Musings on a 1963 Ben Graham Lecture
Yesterday morning as I was shuffling though some papers in preparation for a short trip, I came across a copy of a lecture given by Ben Graham in 1963, titled “Securities in an Insecure World” (originally posted by Jason Zweig). I don’t know about you, but I really enjoy rereading older articles and lectures because they tend to give historical perspective, plus we can see if history has borne out the writer’s thesis, so I shoved it into my travel case for a relaxing read on the train.
Later as I was reading through the lecture I was reminded of an idea that has been haunting me for a while. I’ve been wondering whether there isn’t a viable investing strategy that targets being 100% invested only after a precipitous market decline, then gradually moving to 100% cash as the market stabilizes. This would be the ultimate ‘time arbitrage’ strategy. Intellectually, it seems attractive and even viable in light of the last 10 years of US stock market history, but I suspect the difficulty is in the implementation. What are the exact buy signals, when does one beginning moving to cash and at what point should one be at 100% cash? Furthermore, it has to be considered that such a system might leave one 100% in cash for significant periods of time, making it psychologically difficult to implement (but perhaps because of this, much more intriguing).
The first problem is when to move from cash to equities and vice versa, how much and how fast. Since market bottoms and tops can never be pinpointed except in hindsight, this leaves us to figure out some system that would indicate generally, not precisely, when to begin moving from cash to equities or equities to cash. One might think to use the current market (let’s use the S&P 500 Index as a stand-in) P/E ratio compared to the average market P/E over say the last 100 years, but Graham warns us off this measure quickly noting that P/Es are often highest at market bottoms as earnings disappear rapidly in economic downturns. He then contemplates a P/E measure where the denominator, ‘E’, is the average earnings over the past 10 years. This might help us identify market bottoms by smoothing the earnings; the current price will vary considerably at market inflections whereas the 10 year average of the denominator will be only slightly impacted by the loss of earnings during a recession. To take advantage of this 10 year P/E measure we might first want to see what has happened to the ratio over time. Are there any discernible patterns that we might take advantage of? And, more importantly, using this ratio in comparison to the average, can we identify market bottoms and tops?
Luckily for us non-statisticians Robert Shiller has made the raw numbers of such a 10 year P/E series for the S&P 500 available on his Yale website, and the series goes back more than 100 years. What do we see when we plot the Shiller 10 year P/E for the S&P 500? Well, there is kind of a wave pattern. The P/E expands over a number of years, then contracts over a number of years. Each ‘wave’ is somewhere between 15 and 21 years long, most being in the 15 to 18-year range. That kind of long wavelength doesn’t really bode well for my hypothetical market timing system. I don’t think most investors would be interested in a system where funds sit in cash for 16 years, me included. Furthermore, market bottoms and tops are not consistent over time making it difficult to pinpoint them with any accuracy. The Shiller 10 year P/E bottomed at around 5 in 1921, at around 10 in 1951 and around 8 in 1982, not consistent enough to construct any kind of investment rule.
Where does this all lead me? Well, for one thing away from any macro-based system using P/E ratios to buy and sell indexes or time my equity transactions. That doesn’t, however, mean that I think the Shiller 10-year P/E is a meaningless statistic. Rather, it may be the best macro ‘background’ statistic available. For example, right now we are about 10 years into a contracting P/E cycle, which means that if history is any guide, we should see the contraction continue for another 5 to 8 years. This contraction, of course, could be achieved by either increasing the ‘Earnings’ factor in the equation or decreasing the ‘Price’ factor, or some combination of the two. I will be hoping, naturally that it is the first, that earnings growth drags our P/E down to single digits.
This extrapolation of the P/E waves is not exactly a forecast in the traditional sense, and one which has no immediate application in terms of a buy or sell signal for either the market or any individual security, but it does provide some ‘color’. And I firmly believe that this is about as much market forecasting as can be helpful anyway. As an aside, I am time after time appalled by the number of stock market pundits who find a willing audience for their bombast about where the economy and the stock market are headed. Think about it! If they were good forecasters they wouldn’t be telling you or anyone else what was going to happen; they would be putting their chips down on the appropriate companies to own. It is singular, in my opinion, that anyone actually listens to them. Have their forecasts proved to be consistently right in the past? (not like the stopped clock, right twice a day). If not, then why listen to them? But it is in the nature of human being to seek to ‘see’ the future. Will we never learn?
OK lets get back on track. I’ve disabused myself about using the P/E ratio to time my market purchases. What other words of wisdom does Graham offer in his 1963 leture? In discussing investing guidelines he provides us with some sage advice that is not at all quantitative. He says that we should select investing rules that are in accordance with our own temperament, but in all cases ones that will always provide us with some exposure to common stocks at all times because otherwise we might become so disappointed if we are 100% in cash and the market advances significantly it would “ruin [us] from the standpoint of intelligent investing for the rest of [our] life”. His guidelines appear to be an equity exposure of not less than 25% and not more than 75%, with the balance in bonds or cash. But he doesn’t really address the more critical issue of ‘how much’ and ‘when’ the portfolio weightings should shift. Rather he leaves the definition of these a bit up in the air, saying only that whatever rules one uses should be “suitable for the [investor’s] point of view”. Well, maybe he’s saying that the absolute rules aren’t that important; it is the discipline of following the rules that is critical.
My advice? read the whole lecture. The words of the master are timeless!