Equity valuations – party like it’s 1999… and 2007?
Amidst something approaching a euphoric relief rally in global markets following the UK’s vote to leave the EU, investors should not overlook equity valuation metrics, which have historically provided an excellent guide to returns over the long term. As Exhibit 1 shows, relatively low valuations preceded the bull markets in 1994-1999, 2002-2007 and 2009-2013. However, valuation metrics rarely form part of a market narrative and if they feature at all are often dismissed, usually as “it’s different this time”.
In the 1990s, globalisation and technology narratives were offered as a reason to ignore quite obvious signs of trouble ahead in equities. The reverse applied in 2002-2004 when the resulting stock market correction was mistaken for a long period of economic malaise.
Exhibit 1: Price/book valuations highly correlated with three equity cycles since 1990s
At present the current consensus narrative is that as interest rates and bond yields are very low, equity valuations “should” be high as investors are forced to go out on the risk curve to secure yield. This argument may appear logical on the surface but in our view is mistaken both empirically and theoretically. Empirically, we can see that in Japan until 2014 (when the Central Bank started actively targeting equities under its QE programme) lower government bond yields were associated with lower price/book valuations.
Exhibit 2: Japan – declining fixed income yields not always associated with rising valuations
Theoretically, the link between the bond yields and equities also appears moot. The market dividend yield should represent the sum of the equity risk premium and nominal risk-free bond yield, less nominal dividend growth. Declines in bond yields due to falling inflation expectations should be offset by declines in nominal dividend growth, leading to no change in the dividend yield on equities. Similarly, declines in real bond yields, which represent a fall in expected real economic growth, would logically lead to the same fall in expected dividend growth and no change in equity valuations.
In a period of financial repression, real returns on bonds may be artificially suppressed for a time by central banks wishing to stimulate the economy by providing a subsidy on the cost of capital. We would agree this would raise equity valuations but only modestly, and on the assumption this policy was guaranteed of success. For example, a 2% reduction in the risk-free cost of capital for five years would be expected to generate only a 10% uplift in equity valuations.
The only remaining parameter is the equity risk premium and for a time during the 1990s there was an active debate on whether a reduction in inflation uncertainty post-Volcker could be linked to a sustained reduction in the equity risk premium. That argument may have had some validity at least.
However, in this cycle it is difficult to argue that the combination of very slow economic growth and experimental monetary policy during the current cycle is a reason for a sustained fall in the equity risk premium. To the contrary, it could easily be argued that the growing body of evidence that the effect of ultra-loose monetary policy on the real economy is transient and modest should imply a relatively high equity risk premium, as prevailed in the 2009-2012 period.
Exhibit 3: Current equity overvaluation compared to historic norms
At present, valuation metrics in the US, UK and continental Europe are at the top of their historical ranges, indicating a substantial lowering of the long-term return on developed market equities. Exhibit 3 shows the extent of overvaluation on a number of measures. While the temptation is to focus on the headline numbers – and it is certainly more dramatic to use such data to forecast a crash – we have no reliable model to forecast how valuation metrics might mean-revert in future.
There is for example no reason why any mean-reversion would not be gradual rather than sudden. Therefore, in our view, the key point is that investors should remain focused on what this low level of expected return may mean for equity portfolios on a timescale of five years, similar to the likely timescale of global interest rate normalisation.
Specifically, taking the average current premium to historical valuation metrics of 20-30%, equities in each of the US, UK and Europe appear priced to tread water until the end of the decade as the capital uplift from slowly growing revenues and profits may be offset by the renormalisation of valuation parameters to long-term averages. Furthermore, if not by definition, there is no reason why global equities cannot trade below their long-term valuation averages at some point over the next five years.
Conclusions
1. Watch out for shifts in the market narrative which “justify” high equity valuations.
2. There is at best limited theoretical justification for lower bond yields implying lower equity dividend yields.
3. Market valuations are close to their peaks on a number of metrics in the US, UK and Europe ex UK. While equities are currently supported by improved earnings momentum in the short-term, over the longer-term the risks appear skewed to the downside.
