Equity valuation: no bargain basement
On traditional valuation measures, shares have remained quite highly valued and even after the most recent falls, taking into account lower earnings estimates, they are still not unusually cheap. We think this explains much of the weakness of equity markets, quite apart from the particular problems of financials, and also reduces the scope for sustained recovery from these levels.
Warren Buffet, writing in the Sunday Times on 19th October, says he is personally buying equities and compares the buying opportunity with the 1930’s and the wartime bear market that ended in 1942. Both are instances of investor capitulation in the face of extreme threats to capital which, with hindsight, marked both a short-term selling climax and an excellent long-term investment opportunity.
Buffet wisely admits in the same article that he cannot tell whether the market will be higher or lower in a year’s time. If so, his advice must implicitly be even more dependent on being right about capitulation and, if a sustainable turning point is in order, on the degree of long-term undervaluation in equities. On what does he depend?
Our view is that earnings estimates for companies in several markets (notably including the US) have been much too high, even relative to aggregate forecasts for economic activity. This gap between aggregate and individual forecasts often arises when sentiment deteriorates but there does seem to have been a high level of denial about the impact of the banking crisis on sales and profits of businesses. Though the markets have fallen by around 20% since this denial started to be challenged, estimates have only just started to be cut back realistically.
Our own measure of equity valuation also suggests they are not extremely undervalued. The measure is the ratio of the current index of real total return to its own regression trend, for as long a history as we can rely on. Though agnostic in estimating the inputs to equity value, we feel we can base decisions on the outputs, in the form of deviations from trend.
Deviations above trend will tend to lead to lower future real returns and vice versa.
The matching of expected equity returns to specific time horizons is based on our observations about the slow speed of mean reversion, or the return to trend from any point of deviation.
The uncertainty or forecasting error associated with the mean expected equity return is a function both of the past observed errors or deviations and the amount of historical data we have to work with for each market we forecast (the less history the more uncertainty).
In the chart below we have detrended the real total return indices to highlight the recent levels of deviation, as the key information used in investment strategies. The trend we have removed is about 6% pa. A ratio of 100 means the index is in line with its trend. A flat line means the rate of growth is in line with trend growth of 6%. The ratios have historically erred between twice trend (at bull market peaks) and half trend (at bear market lows) but not all deviations reach extremes before reversing.
These ratios are now (14th October) back to the levels last seen at the bottom of the bear market in 2003. (It is only the trend growth rate that accounts for actual market levels being higher, but that change in level does not tell us about valuation. However, those ratios were themselves well above previous cyclical low points for all but Japan). The fact that the UK, Europe and the USA all bottomed in 2003 without reaching very low valuations was probably due to the scale of monetary stimulus, from which many of the subsequent credit problems stem.
For the current levels (other than for Japan) to prove a cyclical low point might imply either successful attempts at monetary stimulus (which seems most unlikely) or a reversal of the recent trend of analysts cutting their earnings estimates (also unlikely).
The lesson of whole histories of past deviations from normal value is that the current difficult investment climate could persist for many years, as happened in Japan in the 1990s. In that case, ‘calling the bottom’ may be better described as attempting to call a bear market rally. There may be many false bottoms.
Indeed, if markets fall further it will be tempting to view the 2003/7 bull market as a brief interruption to a primary downtrend in equity returns that started with the exceptional overvaluations of 2000, of close to twice normal.
Perhaps Warren Buffet is really only doing what we are doing. We too are agnostic about market timing but we believe we know about long term value. As value increases, we add to positions. Rather more dull than waiting for what may or may not prove to be a selling climax but effective nonetheless.