Benjamin Graham was known for his love of using historic data to analyze companies and historic market trends. Graham like to use long-term data series’, stretching across several decades, to come up with the most reliable figures.
Both Deutsche Bank and Credit Suisse have taken a similar approach to data analysis and the two banks put together research documents every year that look at the long-term asset returns.
Deutsche Bank released its annual Long-Term Asset Return Study a few days ago. The bank used 200 years of data to show that at least three major asset classes, bonds, stock and real estate are all currently trading close to peak valuation levels relative to history.
“…in aggregate, across these three main asset classes, average valuations are close to the highest they’ve ever been relative to their long-term trend. The current reading of just under 80% is similar to that seen at the turn of the twentieth century and during the 1940s when financial markets were artificially repressed around war time…”
On aggregate, all three assets appear to be undervalued. However, when you strip bonds out of the data, it appears as if equities and housing still have some wiggle room.
“Interestingly housing is back to 50% on this measure having peaked at 75% in 2007. As we’ll see in the separate housing section, this masks a wide range of valuations amongst the individual countries. Bonds are quite clearly close to 100% on this measure reflecting the fact that yields are still close to multicentury all time lows. For equities the current figure is 77% (down from 83% earlier in 2015), still high relative to history but shy of the last two peaks of 90% in May 2007 and 87% at various points in 2000.”
However, there is somewhat of a sensationalist slant to Deutsche Bank’s claim that asset prices are now sitting at record highs. For example, the bank’s real estate data only goes back as far as the 1970s. Moreover, Deutsche Bank’s researchers had to estimate equity valuations and bond yields based on assumptions. Specifically, equity valuations were computed at the index level using P/E ratios and where these weren’t available, using Nominal GDP as a deflator of long-term total returns was used. This approach is based on the assumption that long run earnings should broadly be tied to nominal GDP.
But the most obvious fault in Deutsche Bank’s data is its lack of consideration of how markets have changed over the years. A simple example would be the change in industry weightings that has taken place within developed markets since the beginning of the last century.
Credit Suisse’s 2015 Global Investment Returns Yearbook shows how the composition of the UK and US markets have changed since 1900.
[A more detailed summary of the Global Returns Yearbook can be found here]
Low margin, cyclical rail companies, which used to make up the bulk of indexes have been replaced by high-margin, defensive health companies, and in the case of the US, tech companies have grabbed a significant share of the market.
Market theory dictates that investors are willing to pay a higher multiple for cash generative, wide-moat companies with a global presence like Facebook, Google, Coca-Cola and Johnson & Johnson to name a few.
This is by no means a definitive analysis it’s just food for thought. To see a more detailed counterargument see this piece by Josh Brown of The Reformed Broker.