The Destructive Power of Dilution

By VCC Staff

  • 19 Jul 2010

Most investors and analysts assume that a large listed company’s earnings can grow at least as fast as GDP. This rarely happens in the real world largely because of the destructive power of dilution. In particular, in India real EPS growth for the BSE 100 is lower than GDP growth over a three and five year time horizon. In this column I will explore this counterintuitive phenomenon which has particular relevance for a market like ours where new share issuance is ubiquitous. 

EPS underperforms GDP.

Analysts frequently forecast and defend (on various pretexts) EPS growth rates in excess of GDP. However, the iconoclastic strategist, James Montier (now part of the asset allocation team at GMO, a global asset management firm) highlights in his book on “Behavioural Investing” that between 1970-2001 in the US, the UK and Germany, real GDP grew much faster than real EPS (the word “real” here denotes the fact that the impact of inflation is being stripped out both from EPS and from GDP). For example, in the US real GDP grew at 3.1 per annum between 1970-2001 whereas real EPS grew at a mere 1% per annum.


In fact, more generally in the developed world, EPS growth of broad portfolios of companies have been shown to be systematically below GDP expansion over long time horizons. The key words in the above assertion are “broad portfolios” and “long time horizons”:

         Narrowly defined portfolios or indices such as the S&P 100 or the FTSE 100 may not exhibit this behaviour. These indices are actively managed and suffer from survivorship bias (losers are eliminated and winners added).

         It is not difficult to find periods when EPS outperforms GDP but EPS tends to lag the GDP during periods of economic downturns. During periods of good economic growth, EPS plays catch up but does not always recoup the losses of the downturn. As a result in the long term the gap between EPS and GDP tends to grow.


Dilution is the culprit

EPS underperformance is due to net addition of shares over a business cycle. Fresh capitalisation of businesses, akin to what we have seen in FY10, dilutes equity holders. Whilst buybacks in good times partially offset the past dilutions, the shares bought back by companies are often used to feed generous option plans & warrants. EPS is not enhanced as a result of these reissuances. Value is merely transferred from the existing shareholders to the beneficiaries of the options & warrants. As a result the gap between EPS and GDP tends to grow over a number of economic cycles.

The Indian context


In the absence of comparable long term data for a broad set of Indian companies, my colleague, Shantnu Phutela and I based our analysis on Indian companies comprising the BSE 100. A comparison of the EPS growth for the BSE 100 with GDP growth broadly supports the thesis that in India too real EPS growth is lower than real GDP growth for BSE 100 companies. We have tried to partially eliminate survivor bias by basing our analysis on the historical index components i.e. for the three, five and seven year analysis, index components at the beginning of the respective periods have been included. The key findings from our analysis are summarised below:

         Over the three year period, FY07-09, the BSE 100 real EPS CAGR at 4.7%, was much lower than the 8.5% real GDP growth. This is to be expected, as the period under consideration includes an economic slowdown.

         Over the five year horizon, FY05-09, real EPS growth at 8.4% was marginally below the 8.5% real GDP CAGR.


         Over the seven year period, FY03-09, real EPS outperformed real GDP growth. However, the data for the seven year period is distorted by the fact that it includes a long economic boom.

         PAT growth in each time horizon has been significantly better than EPS growth implying ongoing net creation of shares resulting in dilution. Our analysis clearly shows that the best EPS performers in the BSE 100 (eg. Tata Steel, Titan, M&M) dilute less than average whereas the worst performers (eg. Pentamedia, MTNL, Mirc) dilute more than average.

Hence, given that long term EPS growth is a significant predictor of share prices, investors would do well not just to focus on the total profit generation but also on a promoter’s propensity to dilute. 



Share article on