It’s difficult at the top
Markets are grinding higher and many investors are looking to call the top of the market cycle. In this light we ask ourselves what causes bubbles and more importantly how do we spot them? Over time I have collected a range of indicators that seek to have some predictive value in spotting bubbles emerging. I modestly call this the Heiligenberg Index.
Markets are without a 5% correction for the last 198 days and counting. Periods of this length are not unique but they are rare. Little corrections are bought eagerly due to investors' fear of missing out. Contrary to what one may expect though, investor sentiment is not overly optimistic. For instance the American association of Individual Investors Bull Bear indicator is just at average levels (not signalling any extreme investor sentiment). Standalone equity market valuations look rich in our view, especially in the US, while relative valuations (for instance the risk premium of equities relative to bonds) appear ok.
The FT ran an article on the perils of calling the peak of the equities bull run. In my career, over the past 21 years (yes it has been that long), I have analysed market cycles and bubbles: partially as a fascination of the repeating phenomenon, partially because it is a crucial question to make money in asset allocation. What causes bubbles and more importantly how do we spot them? If you can answer this question with some accuracy, the rest of your job as a macro investor is relatively straight forward. My fascination for Bubbles is not unique. One must-read book on this topic and an important inspiration for my work in this field is Edward Chancellor’s: “Devil Take the Hindmost”.
Over time I have collected a range of indicators that seek to have some predictive value in spotting bubbles emerging. Most of them are time series data, like the Senior Loan Officers survey but some of them are very subjective. For instance one of the subjective things I track for years is whether “investment is becoming mainstream party conversation” (my score is currently “not yet” at least not at the parties I go to), another one is whether we see a lot of market talk around “this time is different” (my score is currently “moderate” as we see some signs in technology and around profit margins but nothing like the 1990 internet frenzy) or whether we see market frenzies around IPOs (my current take here is a "moderate" score as we have seen a few crazy stories around twitter and Facebook but nothing recently and nothing comparable with the frenzy we saw during the internet era). As you can imagine these more subjective factors make the quants in my team smirk and mumble something about lack of proper t-tests, useless R-squares and a hindsight bias.
If we ignore the naysayers and we bring all indicators together, we can construct a simple equally weighted leading indicator of bubbles; I modestly call this the Heiligenberg Index:
As you can observe, the Heiligenberg Index would have been a good indicator for the TMT bubble in 1990s and the sub-prime bubble in 2007. The signal in June 2014 was less extreme and with hindsight the index gave a false signal. At the moment my read on the index is that it shows an elevated but not yet completely alarming risk of a bubble. To get an edge, I think we should keep our eyes peeled on the credit markets. Things to look out for are for instance: increased M&A that create further leverage, further decrease of lending standards and rising credit spreads.
In our view the low bond yields – we expect them to remain lower for longer (see this blog and this blog) – and the economic cycle (see Tim’s latest blog) are crucial in explaining the current market grind higher. As mentioned before: relative valuations (equities versus bonds) are not extreme in our view and this matters in a world where investors have minimum return requirements with relatively short horizons.
So as long as interest rates remain low, the cycle doesn’t turn and the Heiligenberg Index doesn’t raise the alarm bells, we believe it’s too soon to call the top.