Looking at the Nifty P/B or The Curious Case of The Falling ROE

So another widely tracked metric on market valuations is P/B, or price to book value. I decided to also chart out this metric and see what it says about market valuation. This is what I see:

Once huge caveat here - from what I see, the Nifty book value per share (BVPS) get updated only approximately once a year. This seems to be because in quarterly results companies only declare their P&L and EPS, and not balance sheet numbers. Thus, only when audited financials are reported does the BVPS number seem to get updated. This means there is a huge lag in this data as opposed to P/E (based on EPS), which gets updated every quarter. 

The median P/B for this period is 3.45.

The max was 6.55, just before the 2008 financial crisis.

The min was 1.97 in Sep 2001, when the market bottomed out after the dot-com crash.

The 25th percentile is at 2.96 and the 75th percentile is at 3.92.

This puts the P/B of 3.37 as of Jun 30, 2016 just below the median P/B for the period. 

Hmm, this is interesting. While the current Nifty P/E (as I had discussed in my previous post) is above the 75th percentile (actually at the 90th percentile) based on historical data, the current P/B is below the median - implying that the Nifty is not expensive by historical standards going by this metric.

Mathematically, if P/E is higher than average and P/B is lower than average, this should mean that the ROE is lower than average (since ROE is E/B = (P/B)/(P/E)). Indeed, if we look at the implied ROE for the Nifty over the period, we see that it has drastically reduced since the 2008/09 crash. See chart below:

Again, because the P/B lags, this data is not a completely accurate representation of the ROE, but is still directionally correct over the period we're looking at.

We see that the ROE for Nifty 50 companies over the last year is actually at the lowest levels since 1999. ROE should also correlate with EPS growth, and we do see that implied annualized EPS growth for Nifty for the period from Jul '09 - Jun '16 is only ~8% whereas the number for the period Jul '02 - Jun '09 is actually much higher at ~19%. See the annual chart below (Jul-Jun periods) for EPS growth for the Nifty.

I'm pretty intrigued by the dramatic fall in ROE since 2009 and wondering what the reasons could be:

• One obvious reason could be lower returns for companies over the last few years as the global economy hasn't fully recovered.
• The lower returns issue might also be compounded by higher debt on companies' balance sheets, further depressing returns, though I currently have no data to support this.
• Another reason however could be change in Nifty composition. For e.g., banks tend to have lower ROEs than consumer or tech companies and if Nifty composition has changed more towards these lower ROE industries then we may have to adjust our analysis for that. However, I do not have access to historical Nifty composition to be able to check if this is the case or not.

So, to summarize, while the Nifty P/E currently is well above the median P/E since 1999 (implying an expensive market), the current P/B is just below the median (implying a not so expensive market). However, ultimately I think investors care about earnings (and its growth) since that is the basis for equity valuation of stocks on a going concern basis. If we normalize the P/B for the lower ROE and earnings growth that we have seen over the last few years, then the market doesn't look cheap by that metric too. This analysis may need to be revised if Nifty composition has fundamentally changed since 2009.

It is totally possible though that investors are expecting a significant pickup in earnings going forward and are thus willing to buy into an expensive market by the P/E metric, assuming that the lower than median P/B will protect their downside. 

However, does this analysis change my view on current Nifty valuations? No.

What do you think?