Market.view

A riddle answered

Jun 15th 2008
From Economist.com

How inflation affects bond and equity yields

WHAT are the implications of rising bond yields for equity markets? Answering this question requires grappling with one of the most fundamental debates in modern finance.

One approach sees equities and bonds as assets that compete for a place in investors' portfolios. If one asset class becomes overpriced, then investors will flock to the other. The conventional way of comparing the two is to look at the yields.

But which equity yield? In more conservative times, investors looked at the dividend yield. Until the late 1950s, it was common for the dividend yield on the stockmarket to be higher than the government-bond yield; after all, since dividends could be cut, they were more risky.

But pension funds and other institutional investors noted that, in a broadly diversified portfolio, dividends would tend to rise pretty steadily over time. That growth meant that equities could trade at a lower dividend yield than bonds. The wisdom of this shift into equities seemed to be confirmed when the real value of bonds was devastated by the inflation of the 1960s and 1970s.

By the 1990s, investors had turned to the earnings yield, the inverse of the price-earnings ratio (so if the p/e is 20, the earnings yield is 5%). A paper by Federal Reserve economists led to a widespread belief in the Fed model—that the market was fairly priced when the prospective earnings yield on the stockmarket was equivalent to the ten-year Treasury-bond yield.

In this world, falling bond yields were seen as unequivocally good news for equities. And it seemed to work, thanks to the long period of disinflation from 1982 onwards, which saw nominal bond yields fall sharply and share prices rise substantially.

Believing in this model, however, required one to ignore some powerful contrary bits of evidence. After all, the model did not work prior to 1960. Companies only pay out a fraction of their earnings as dividends, so the earnings yield was then much higher than the government-bond yield. And the model did not work in 1990s Japan, where sharply falling government-bond yields were no help to the stockmarket.

Disillusionment came in 2000-02. Bond yields fell sharply, but this was not good news for shares. Indeed, investors were switching out of shares and into fixed income in the wake of the dotcom bubble. Equities looked cheap on the basis of the Fed model and got cheaper. A lot of people argued that the model was rubbish; after all, if lower bond yields were the result of lower inflation, then why should equities benefit? Profits forecasts would have to fall as well.

The answer to the riddle seems to show up in some figures (see chart) highlighted by Richard Cookson, a strategist at HSBC (and a former writer for The Economist). Bonds may move in sync with equities at some parts of the economic cycle but not at others. When recession or deflation looms, government bonds are relatively unaffected but equities suffer because profits are likely to fall.

Since 1900, when inflation has been in the 1-4% range, price-earnings ratios on the stockmarket have averaged between 17 and 19; in other words, an earnings yield of 5-6%. But when inflation has been below 1%, p/e ratios have averaged just 14 (an earnings yield of 7%). In such circumstances, bond yields fall but earnings yields rise.

In the current situation, bond yields are rising because of inflation: the headline rate is above 4% in Britain and America. And that is a problem for equities, since 4% seems to be a key figure. Average p/es when inflation has been in the 4-5% range have been 15; by the time inflation reaches 6-7%, the p/e drops to 11.

The problem for stockmarket investors is that the economy currently seems posed between extremes. If inflation returns, that will be bad for valuations; if recession wins out, that will be bad for profits. It is an unpalatable choice.