World equity markets – bubble territory?

The recent run by equity markets has been unbelievable compared to previous recoveries.  From 1997 to 2000 the S&P500 charged ahead by 106%, only to collapse by 49% over the following 2 years.  In the 5 years to October 2007 it rose strongly (+101%) only to fall 57% to its 2009 trough.  So a run skywards of 193% to new all-time highs must send us into overvalued territory, right?

Based on a range of measures we’d say no, world market valuations are above fair value but not stretched.   We like four simple market valuation metrics which over time consistently and objectively help form a view of hype vs fair value.  These indicators are:

1. A simple long run trend analysis
2. Tobins Q (comparing the market cap of all listed companies with the total replacement cost of their assets)
3. Warren Buffet’s favourite valuation metric (market cap as a percentage of GDP)
4. Forward P/E  ratios compared to the 5 year average

One of the critical decisions with market valuations is not the measure you use but the starting point you measure from.  Some go back a long way – for example the highly regarded Shiller Cyclically Adjusted P/E  ratio (also known as the CAPE) uses a start date of 1880.  We question the relevance of going back that far – companies are very different beasts compared to 30, 50 or 100 years ago.  We prefer a more recent data set starting in June 1987 for 3 reasons: (1) this is roughly when markets became more globalised, heavily traded and liquid (2) this is after the last bout of out of control inflation and (3) this start date predates the first of the modern stock market crashes in October 1987.

So armed with our June 1987 start date what do the indicators tell us?  Firstly we are nowhere near the overvaluation levels of the late 1990s.  Although the market is higher than at any point in the last decade, we are far from bubble territory.   When past markets were at the same level as now the following two years returned an average of 6% p.a. (although with considerable variation around this average).