In an extensive report published this morning by Deutsche Bank’s Jim Reid, the credit strategist looks at the “Next Financial Crisis”, and specifically what may cause it, when it may happen, and how the world could respond assuming it still has means to counteract the next economic and financial crash. While we will have much more to say on this study in upcoming posts, we wanted to bring readers’ attention to one observation made by Reid, namely that “we’re in a period of very elevated global asset prices – possibly the most elevated in aggregate through history.”
Here are the details on what appears to be the biggest asset bubble ever observed, courtesy of Deutsche Bank:
Figure 57 updates our analysis looking at an equal weighted index of 15 DM government bond and 15 DM equity markets back to 1800. For bonds we simply look at where nominal yields are relative to history and arrange the data in percentiles. So a 100% reading would mean a bond market was at its lowest yield ever and 0% the highest it had ever been. For equities valuations are more challenging to calculate, especially back as far as we want to go. In the 2015 study (‘Scaling the Peaks’) we set out our current methodology but in short we create a long-term proxy for P/E ratios by looking at P/Nominal GDP and then look at the results relative to the long-term trend and again order in percentiles. Nominal GDP data extends back much further through history than earnings data. When we have tracked the two series where the data overlaps we have found it to be an excellent proxy. Not all the data in Figure 57 starts at 1800 but we have substantial history for most of the countries (especially for bonds).
As can be seen, at an aggregate level, an equally weighted bond/equity portfolio has never been more expensive. Figure 58 shows that bonds are much closer to 100% than equities though and Figure 59 then looks at the raw data for bonds showing average G7 yields back to 1800.
It’s easier to be black and white in terms of bonds long-term value. In short there isn’t any relative to history. For equities it’s more difficult to assess partly because they are a real asset and therefore today could be a good time to buy if one felt that despite relatively high valuations, inflation may permanently increase (or better still real GDP growth) and thus lead to eventually permanently higher earnings notwithstanding any short-term negative implications of the inflationary transition. However our technique looks at valuations relative to what we know now and where we are relative to history.
For equities, current valuations are certainly stretched relative to nominal GDP through history. We have been more expensive but we are approaching the peaks of 2000 and 2007 and are in line with the most stretched valuations from the 1930s on this metric and higher than the 1929 crash point.
Given how weak nominal and real GDP has been post GFC (Figure 60), and how much of a downward trend both have been for several decades now, this shouldn’t be a surprise.
Nominal and real GDP growth rates have been trending down and unless equity returns slow relative to the past, then valuations on our measure will go up. Obviously if profits take up a bigger share of GDP for a period of time our method will look more stretched than traditional P/E ratios. However over the longer-term, this should be mean reverting as profits can’t permanently outstrip nominal growth – especially at a global level. Currently there is some evidence that the US is one area where actual earnings have outstripped nominal growth in recent years for various reasons that include their large global players gaining excess overseas earnings (must be a zero sum game globally), a more shareholder friendly and focused culture and perhaps higher inequality and therefore more spoils to capital over labour.
However we’d repeat that history suggests all this is mean reverting over the medium to long term. If we look at more detail on the US which has the most developed history of equity data, including the longest series of earnings data through history we can see the longer term issues with equity market valuations.
Indeed the US CAPE ratio (Figure 61) has only been higher before the 2000 equity bubble bursting and was only slightly higher ahead of 1929 crash. CAPE analysis cyclically adjusts earnings by using the average of the last 10 years so you would have to believe the higher earnings of the last decade represent a new paradigm to not be concerned by this graph.
DB’s conclusion: “While there are no obvious triggers for historically high global asset valuations to correct, while they remain this high there is always a risk of a sudden correction that could be destabilising to a financial system and global economy that seems to require such elevated asset prices.”