The chart below is an updated version of one presented by Goldman Sachs chief strategist David Kostin before the pandemic that I haven’t been able to get out of my mind. U.S. sector valuations seem to have organized themselves using a largely outmoded accounting measure in the form of book value.
First, the chart. The x-axis is price-to-book value (P/BV), and the y-axis is profitability using return on common equity (ROE). Each industry falls closely along a trend line (the variance math backs this up) that, depending on your beliefs about correlation and causality, might imply an association between the two accounting measures.
It is an issue here that few analysts or investors use book value any more. The measure was ideally suited to Depression-era industrials for capturing the liquidation value of a company that almost certainly owned machinery that could be auctioned off.
Now, the assets for the companies that dominate the index are intangible – things such as patents, installed software bases, brands, client relationships and sheer scale. The value of intangible assets in the event of liquidation is much more difficult to assess, so book value is largely ignored.
Tangible asset-heavy market sectors are all grouped in the chart’s bottom left – low P/BV and low ROE territory. This includes utilities, real estate, materials (the U.S. materials sector, unlike Canada’s, is not dominated by precious metals exposure) and possibly energy.
The high P/BV, high profitability upper-right territory on the chart features technology, unsurprisingly, and also Amazon.com-dominated consumer discretionary. The communications services sector, in the lower third of the trend line, is an interesting case. It includes the low-asset, high-profitability Alphabet Inc. and Meta Platforms Inc. but profitability is dragged down by asset-heavy AT&T Inc. and Verizon Communications Inc., along with various and sundry legacy media.
This is more a chart to ponder rather than use as the sole source for asset allocation decisions; it’s too crude, and there are the accounting issues mentioned above. But it appears that despite the lack of interest in book value multiples (bank stocks might be an outlier here), the market sectors in the world’s largest equity market have organized themselves according to it, driving higher P/BV multiples in proportion to ROE.
There may be a reason behind the apparent relationship between P/BV and ROE that I haven’t considered, making the chart eye-rollingly predictable. Otherwise, it may be the case that investors in aggregate have unwittingly created a new equilibrium that could help approximate fair value for equities.