Forecasts suggest a turning point in the global economy
December 9, 2012 4:50 pm
by Gavyn Davies
This is the time of year when the major teams of macro-economic forecasters in the financial markets produce their annual outlooks for the next 12 months. Today, I would like to discuss what these forecasts are telling us about a key question facing policy makers and investors, which is whether the 2011-12 downturn in the global economy has now touched bottom.
Although long and painful experience suggests that these year-ahead economic projections will need to be revised considerably in the course of the coming year, they have been shown to contain information which is better than can be derived by naive rules (like statistical extrapolations, for example). Furthermore, economic forecasts are very widely used to determine economic policy. Finally, investors need to know what is “priced in” to the economic consensus so that they can gauge the likelihood of future surprises which will have an impact on asset prices.
The first collection of graphs below shows the mean projections for GDP growth derived from a selection of the best known forecasting groups in the financial markets . There have been persistent downward revisions to growth projections for both 2012 and 2013 over the past 18 months. These have reduced the predicted level of GDP in the 2013 calendar year by about 2 per cent since the autumn of 2011, with the revisions applying almost equally in 2012 and 2013 respectively.
Most countries have experienced downward revisions, though the US is a major exception. Unsurprisingly, the largest downward revisions have come in the peripheral eurozone economies and the UK, though there has also been a significant downward shift China and the other emerging economies.
Click to enlarge
The behaviour of the financial markets has, to some extent, followed the pattern of these revisions, with US equities performing much better than equities in the eurozone, UK and China. However, the fact that global equities as a whole have produced positive returns of 14.2 per cent this year, compared to 5.7 per cent from global bonds, is something of a surprise, given the extent of the disappointments about GDP growth.
It is clear, in retrospect, that the reduction in risk aversion, which has followed the actions of the major central banks, has been the dominant force in the markets. Although the expected path for GDP has come down, the probability of an outright recession has diminished, and the discount rate to be applied to future profits has dropped with the real bond yield. Consequently, equities have out-performed bonds in a declining growth environment.
That pattern seems unlikely to continue. To judge from options prices in the equity market (the level of the VIX, for example), risk aversion has now returned roughly to its long term average level, and seems very unlikely to fall much further. In addition, the real bond yield is unlikely to drop much from present levels, even if the central banks undertake a lot more quantitative easing. This means that further gains in equity prices will need to be earned the hard way, via an improvement in GDP growth and corporate earnings. This could also unlock valuation gains, via a decline in earnings and dividend yields, which are very high relative to the artificially depressed level of bond yields.
The first set of graphs shows that the global GDP growth rate projected for the 2013 calendar year is now very close to the out-turn for calendar 2012. However, this implies that the quarterly rate of growth will start to improve from now on. In order to achieve the calendar year averages shown in the graphs, the annualised growth rate in global GDP would need to bottom at about 2.4 per cent in the current quarter, and then gradually rise to 3.5 per cent by 2013 Q4. In other words, the global economy would now be at a critical turning point.
This recovering growth pattern is also shown in the IMF projections which appear in the panel below. There have been successive downward revisions to growth in 2012 and 2013 in recent forecasting rounds, but most economies are expected to return to their trend growth rates, or even above trend, by 2014. Medium term growth rates have not been revised by the IMF, except in the case of China and other emerging economies, where five year ahead growth expectations have dropped by about 1.5 percentage points since the financial crisis.
Click to enlarge
What do these forecasts imply about economic behaviour? Crucially, they imply that the dip in growth in 2011-12 has been due to a series of temporary shocks which will not have permanent effects on global growth rates (though there clearly is a permanent hit to the level of GDP, relative to prior trends). The most obvious of these shocks are the eurozone financial crisis, and the fear of sharp tightening in fiscal policy in the US, UK and Japan.
The projections show (or, more accurately, assume) that the euro crisis will gradually abate, which seems reasonable in view of the much greater involvement of the ECB balance sheet in addressing the crisis this year, and the progress which has been made towards banking and fiscal unions. They also assume that the global fiscal tightening will proceed at a rate of about 1 per cent of GDP per annum once the US fiscal cliff has been resolved, and that this pace of fiscal tightening will eventually be offset by aggressively easy monetary policy.
Upside surprises would of course occur if demand in the global economy were to recover more rapidly than expected, probably led by the end of private sector deleveraging in the US. Downside surprises would stem from higher than expected fiscal multipliers, or the impact of dysfunctional banking sectors, probably led by the UK and southern Europe.
On the central economic projections, we might expect that a “par” return on global equities in 2013 could be slightly higher than the 7 per cent projected growth in nominal GDP. This excess would be triggered by the fact that equities are undervalued relative to bonds, as explained above. Unlike in 2012, deviations from that “par” outcome may well be triggered by changes in GDP forecasts in the course of the year. As we pass the turning point for global activity in the current quarter, investors need to watch changes in consensus forecasts even more carefully than usual.
There have been many studies of the track records of economic forecasters over the years. They tend to show that it is very difficult for any individual forecaster consistently to out-perform the median of professional forecasters, though there is a surprising tendency for a few forecasters to survive while persistently performing worse than the median. Studies also suggest that forecast errors are likely to be much larger when the economy is in recession, which unfortunately is when the forecasts are most needed. (See this study by the Cleveland Fed, and this one by the St Louis Fed, as examples of this type of research.) Studies also show that one year ahead forecasts for GDP growth tend to be biased on the high side.
If these conclusions are valid, then it follows that the median or mean forecast may be the best available information on the future course of the economy, although users should recognise that the standard error of such forecasts is fairly high, especially near economic turning points.
Many FT readers will probably want to dismiss economic forecasts altogether. Given the track record of formal economic forecasters from 2007-12, this is not too surprising. But if we want to have an intelligent debate about the decisions which need to be made in economic policy and investment strategy, is there an alternative to the use of forecasts? People who suggest alternative methods are often in fact suggesting other ways of making forecasts, without always realising it. (For example, the “value” approach to investing amounts to assuming that assets with attractive valuations will rise in price in the future. This is just a different way of making an implicit forecast of future asset price changes.)