In his latest note, Gluskin-Sheff’s David Rosenberg explains what indicator he would like to see that would make him bullish.
His favourite indicator: Revisions to analyst earnings projections.
The 12-month forward S&P 500 earnings estimates, for example, peaked at $103.61 per share right when the market peaked in October 2007. It bottomed, believe it or not, in March 2009 (right at the market lows), at $60.08.
The consensus analyst earnings estimates on a 12-month forward basis peaked this year in April — again, right when the S&P 500 did — at $94.79. It has since declined for five months in a row and so far in September is down 1.3%, to $86.74.
Based on this indicator, we’ve still got a way to fall:
This means that the time to start turning optimistic is when these earnings estimates bottom out (ie, the downgrading process comes to an end). Now at the profit cycle peaks, the analysts are typically about 20% too bullish on their earnings estimates for the coming 12 months, which would then mean a trough somewhere around $75 EPS this time around — if past is prescient. So this is a level I would be looking for before sounding the all clear that a lot of the “double dip” news is priced in ($75 would be a 4% earnings haircut from where we are now on a 12-month trailing basis for operating earnings). The consensus, meanwhile, is still at +11% in terms of profit growth expectations for the coming year, which is unlikely to occur if nominal GDP growth is flat-ish, as I expect. But at the peak, the consensus was looking at +43% EPS growth for the next 12 months, so at least we are well on our way in this process of eradicating the excess optimism in consensus profit projections.
In any event, I always say that there is no alarm bell that rings at the lows or the highs but this seems like a metric we can fall back on at some point to turn more bullish. It worked like a charm in March 2009 (though admittedly, it would have got us into the market too early in 2002 … but by mid-2003 the patience would certainly have paid off).