Two weeks ago, PIMCO's Bill Gross stirred up a few ivory-tower academics, [6] permabull sell-side commission-makers, and bloggers [7]pressured by Series [X] investors to generate maximum page views when he called for the death of the cult of equities [8]. His main point was the apparent paradox that the total return on equities can outpace GDP growth over long periods. While there has been much gnashing of teeth over this comment, Morgan Stanley has very succinctly clarified and confirmed that this is not so much a paradox as a Catch-22. The key point is that, in aggregate, investors do not typically reinvest their dividends (or coupons); and akin to Keynes' paradox of thrift, if investors actually tried this en masse then the historical returns reported in total return indices would be unachievable. So here’s the Catch-22: over the long run theoretical total returns can exceed GDP so long as investors don’t actually try to capture those returns. But if ever investors try to achieve such GDP-plus total returns, it will be impossible for returns to stay above GDP growth. Hence, equities for the short- and long-term, are essentially a Ponzi scheme - as long as everyone buys-and-holds - but if 'someone' decides (or is forced) to take-profits, equity ROE will rapidly game-theoretically collapse to GDP growth.
Why Gross Is Right...
Morgan Stanley: Mind The Gap
High-trend GDP growth still underpins some investors' focus on EM equities. However, there's no correlation - even over very long periods - between GDP growth and equity returns.
[9]
This is another example showing that economic growth and equity returns are unconnected. This is true over even very long periods. Exhibit 2 shows the correlation between equity returns and GDP (and GDP per capita) growth over periods of up to 105 years. Ironically, many of the correlations are negative. But the more important point to note is that the explanatory power is very low.
[10]
Several factors can drive a wedge between GDP and equity returns. Exhibit 3 shows a long history of equity returns in the US. The returns are calculated over a rolling ten-year period, adjusted for inflation, and including reinvested dividends. Returns are driven by two factors: dividends and changes in share prices. The change in share prices can be attributed to either change in earnings per share or changes in valuation.
[11]
It is clear that historically valuation changes account for much of the variation in returns.
It is because dilution rates are correlated with GDP growth that EPS growth is not correlated with GDP growth. Exhibit 5 shows real GDP growth and real dividend growth for 16 markets over 100 years to 2000. There is no correlation. In short, even if you are far-sighted enough to look through valuation changes, and swings in profit share of GDP, economic success still doesn’t guarantee investment success.
[12]
Finally, a comment on the apparent paradox that the total return on equities can outpace GDP growth over long periods. In the US, for example, nominal GDP has increased around 850-fold from 1900, while $1 invested in equities over the same period would have multiplied 25,000 times (Exhibit 6). (This huge gap is the result of nominal GDP compounding by 6¼% per year while the nominal total return on equities has averaged 9½% per year.)
[13]
Note that this paradox is not peculiar to the US, and is not peculiar to equities. Indeed, any debt asset that, on average, provides a yield higher than nominal GDP growth will generate a total return that persistently out-strips nominal GDP. For example, Moody’s BAA US corporate long bonds have averaged a 7% yield since 1920, higher than the 5¾% nominal GDP growth. So while nominal GDP has increased 170-fold since 1920, the total return on an index of BAA bonds has increased by almost 650 times. To recycle Bill Gross’s phrase, on this basis, debt markets are as much a Ponzi scheme as the equity market is.
Note also that these GDP-defying returns only appear when returns are calculated on a ‘total return’ basis – that is, on the basis that the dividends (in the case of equities) or the coupons (in the case of debt) are reinvested back in the asset class.
However, this isn’t so much a paradox as an investment Catch-22. The key point is that, in aggregate, investors do not typically reinvest their dividends (or coupons). If investors actually tried this en masse then the historical returns reported in total return indices would be unachievable.
For equities, the surfeit of capital coming from a compounding cascade of reinvested dividends would, sooner or later, crush margins, hence earnings, hence capital values (and hence future dividends). Reinvesting in debt indices would be likewise unsustainable. For example, to reinvest in US corporate bonds would mean rolling back 7% of the debt stock each year as new debt. Corporate debt in 1920 was around 70% of GDP; it would now be 464% of GDP if it grew in line with ‘reinvested’ corporate bond returns.
So here’s the Catch-22:
over the long run theoretical total returns can exceed GDP so long as investors don’t actually try to capture those returns. But if ever investors try to achieve such GDP-plus total returns, it will be impossible for returns to stay above GDP growth.
Source: Morgan Stanley