Inaccurate 2013 growth estimates unsettling
WHEN the Ministry of Trade and Industry (MTI) first introduced quarter-on-quarter advance gross domestic product (GDP) estimates in Q2 2000, it aimed to provide more timely economic data to the public and business community.
By all measures, these flash estimates have been a boon for investors and businesses alike, giving an early and largely accurate indication of sequential GDP growth each quarter - until last year.
In 2013, three out of four GDP estimates (Q1, Q3 and Q4) were off the mark in directional terms - the economy was projected to contract, when it actually expanded quarter-on-quarter.
Because flash estimates are compiled largely from data from the first two months of the quarter, some degree of variation is to be expected, once more comprehensive and detailed data becomes available.
But last year's growth projections were uncharacteristic, and gave economists pause. Some even referred to the changes as "massive revisions" - not because the differentials were huge, but because the estimates and actual figures bounced in different directions.
The biggest surprise came from Q4's data - MTI had said in January that it expects the economy to contract by 2.7 per cent. But, in February, this number was revised upwards by 8.8 percentage points, to a quarter-on-quarter annualised rate of 6.1 per cent growth.
And while revisions to Q1 and Q3 data weren't as large in magnitude - upward adjustments of 3.2 percentage points and 2.3 percentage points respectively - the change in direction has been unsettling.
This is especially so when viewed historically: Over the last 10 years, MTI's quarter-on-quarter flash GDP estimates haven't been off the mark in directional terms - not even once.
How is it, then, that the Government projected - incorrectly, as later data subsequently showed - sequential contractions for three out of four quarters last year?
First, the nature of Singapore's small and open economy could be a factor. With total trade about 2.65 times GDP, the Republic is especially vulnerable to financial shocks in other countries, and would likely feel the impact more than larger, more domestically driven economies.
Second, with Singapore's shift to more volatile growth sectors - such as the biomedical manufacturing cluster - accurate advance estimates could be increasingly difficult to hit, simply because of the country's growth profile.
Because the early projection is based on data from the first two months of the quarter, if, for instance, production in the biomedical segment were to surge in the final month, the flash estimate could very well be blown out of the water.
Third, some have theorised that Singapore's growing ties to China's economy could have an impact, especially when seasonal factors are in play. If manufacturing plants in China shut down, for example, Singapore's production cycle would be affected too - impacting economic growth here.
By April 14, MTI will release its flash estimates for Q1 GDP growth. The median forecasts of private-sector economists polled by Bloomberg stand at 5.4 per cent year-on-year, and an annualised 0.7 per cent quarter-on-quarter, after seasonal adjustments.
With so many unpredictable factors at play, there is no doubt that MTI's task is a difficult and tricky one. Some economists say that the Government's data collection processes ought to be reviewed, to see if more accurate estimates can be garnered from the get-go.
Still, regardless of the reasons behind them, the swings between projected figures and actual sequential growth numbers have led some to question whether flash estimates are even serving their purpose anymore.
If advance estimates were introduced to give investors, businesses and the public a heads-up on where the economy is heading - certainly a laudable and important endeavour - last year's reversals were misleading and confusing.
At least in theory, when negative headline advance estimates are released (by the Singapore Government, no less), a knock-on effect on consumer and business behaviour can occur, impacting investment and spending decisions - only for this to have been for nought when the actual figure turns out positive.
Looking at things through an optimist's lens, it's true that last year's news wasn't all bad: At least GDP growth kept surprising on the upside (and not the other way around), and the economy escaped landing in contractionary mode multiple times.
But it's also true that when government projections veer off in the opposite direction altogether, the quarter-on-quarter flash estimates start to lose their relevance.
And, with numbers so different from the final outcome, the credibility of advance GDP numbers will inevitably be called into question - defeating the purpose of having them in the first place.