Which value metric works best?

I recently undertook a quick and dirty backtest on NIFTY stocks to verify whether buying a portfolio of the cheapest index stocks and rebalancing annually would beat the broader index. In spite of the fairly short period under test (nine years from 2008 to 2017), if you’re a value investor, the results are encouraging.

From Sep 2008 to Aug 2017, ignoring transaction costs and dividends, a 10 stock value portfolio returned 18% annually compared to 9.8% for the NIFTY.

The details of the test and outcomes are in the post ‘Can buying cheap NIFTY stocks beat the index

For this test, I defined “cheap” using Greenblatt’s ‘Magic Formula’ definition of price, i.e. EBIT / Enterprise Value.

But what if we define “cheap” using other metrics like Sales, Earnings, Book Value, Cash Flow, in relation to Price? How do value portfolios constructed from these metrics compare against the NIFTY?

In this post, I compare results from the EBIT / Enterprise Value measure against other price-related metrics like Price-Earnings, Price-Book Value, Price-Cash Flow. The results are interesting.

Value portfolios comfortably outperform the NIFTY

The Calm Investor | NIFTY Value Portfolio Comparison

Irrespective of the metric we use to identify “cheap” stocks (EBIT / TEV, Price-Earnings, Price-Book, Price-Cash Flow from Operations), all value portfolios seem to comfortably outperform the 9.8% annual return of the NIFTY. Looks like buying cheap works, irrespective of how we define “cheap”

For a concentrated 5-stock strategy, P/E, P/B, P/CFO portfolios knock it out of the park

The Calm Investor | 5 stock value portfolio growthBad news up front for the EBIT / EV metric. It’s respectable 18% return performance is trounced by well-known valuation metrics, P/E and P/B. Though it’s the Price-Cash Flow from Operations value portfolio that ends with the highest returns.

A ₹100 investment in Sep 2008, in any of the P/CFO, P/B, P/E value portfolios would be worth more than 6x by Aug 2017 compared to a modest 4.2x for the EBIT / TEV metric, and far ahead of the 2.3x in the NIFTY.

Before drawing any conclusions though, let’s look at overall metrics.The Calm Investor | value portfolio performance comparison All value portfolios show better Sharpe Ratios than the NIFTY, but the P/B and P/CFO showed far higher volatility than their peers. Seeing one-year drops in absolute value higher than your starting portfolio amount is not for the faint-hearted.

All things considered, the Price-Earnings value portfolio does the best when considering a relatively concentrated portfolio of 5 NIFTY stocks.

Interestingly, the 3 top performing portfolios (P/E, P/B, P/CFO) are identical for the first two years before diverging in subsequent years. (you can see the identical movement on the chart above)

2008: BPCL, ONGC, GRASIM, TATA MOTORS, TATA POWER

2009: SUN PHARMA, ONGC, TATA POWER, BPCL, GRASIM

Diversified value favours EBIT / EV over the other metrics

Consider results from 10 and 15 stock portfolios respectively

The Calm Investor | 10 stock value portfolio comparison

The Calm Investor | 10 stock value portfolio

The Calm Investor | 15 stock value portfolio growthThe Calm Investor | 15 Stock Value Portfolio ComparisonNotice how the returns from the EBIT / EV portfolio don’t drop dramatically when going from 5 to 10 to 15 stock portfolios; 4.3x to 4x while volatility is steady between 22 and 24%

The other portfolios see bigger swings in value: P/B (6.2x to 2.6x), P/E (6.5x to 3.0x) and P/CFO (6.7x to 3.2x). They see reduction in volatility with increase in number of portfolio stocks.

Conclusion: The best of the value metrics

Based on their relative performance on returns as well as volatility and drawdowns, the Price-Earnings and EBIT / EV value portfolios look superior in this lot. The Price-Book Value portfolio looks particularly unremarkable as it shows middling returns at high volatility. The caveats from the previous post also apply here, mainly that the time period considered is a short ten-year window and it would take a lot more historical data to consider these findings conclusive.

Here are stocks identified for the Sep 2017 portfolios.

The Calm Investor | 2017 NIFTY Value PortfolioDisclosure: I currently own stocks in red bordered cells. Do not consider any stocks mentioned here as recommendations to buy / sell. Please do your own research before making any investment decisions.

2 thoughts on “Which value metric works best?

  • September 29, 2017 at 10:21 pm
    Permalink

    Nice follow up post.
    please include bank stocks also in the study.

    Thanks for sharing.
    I came to know about this blog from forindianinvestorcom

    I thank them as well…

  • October 2, 2017 at 7:11 pm
    Permalink

    Thanks. Financials of banking stocks are not comparable to other sectors, hence the trouble with including them. Will see if there’s a way to work around that.

What do you think?