As the ongoing stock rally gathers pace, experts are increasingly becoming hopeful that this is a harbinger of a multi-year bull run that should create wealth for investors, and themselves.
A good sign of any bull market is the increasing number of lofty forecasts that are made in a bid to attract apathetic investors to the great run. But very often, such predictions fall flat in the face of reality as the years go by, leading to much disillusionment.
In the history of forecasts, the most bizarre appeared in the US at the height of the dotcom bubble in 1999 when three books titled Dow 30,000, Dow 40,0000 and Dow 1,00,000 were published in quick succession — with the numbers outlining the targets the Dow was expected to reach — when the Dow was trading at around 11,000 (having already doubled in less than five years). Fifteen years on, the Dow is at around 17,000 levels.
Even in India, when the Sensex reached 21,000 in a crescendo-like move between 2005 and early 2008, there were several forecasts that pegged the next target for the benchmark at 30,000. The Sensex treaded water for the next six years, taking out the 21,000 comfortably only recently.
In the recent upmove, the Sensex has risen from about 18,500 levels to 27,000 currently in exactly a year (a near 50 percent move). Expectedly, analysts have upgraded both their earnings forecasts as well targets for the Sensex sharply over the past year or so.
Recently, Raamdeo Agrawal, joint managing director of Motilal Oswal, told CNBC-TV18 he expects the Nifty to reach 10,000 around the time of the Budget. Given that the Nifty and the Sensex benchmark values have historically had a ratio of about 3.3, this works out to a target for the latter at 33,000.
That’s not all. Several brokerages and advisory firms have recently put out various long-term targets for the Sensex.
HDFC Bank’s investment advisory division recently said it expects the Sensex to reach 40,000 by March 2017.
In an interview with the Business Line, BNP Paribas managing director CJ George said he expects the Sensex to reach 60,000 by March 2018.
While Karvy recently put out a much-discussed Sensex target of 1,00,000 by March 2020.
Most of these forecasts are based upon a specific framework of assumptions of earnings growth and valuations.
For instance, the earnings-per-share (EPS) for the Sensex stood at about Rs 1,300 in FY14. In March 2014, the Sensex was trading at about 22,500. This implied a trailing price-to-earnings multiple of about 17.2.
While forecasting long-term targets for a benchmark, analysts assume its earnings would grow at a certain percentage and it would trade at a specific multiple.
The earnings growth themselves are often predicated upon certain assumptions for GDP growth (so, many analysts expect India’s nominal GDP growth to climb about 15 percent for several years ahead: 7-8 percent real growth plus 6-7 percent inflation) and assume that earnings of bluechip companies could top GDP growth by some percentage points.
For instance, in the case of Karvy’s forecasts, it assumes earnings to grow at about 25 percent while forward earnings multiples (projected for the year ahead) would stay at 17 times.
If the Sensex’s earnings was to grow 25 percent each year from FY14, it would stand at about Rs 6200 in FY21. A forward multiple of 17 times would ensure the Sensex has risen to a little over 1,00,000 by March 2020.
But here’s the problem. Just as the stock market tends to not move up in a straight line, earnings of companies too do not.
For instance, below is the chart for EPS growth for the Sensex since 1990. Over the long term, EPS has grown at an average 15 percent but this performance is marked by several years of superlative and underwhelming growth, alternatively.
As seen in the chart, earnings growth for the Sensex has been very volatile, often bearing little correlation with actual GDP growth, which tends to be more linear. In each of the six-year periods between FY90-FY96, FY96-02, FY02-08 and FY08-14, growth has averaged 28.7 percent, -0.96 percent, 23.4 percent and 7.7 percent, respectively.
While 25 percent earnings growth is not unachievable and Indian companies have grown by that much in the past as can be seen above, that it may on average grow by exactly that much in the years ahead could be a risky bet.
Also, as forecasts go further far out into the future, it becomes greatly difficult to accurately take into account any of the number of variables that could end up affecting the forecast.
For instance, in this 2010 interview, an expert who had ‘seen the 2008 crisis’ predicted the Sensex could reach 50,000 levels by 2015 . This prediction was predicated upon a certain set of assumptions: “that the Indian economy would grow at an average of 8.5 percent per year until December 2015, if there is no double-dip recession in developed countries, if the Indian rupee continues to trade around Rs 46 to the US dollar”.
Four years later, none of the scenarios panned out as they were seen while the Sensex is almost half-way to the predicted mark.
The problem is compounded further by the fact that the Sensex, or any stock index in the world, gyrates well below and above its mean average PE multiple.
As seen in the chart below, even if we do not take into account the frothy valuations during the 1992 Harshad Mehta bubble (where one could consider an inefficient market was driven by an outright fraud), the stock market has gyrated between PE ratios of 11 and 29.
The long-term average PE ratio for the Indian market also stands at 15 times.
Investors generally consider any level between 12 and 18 as “fairly valued” and the market can very well gyrate between just about any PE ratio in the short term driven by factors such as liquidity and sentiment.
Thus, even if an analyst were to accurately forecast that EPS would grow by exactly 25 percent from FY14 to FY16 (from Rs 1300 to Rs 2030), it would be difficult if not impossible to predict at which multiple the market could trade at in March 2015.
Thus, the Sensex could be at 24,000 levels given a forward multiple of 12 and 36,000 given a multiple of 18. Valuations, though, would be considered as “reasonable” for both levels.