Value investing doesn’t necessarily mean cheap – in stocks or markets

By Derry Pickford*  Macro Analyst at Ashburton Investment

In some quarters the term 'value investing' has acquired misleading connotations. The concept of 'value' has become synonymous with 'cheap'. In the same way a cheap coffee might be undrinkable and be a bad buy, cheap does not confer value. We would argue that the distinction between growth and value investing has always been somewhat of a false dichotomy. A good value stock is one that will give great returns – it doesn't matter whether those returns are driven by an incredible growth opportunity or a recovery story. We trace the history of attempts to value markets, and survey the use of valuation ratios to forecast returns.

Valuing markets is a controversial subject in the financial markets literature. The Efficient Market Hypothesis (EMH), which has dominated finance theory for four decades, implies that such attempts are futile as value will always equal price. If a stock was expected to deliver excess returns, then arbitrageurs would soon buy it up, pushing the price higher to eliminate that excess return. For value to differ from price implies that there must be some barriers to arbitrage.

Whilst arbitrage works if investors can forecast short-term returns with high degrees of certainty, it becomes more difficult the longer it is expected to take for value to reassert itself and the more uncertainty we have about returns. One of the pioneers of EMH theory, Nobel Laureate Paul Samuelson, argued that stock markets are micro efficient but macro inefficient: that the value of a market as a whole can deviate substantially from fundamentals, whereas the relative price of individual securities would be relatively efficient. We survey the field of valuation ratios for markets as a whole rather than individual stocks and find that returns can indeed be predicted by more than would seem reasonable under the EMH. However, not all valuation ratios are equal and some have substantially better forecasting power and theoretical robustness than others.

The concept of intrinsic value is nearly as old as financial markets themselves, beginning with the South Sea Bubble of 1720, where wise observers recognised that shares were traded at substantial premiums to their 'intrinsick value'. Throughout the 18th and 19th centuries, the dividend was the mainstay of valuation methodologies: profit numbers weren't to be trusted and to the extent they were used, it was in the context of the sustainability of the dividend.

More formal and quantitative models for share valuation had to wait until the early 20th century. The dividend discount model (DDM) developed by John Burr Williams (JBW) argues that a share is worth the sum of all its future dividend payments discounted back to the present value. For investors who are more immediately focused on the prospect for capital gains this might seem anomalous. JBW's insight was that in the end, stocks only have value because they will return cash shares to shareholders.

If the expected returns on stocks were constant then the EMH and the DDM would imply that when dividend yields were low, subsequent dividend growth should be high and vice versa. With increased data availability and advances in data and computing, Professors John Campbell and Bob Shiller in the 1980s and 90s tested this. The surprising result was that when the dividend yield on the S&P500 was high, dividend growth was also subsequently high. Instead a low dividend yields forecast that returns were likely to be low instead.

Despite its academic literature dominance, the dividend yield is far from perfect. Changes in the payout ratio can change the equilibrium dividend yield and distort the message. We thus need to consider more appropriate measures that have sound fundamental principles such as simplicity; consistency with economic theory; measurability; stability and forecasting power.

Probably one of the most popular methods of calculating value is the Cyclically Adjusted Price-Earnings Ratio (CAPE ratio), which recognises that earnings are in some respects a better indicator of value than dividends. Retained earnings are also important to value as they help grow the business and provide the ability to higher dividends in the future. However, without some form of cyclical adjustment it could end up give misleading signals: no one would claim that the stock market was valueless if in the depths of the recession earnings in aggregate are negative. CAPE was originally developed by Ben Graham and David Dodd in their seminal work Security Analysis. Bob Shiller then highlighted this ratio in 1999 when it suggested that the US stock market was at the most expensive it had ever been.

Chart 1: Annualised Subsequent 10 Year Real Returns on the S&P 500 versus the level predicted by the CAPE Ratio and a Simple Trailing P/E 

The current level for Shiller's CAPE doesn't suggest particularly exciting returns from equities over the next 10 years, with returns forecast to be around 2%. This still suggests real returns better than can be achieved on index-linked instruments such as US Treasury Inflation Protected Securities (TIPS), where the yield is currently only 0.46%.

We have found that the best performing variable is perhaps the least well known: a monthly proxy for the q ratio. The q ratio is a measure of the total market value to the net worth of companies.

Chart 2: Annualised Subsequent 10 Year Real Returns on the S&P 500 V Returns Predicted by q Ratio 

Unfortunately q doesn't suggest a good outlook for the US equity market suggesting that returns will be not much more than 1% in real terms.

All valuation models can be criticised – if they weren't, consensus would rapidly converge on the best model. However, we shouldn't ignore the message from those measures which are empirically and theoretically robust. Unfortunately current valuation suggest that real returns on the S&P are likely to between 1% and 5%, over the next 10 years. This might seem low, but must be put into the context of low returns achievable on other fixed income instruments.

We must therefore look beyond the US market. Other equity markets look more attractive, particularly in India and Africa. We have also identified opportunities in alternative assets such as infrastructure where prospective returns versus US equities are higher despite greater certainty about futures cashflows.

Whilst it is not our central scenario, the more robust indicators of value are starting to look high and the chances of a correction to equity prices are starting to grow, albeit from low levels. The subdued volatility in equity markets that we have enjoyed over the last few years may soon be coming to an end.