Stock market allocations should be reduced

SERIOUS MONEY:It is a year since major indices fell to crisis lows, but do current stock values merit long-term investment?

IT IS A year since major stock market indices tumbled to their crisis lows on fears that the overleveraged global economy would succumb to a prolonged and painful debt deflation.

The 17½-month decline wiped almost $8 trillion (€5.87 trillion) from the market capitalisation of the SP 500 alone and the cumulative 57 per cent drop was surpassed in magnitude only by the savage bear market that extended from the autumn of 1929 to the summer of 1932.

Government and central bank support on an unprecedented scale prevented the worst- case scenarios from becoming reality. In response, stock prices bounced almost 70 per cent and registered the most rapid advance in market indices since the autumn of 1932.

It is clear that stock prices had dropped to relatively attractive levels 12 months ago. One year on, it is reasonable to ask whether current values merit long-term investment.

The valuation of common stocks is the cornerstone of successful investment for long- term investors. However the output arising from a sensible valuation exercise often proves controversial and is frequently ignored, to the detriment of long-term performance.

The decision prescribed by valuation models, either to reduce or increase weightings, is typically avoided because factors other than valuation drive short-term performance and overconfidence leads many to believe that they can time the market, despite overwhelming evidence to the contrary. It is clear, therefore, that a valuation model should be used to provide an anchor for fair value and to shed light on the risks that are being borne via current allocations.

John Burr Williams concluded in his seminal work, The Theory of Investment Value(1938), that "a stock derives its value from its dividends, not its earnings. In short, a stock is only worth what you can get out of it. Even so, spoke the old farmer to his son: a cow for her milk, a hen for her eggs and a stock, by heck, for her dividends."

The resulting valuation model though did not garner widespread appeal as the stock market was largely shunned by the population at large after the Great Depression, while the complexity in calculation ensured that the relatively simple model did not come to the fore until the advent of the PC decades later.

The dividend discount model is the most basic model for valuing stock today. It has intuitive appeal because it assumes that asset value is a direct function of the cash flows expected in the future.

When investors purchase stock, they expect to receive two forms of cash flow: the dividends during the period they hold the stock and the expected price appreciation over the holding period.

If an investor sells the stock, the purchaser of the stock is buying the remaining dividend it pays. Thus, the relevant cash flows are the dividends paid plus the value of the stock when sold. In practice, this requires an investor to estimate the discounted stream of dividends over a specified forecast horizon – typically, seven to 10 years – and a discounted terminal value determined by the cost of equity capital and long-term growth expectations.

The first critical input to the model is an estimate of normalised or cyclically adjusted earnings per share. The use of relatively simple statistical techniques reveals an estimate of $63 a share using reported earnings or $78 a share using operating profits.

Exceptional and extraordinary charges should be included in any valuation analysis, not only because they are borne directly by shareholders, but because the supposedly once-off charges happen year in year out and have averaged close to 20 per cent of operating earnings annually over the past 20 years, or roughly two cents a year for every dollar of existing capital stock.

However, the choice of earnings figure matters little when using the dividend discount model, because the respective long- term payout ratio for either operating or reported profits will yield an equivalent estimate for normalised dividends.

Market analysts expect SP earnings to reach our normalised estimate this year and our model assumes that profits will grow at a rate of 6 per cent a year over the 10-year forecast horizon. This number assumes an inflation- adjusted annual earnings growth rate of 3.5 per cent. This is well above the long-term trend but it is possible, given the earnings accretion that may arise from share repurchases conducted via employment of the almost $1 trillion in cash sitting on non-financial SP 500 balance sheets. It also incorporates an inflation rate of 2.25 per cent as reflected in current US treasury yields.

Additionally, the long-term earnings growth rate applied in determining the terminal value is reduced to 5 per cent to match the economy’s trend rate of growth, even though history shows that it is typically one percentage point lower than economic growth due to the dilutive impact of share issuance arising from new business enterprises.

The discount rate or cost of equity capital applied in the analysis is determined by adding a risk premium of 3½ percentage points to the current yield on 10-year treasuries. History supports the use of a risk premium in the range of 300 to 400 basis points (three to four percentage points) and the estimate employed attempts to capture normal market conditions, which should prevail over time.

The analysis reveals that the SP 500 is fairly valued at current levels. However investors should note that, not only are the estimates of dividend growth based on relatively optimistic assumptions, but they are inconsistent with current treasury yields of 3.6 per cent. The employment of consistent inputs via either higher yields or lower dividends points to an overvalued market. Even though the cyclical bull market rolls on, allocations should be reduced. Caveat emptor.