Chart 1

Chart 2

Chart 3

Chart 4

Key Points

• As of 21 July 2009, the 12-month target for the FTSE STI calculated from bottom-up consensus analyst estimates was merely 2,456 points
• Consensus estimates tend to be momentum-driven and will be extremely inaccurate at economic inflection points
• Current estimates are understandably conservative, but too conservative in our opinion
• Earnings could decline by 10.1% in 2009, then grow 28.9% in 2010 and 13.7% in 2011
• We arrive at a 3,670 point target for the STI on a 16X price-earnings multiple based on forecast 2011 earnings

The stock market has always attributed considerable importance to consensus earnings estimates. Stocks react positively when corporate earnings beat expectations, but slump on failure to meet what stock analysts declare as their best estimate of a company’s earnings for any particular quarter.

Understandably, such high importance is placed on the estimates by stock analysts because they possess superior knowledge and understanding of the companies (through meetings with the management, rigorous study of financial statements, job expertise etc.). While stock analysts may be extremely accurate in their earnings forecasts for the subsequent quarter, it is a sobering fact that longer-term forecasting by the consensus is nothing short of atrocious at economic inflection points. The recent market correction which started in late 2007 provides us an opportunity to examine the behaviour and reaction of the consensus.

Consensus estimates revised down as the market tanked

Chart 1 shows how the consensus 2009 and 2010 estimated earnings (for the current index components) have changed along with fluctuations in the FTSE STI (Straits Times Index). The STI peaked in October 2007, and earnings estimates for 2009 peaked only two months later, in December 2007. As the market collapsed, earnings were continually being revised downwards, with 2009 estimated earnings 37.8% lower by the end of April 2009. This was, of course, a full month after the STI bottomed in early March 2009.    

Chart 1 clearly shows that earnings estimates and the stock market are closely related, with the former tending to lag the latter. What this suggests is that the majority of analysts are adjusting their estimates according to the performance of the stock market, which in our opinion, is counterintuitive. We believe that earnings estimates are meant to forecast the market, rather than the market being used as a tool to forecast earnings.

Consensus analyst target prices track the market’s momentum

Another example of the cardinal sin of closely tracking market momentum in relation to making estimates is reflected in how the target prices of stocks tend to track the actual stock prices, albeit with a small lag.

Chart 2 shows how closely the consensus 12-month target price for Singapore Exchange Limited (SGX) tracks closely the actual stock price at the time of forecast. The stock price peaked in October 2007 while the consensus target price only peaked in December 2007, evidence of consensus’ forecasts lagging the actual stock price.

A look at Chart 3 tells a similar story. By compiling target prices of STI component stocks on a monthly basis, we derived a target index level which we plotted against the actual STI level. Once again, the target index level tracked the actual index on the way down, which meant that most analysts were continually revising down target prices as the market tanked. Target prices are 12-month forward targets, you say? We brought forward target index levels by 12 months and plotted them against actual index levels (see grey plot on Chart 3), and the results clearly suggest that the consensus knew little about where stock prices would be 12 months down the road.

Not a matter of incompetency, but rather, differing time frames and differing ideals

We are not out to condemn equity research analysts; in fact, we appreciate the work they do, which is the basis for forward market valuations. Rather, we are trying to highlight the difference between the time horizons which most equity analysts base their work on, and the time horizon which we feel that investors of equity and equity funds should be looking at (at least two to three years).
This lack of compatibility between investment time horizons is an important issue because this leads to consensus analyst forecasts being extremely momentum-driven. Most equity research is done with a view towards a 12-month holding period, and some tend to focus on even shorter periods (eg. 3 or 6 months). What this means is that huge importance is placed on earnings expected over the next quarter or two, as that should determine the direction of stock price in that short period.

Moreover, many equity analysts are compelled to downgrade companies and target prices as markets tank, while continually revising up earnings and target prices as the market turns bullish. This stems from the fact that being right over the next week or two is much more important than being right over the next year or two. By revising target prices downwards when markets go into free-fall, the analyst thus has a higher chance of being correct over the near-term.

With an inflection point in the economy reached, the consensus is slow to react

While we have criticised the work of the consensus so far, we still believe that most equity analysts will be largely right when the economy is on a stable uptrend or downtrend. When the economy takes a sudden turn (an inflection point) and earnings become notoriously difficult to predict, the consensus analysts’ forecasts tend to be less accurate. It is apt to highlight here that there are always some equity analysts who see things differently, and have been proven correct with contrarian calls on equity markets while the consensus has been looking in the same direction.

In our period under study (2007-2009), we have seen an inflection point in late 2007 where analysts were slow in revising down earnings estimates and target prices. We believe that a second inflection point has occurred in early 2009, which is indicative of a recovery in the global economy. We have written that the Singapore economy is officially out of a recession (Quick Update: Singapore Officially Out of Recession!), but we would like to highlight that given the cyclical nature of the bio-medical sector which boosted 2Q 09 figures, 3Q 09 may show a contraction from 2Q 09’s high base. Nevertheless, we believe that an economic recovery remains on track.

This should entail much higher levels of earnings over the next two years, but coming after a severe recession, the consensus is understandably conservative. We thus feel that the consensus is underestimating the recovery of earnings and will likely revise up estimates over the next few quarters as actual earnings recover.

Banks earnings set to surprise, property outlook not as bleak

Current estimates understate the potential recovery of earnings of the three local banks, which have estimates based on distressed economic conditions. Non-performing loans remain a far cry from levels in 1997-1998, and with credit losses expected to lessen considerably, earnings have much potential to surprise on the upside.

Improving earnings of the banks are expected to go hand-in-hand with a recovery in the property sector, which is currently forecasted to have a gloomy outlook. Excessive liquidity appears to be absorbing the much-touted “excessive supply” in the property market, which should boost the earnings of property developers.  

As at 21 July 2009, the consensus expects earnings of STI companies to decline 19.4% in 2009, then grow 13.3% in 2010 and a further 13.9% in 2011 (shown by the red bars in Chart 4). We believe that the potential recovery in earnings is currently underestimated by the consensus and have revised up 2009, 2010 and 2011 earnings estimates.

For 2009, we are expecting earnings to decline 10.1%, before spiking 28.9% in 2010 and growing a further 13.7% in 2011 (see Table 1).

Revising up earnings estimates to give us a target of 3,670 points for the STI

Various indicators have shown that things are improving on the economic front, in-line with our view that the bottom of the economic cycle is behind us. The consensus is still besotted with the recessionary conditions which have plagued the global economy, and are unsurprisingly conservative with forward estimates. As we pass this inflection point in the economic cycle, we believe that the consensus is slow once again in their forecasts, and will likely revise earnings and target prices upwards a few quarters down the road as corporate earnings and economic data surprise on the upside.

As at 21 July 2009, the compiled Bloomberg consensus target prices for STI companies yielded a paltry 2,456 point 12-month target for the Singapore benchmark index. Like estimates before, this will prove to be flawed and we can expect upward revisions over the next few months. Based on our forecast 2011 earnings and an undemanding 16X price-earnings multiple, we arrive at a 3,670 point target for the STI by 2011. We maintain our 4.5 star “Very Attractive” rating on the Singapore equity market.

Related Articles:

Economic Recovery Funds Promotion
Why the Singapore Market is taking a breather
Can The Singapore 2009 Budget Save The Economy?
Expect A Weaker Singapore Dollar In 2009
Singapore Budget 2009 – A Preliminary Examination
Singapore in a Technical Recession: About the Technicalities and the Possibilities

Singapore Equity Funds (at 1.5% Sales Charge Until 17 AUGust 2009):

SGAM Singapore Dividend Growth
Aberdeen Singapore Equity