to me.
Rajat
Jun 12 2024 at 4:41pm
The second-last sentence should start: “Loosening or changing the timeframe for hitting a level target is probably seen as more radical…”
Scott Sumner
Jun 12 2024 at 7:25pm
Even worse, the RBA is not level targeting. They are merely trying to get the rate of inflation back to 2%.
Andrew_FL
Jun 13 2024 at 9:13am
What one should have learned over the last three years is that in practice, the last scenario might happen but the former two would never, under any circumstances, be allowed to happen.
Comments are closed.