In the cacophonous analytical scrutiny of, and political volleyball over the efficacy of the government’s recent marathon five-part series of measures, an important abdication of responsibility by the monetary policy committee (MPC) has been paid almost no attention. While unexpectedly announcing an outsized cut in the policy repo rate for the second time in less than two months, the MPC last week shied away—yet again—from offering a number (or range) forecast for India’s real GDP growth and CPI inflation. The decision is disappointing and shows the Reserve Bank of India (RBI) in poor light, relative to its own institutional history and in comparison with other forecasters, including central banks.
Economies, like weather systems, are complex frameworks. However, unlike weather systems, economies rely on human behaviour that isn’t always easy to map accurately, even in the best of stable conditions. Supply and/or demand shocks can alter that behaviour quickly and, sometimes, the changes can last for extended periods, if not permanently.
Public projects: Why states should go beyond ‘heads I win, tails you lose’Public projects: Why states should go beyond ‘heads I win, tails you lose’ISRO’s love-hate relationship with private sector – A look back at historyISRO’s love-hate relationship with private sector – A look back at history
It is in the above context that macroeconomic forecasting remains an evolutionary process that is a subjective mix of science and art. Assessments about the economic outlook are affected by data inputs, and the trustworthiness of economic models. That is why reliable high-frequency economic data and surveys have a useful role in economic modelling, and successive forecasts are best seen as an evolving economic tapestry. But, for all the advancement in the tools and techniques used, forecasting is an inexact science—an economic model could be high on complexity and sophisticated mathematical response functions, but the forecaster’s artful judgement, honed by experience, still matters in the final tally.
For the uninitiated, the Reserve Bank’s forecasting team is responsible for the macro forecasts used by the MPC. The latest policy statement is missing the fan charts and the numerical forecasts for inflation and real GDP growth for the second consecutive meeting. Following the late-March measures, Governor Shaktikanta Das cited “heightened volatility, unprecedented uncertainty and extremely fluid state of affairs” as the reasons for the missing forecasts. It is unclear if the fan charts—or their truncated versions, or some forecast scenarios—were shared with the MPC.
On the outlook for inflation, the latest MPC statement concludes: “…pull down headline inflation below target in Q3 and Q4 of 2020-21.” The Governor’s statement, an unnecessary addition to the list of documents economists sitting in the trenches, and investors, have to rush to digest quickly following the policy announcment, states: “…the MPC’s forward guidance on inflation is directional rather than in terms of levels. Going forward, as and when more data are available, it should be possible to estimate the path of inflation with greater certainty.”
How much lower than the official inflation target of 4%, is left to the subjective assessment of individual forecasters. That gap contains relevant information about the persistence of the evolving disinflationary shock that assists in the mapping of expectations about the MPC’s response function. But, RBI prefers not to stick its neck out; that responsibility has been outsourced largely to the private sector. And, never mind that even concluding that inflation in Q3 and Q4 of FY21 will be below target would have required a numerical forecast or assessment—unless they are winging it.
It gets better. The forecast for GDP growth is not only missing from the MPC’s policy statement, there isn’t a whiff about the extremely high likelihood of contraction in output in FY21. This crucial exclusion in the MPC statement is bizarre because the Governor’s statement reads: “Given all these uncertainties, GDP growth in 2020-21 is estimated to remain in negative territory…” The choice of the word “remain” is intriguing because the prior Governor’s or MPC statement didn’t indicate expectation that GDP would contract.
The above oddity raises questions about process, transparency, and communication: First, when did RBI conclude that GDP will indeed contract in FY21, and why wasn’t it announced clearly instead of last week’s “quietly-slip-it-in” approach. Second, why is the MPC statement silent on the contraction in GDP? Surely, Das isn’t working with a different forecast set from the one shared with the MPC panel he chairs.
It is unclear whether the omission about GDP contraction in the MPC statement was an oversight, a casualty of poor drafting, or a conscious choice by the powers to be. Whatever the reason or the motivation, the MPC should clear the air and also improve its communication. Indeed, the Governor’s statement, which is effectively meant to be a grand summary of the MPC and non-MPC announcements, needs to avoid carrying potentially market-sensitive guidance or comment(s) that should be included in the MPC statement in the first place.
Unfortunately, economic forecasts are least reliable during periods of high uncertainty, when they are, perhaps, most needed. Any seasoned and confident forecaster will confess that forecasting becomes less accurate when an economy experiences a destabilising shock. The behaviour of economic agents is altered depending on the nature and longevity of the shock, which also impacts the time taken for various response functions to return to normal. Consequently, in such settings, the forecasts for key macroeconomic variables, such as GDP growth and CPI inflation, used for making policy decisions understandably carry lower predictability.
In defence of forecasters, turning points are inherently difficult to forecast. But, that is where the commercial reality of forecasting partly comes in. Since forecasts typically converge over time, investors prefer to have the first mover advantage to redesign their asset allocations and/or restructure their portfolios. The volatility in forecast errors, and the direction and magnitude of forecast revisions carry useful signals for investors and market economists. And, given the strong herd behaviour in forecasts as a result of which they are normally concentrated within a narrow band, sensible outliers attract attention, often amplified by media coverage.
Frankly, even if one is forgiving of the missing fan charts because of the unprecedented volatility, multifaceted uncertainty, and issues about the reliability of data and surveys because of the lockdown, the egregious absence of numerical guidance shouldn’t be overlooked. RBI could have announced, say, a forecast range.
But giving up forecasts all together? That is embarrassing for RBI for two reasons: First, just about every professional economic forecaster—all of whom work with fewer resources than those at the disposal of RBI’s forecasting team—are publishing forecasts, several with some quarterly granularity as well.
Admittedly, most private forecasters don’t use the more time- and data-intensive econometric models, but the IMF and other central banks do, and they surely haven’t given up publishing forecasts despite facing similar lockdown-related challenges. The current range of economic forecasts by the private sector is wider than what it typically is, and will likely be susceptible to sizeable revisions. That shouldn’t be surprising given the unprecedented nature of the shock, the resulting heightened uncertainty, and the still-evolving policy response. But, and this is important, investors are used to digesting forecast revisions.
The Reserve Bank shying away from publishing forecasts, or even numerical guidance, for two key macro variables that form the foundation of the inflation-targeting framework is an abdication of a key responsibility. Indeed, RBI should explain what is so unique about it that it is incapable of publishing key economic forecasts when other forecasters in India and abroad are routinely able to do so.