analysis of the australian stock market using the shiller index

425kB Size 13 Downloads 14 Views

Shiller warns that a market that is overvalued by historical standards, as it was in the ... calculating the price stock/earnings over 12 months, the Shiller PE ratio ...
ANALYSIS OF THE AUSTRALIAN STOCK MARKET USING THE SHILLER INDEX The term "irrational exuberance" was first used by Alan Greenspan, chairman of the Federal Reserve Board in Washington, in a speech entitled “The Challenge of Central Banking in a Democratic Society" in 1996. In this speech Greenspan posed a question: “... how do we know when irrational exuberance has unduly escalated asset values, which then become subject to unexpected and prolonged contractions as they have in Japan over the past decade?" Taking up this point in his seminal work Irrational Exuberance in 2001, Robert J. Shiller, a respected expert on market volatility and behavioural finance, provides a methodology to examine the phenomenon, and identify its impact on market valuations. In the five year period spanning 1994 to 1999, the market capitalisation of the Dow Jones Industrial average tripled. Traditional metrics of the broader economy could provide little explanation for this growth and at the time even respected market observers like Greenspan* who were originally sceptical of the boom, justified market valuations based on the explosion of the Internet and the productivity improvements that it promised. Conversely, Shiller diverges from the traditional economist assumptions of rational economic agents and efficient markets to identify a number of behavioural biases and structural changes that help to explain the growth and subsequent crash in market valuations. Shiller warns that a market that is overvalued by historical standards, as it was in the late 90’s, is inherently precarious. He directs some attention to the need for caution especially in pension plans, charged with managing retirement savings. At the heart of Shiller's analysis is the Shiller PE ratio. Whereas the traditional trailing PE ratio assesses the performance of an individual stock and "how expensive" a stock is by calculating the price stock/earnings over 12 months, the Shiller PE ratio examines price in relation to long term earnings averaged over a 10 year period, adjusted for inflation i.e. Shiller PE = price stock/ave. over 10 yrs (adj. inflation) Compared with the regular PE ratio, therefore, the Shiller PE thus smooths out fluctuations in earnings over a business and economic cycle. Consequently, it is less susceptible to overvaluing stocks at the peak of the business cycle and undervaluing stocks at the bottom of the cycle. While critics of the measure point to the potential for non-stationarity in mean earnings over a ten-year period, research generally shows that the Shiller PE is a good predictor of long-term stock market performance.1 A requirement of calculating a Shiller PE ratio is that both price and earnings data for a stock market is available for an extended period of time. This has generally constrained any estimate of the ratio in the Australian context prior to the early 90’s. 2 Using the ANZT-ACFS

1 2

See for example, Value Matters: Predictability of Stock Index Returns (Angelini et al, 2012) This is because earnings data for the Australian market is only available from the early 1980s.

Australian Equities Database which has earnings data beginning in 1964, this paper examines movements in the Shiller PE ratio for the period 1975 to 2011. First, we compare the standard PE ratio with the Shiller PE for all listed Australian companies over this period, and second, the Shiller PE for the Australian Market is compared with the Shiller PE for the US stock market.

DATA The ANZT-ACFS Australian Equities Database (AED) has been developed by the Australian Centre for Financial Studies (ACFS) in conjunction with ANZ Trustees to fill a gap in data available for long term sectoral analysis. It is a project that digitises monthly equity data on a stock by stock basis back to 1948. The aim is to fill the current gap in historical Australian equity data where stock by stock data has previously been unavailable prior to the 1980s. The AED contains stock market data for almost 6000 companies between 1948 and 2013 for the following variables: Capital data

Earnings and Dividend data

Paid up capital

Dividend pay date

Number of issued shares

Interim dividends

Nominal price per share

Final dividends

Net tangible assets per share

Dividend yield

Share volume traded

Book earnings (Beginning in 1964)

Data on the US Shiller PE ratio was obtained from Robert Shiller’s website.3

METHODOLOGY Developing the Shiller PE for the Australian market was a three stage process. First, while the AED has “monthly” EPS, these are based on the last reporting date, and as not all stocks report on the same date, the reported earnings of each company are divided by 12 to create a 12 month rolling average earnings series for each company. Second, the monthly earnings are then multiplied by the number of shares for each stock, and then summed to get total market earnings for each month. Multiplied price per share for each stock by number of shares for each stock, then summed these to get the market value of the entire market for each month. Third, the inflation adjustment is applied by downloading quarterly CPI figures from the ABS, estimating monthly CPI figures and then adjusting the total market earnings and market value using the CPI data. The real market value of the total market is then divided by the 10 year average of the rolling 12 month average inflation adjusted earnings. 3

http://www.econ.yale.edu/~shiller/data.htm

RESULTS When comparing the Shiller PE to the traditional trailing PE ratio it is interesting to note that the Shiller PE ratio has remained above the trailing PE ratio since 1985. This is reflective of the fact that the Australian stock market has not had a period of sustained negative inflation adjusted earnings growth for more than 20 years. However, the recent convergence between the two ratios highlights the negative earnings growth experience post-GFC. Figure 1 Standard Trailing P/E ratio compared to Shiller P/E ratio: Australia – 19752012 50 45 40 35 30 25 20 15 10 5 0

Standard Trailing P/E Ratio

Australian Shiller P/E Ratio

The comparison between the Australian Shiller P/E and the US Shiller P/E is even more interesting. The greatest divergences between the two time series occur just prior to the 1987 crash and before the Dot-com Bust. The results suggest that speculation in Australian stocks was much more pronounced in the lead up to 1987 while the US stocks were more overvalued at the turn of the century. This is consistent with tech stocks having a much greater weighting in the US. Divergence between the two ratios is occurring once again, a result of continued quantitative easing by the Federal Reserve perhaps? Figure 2 Australian Shiller P/E ratio compared to US Shiller P/E ratio: 1975-2012 50 40 30 20 10 0

Australian Shiller P/E Ratio

US Shiller P/E

CONCLUSION Market valuation measures like the Shiller PE can be useful a tool in determining the relative valuation of a stock market at any point in time. However, metrics of this nature provide no guidance to the user on what the true or intrinsic value of the market is and require a benchmark such as an average value or an international market against which to estimate relative over or undervaluation. In June 2012, the final month in which this analysis is performed, the Shiller PE ratio stood at 11.89, well below the 37 year average of 19.94. Furthermore, Figure 2 suggests that the Australian market had diverged from its long-term trend with the US market. Given these indications of relative undervaluation, perhaps we should not be surprised at the strong performance of the Australian stock market over the past 12 months.

Comments