Below is an extract from a commentary posted at Last year at around this time we asked the question: "Has a bear market started yet?" We thought this was a reasonable question because the NYSE advance-decline line had been trending higher for three years and several important sectors of the market were clearly not in long-term downward trends. And the question seems even more relevant today than it did 12 months ago because the NYSE advance-decline line hit an all-time high early this year, and although the senior stock indices have never looked like challenging their 2000 peaks many sub-indices have moved to all-time highs over the past 12 months. The question really doesn't exist, though, if we make the sensible decision to define the secular trend in terms of valuations rather than prices. When we try to define secular stock market trends in terms of prices we run up against two problems. First, nominal prices are expressed in terms of a currency that is constantly changing. Second, when a currency depreciates as a result of inflation an effect, at least during the early stages of the inflation, can be higher earnings and higher stock prices. It is possible, however, to eliminate the 'smoke and mirrors' effect that massive monetary stimulation can cause because over long periods the stock market has always differentiated between real earnings growth and earnings growth caused by inflation. Specifically, investors have historically paid less for earnings that stem from inflation than for earnings that stem from real (productivity-driven) growth. In other words, if earnings are being artificially boosted by currency depreciation then P/E ratios will trend lower.
Further to the above, if we define secular trends in terms of the average P/E ratio, as opposed to the nominal level of a stock-market index, the secular trends will not be obscured by changes in the value of the currency. For example, the below chart shows how long-term trends can appropriately be defined by the trends in the S&P500's P/E ratio.
The logical conclusion is that a secular bear market began following the major peak in the P/E ratio during the first two years of this decade and will, if history is any guide, continue until well into the next decade.
The one problem with defining secular trends in terms of P/E ratios is that the P/E ratio will sometimes fluctuate wildly during any given 1-5 year period. As a result it is really only possible to 'see' the trend in the P/E ratio when looking at very long-term charts, whereas during a normal investment timeframe the secular trend in the P/E ratio might not be evident or even relevant. Fortunately, there is a method of defining the secular trend in the stock market that gets around the problem of nominal prices being substantially altered by changes in the currency AND is of more practical use than the long-term trend in the average P/E ratio. The method, which we've illustrated on the below chart, is to let the trend in the Dow/gold ratio define the secular trend in the stock market.
Note: Gold and the US$ were officially linked prior to 1971 and this resulted in a very short and very sharp downturn in Dow/gold during the early 1930s followed by a very lengthy recovery into 1966, but if the US$ had not been convertible into gold at a fixed rate we suspect that the downward trend in Dow/gold that began in 1929 would have been more gradual and would have bottomed during the first half of the 1940s.
The above chart suggests that a secular bear trend in the stock market is in its early stages in terms of both time and price. In fact, if history is any guide the bear trend will end sometime next decade at a Dow/gold ratio of between 1 and 5. With the ratio presently at 27 this means that it would be reasonable to expect an investment in gold bullion to perform at least 5-times better, and perhaps as much as 27-times better, than an investment in the Dow Industrials Index if both investments were held over the next 10 years.
