May 6, 2016
By Steve Blumenthal
“The Fed has borrowed from future consumption more than ever before. It is the least data dependent Fed in history. This is the longest deviation from historical norms in terms of Fed dovishness than I have ever seen in my career… This kind of myopia causes reckless behavior.”
At the beginning of each month, I like to look at equity market valuations. The stock market moved higher in April, yet for the fourth quarter in a row, corporate earnings were down. The good news about market valuations is that they can tell us a great deal about the annualized returns we are likely to get over the coming 10 years. The bad news is they tell us little about returns over the coming two years.
Buy low — sell high, they say. Pretty simple actually, but not many can actually do it. This is where valuations can help us get centered. Below I provide examples from several periods – low valuations in February 2009 and the great gains since.
Valuations were high in 2000 and 2007 and, as you will see, the subsequent returns from those starting points were not good. There were some better entry points, like in early 2009 but, really, who do we know that bought back into stocks then? It was pure panic, margin calls and forced liquidations. I wrote a piece that prior December entitled, “It’s So Bad It’s Good.” I didn’t get many positive responses.
Yet, here we are today — valuations are now higher than they were in 2007. As you will see on the coming chart, they are higher than every other period in time with the exception of time surrounding the great tech bubble.
We are getting more than a few “you didn’t beat the market” phone calls like I’m sure you are getting as well. The reality today is much like 1999 and 2007, far too many investors with false confidence in hand are buyers – not sellers. Cited are the returns of the market over the last handful of years. Behaviorally, this feels hauntingly similar to those prior cyclical peaks. Lessons not learned.
I have been writing about the global overcapacity glut in the last several letters. Let’s pass on that, let’s pass on the debt mess, let’s pass on underfunded pensions and let’s pass on the Fed. Valuations do matter, so let’s go there.
Before we do, I want to weave in a few recent comments that crossed my desk this week – and likely yours – from Stanley Druckenmiller. Stan spoke this past week at the Ira Sohn Investment Conference. The CliffsNotes version of his presentation follow:
- He is negative on China’s economy going forward
- He believes further stimulus in the Asian country will not work as it has “aggravated the overcapacity in the economy” (SB here – to which I add globally aggravated along with the other central banks)
- “Get out of the stock market and own gold.”
And here is the big point as it relates to today’s overpriced stocks:
- Stan added, “U.S. corporations have not used debt in productive investments, but instead relied on financial engineering with over $2 trillion in acquisitions and stock buybacks in the last year.”
- “Corporate books show that operating cash flow growth in U.S. companies has gone negative year-over-year, while net debt has gone up.” (SB again – big buyers of U.S. stocks have been the corporations themselves. I believe they are near the end of that runway due to negative operating cash flow.)
- “Higher valuations, limits to further easing… the bull market is exhausting itself.”
Druckenmiller is the Babe Ruth of investing. One of the all-time greats. His performance record, I believe, may be unmatched. “Get out of stocks”… “Buy gold”… Try explaining that one to your clients.
Yet, like all of us, he is far from perfect (as he’ll tell you); however, we should pause and give consideration to what he is saying. Now, on to valuations.
Valuations can tell us a great deal about what the forward annualized returns over the coming 10 years are likely to be. My favorite metric is a price-to-earnings ratio measure called median P/E. I believe it to be a fair and consistent way to look at current valuations and compare on the same basis to historical valuations (email me if you want a deeper explanation). In this way, we can see what forward returns might likely be.
The rules remain simple. If the price is excessively high relative to earnings, then you are not getting a good bargain and won’t make as much money. If price is low relative to earnings, you get really good bang for your buck.
Here is a look at a few sample periods in history (you may have seen this chart from me before):
Take a look at the median P/E of 11.0 the month the market made its low in February 2009. Did most clients get 16.69% annualized return since then? Well, for those of you that had little money in the market or were in high school or college back then, here is what happened: MOST PEOPLE PANICKED AND GOT OUT at the LOW. The single biggest month of outflows on record occurred that month.
Show me the brave few who bought then. I don’t recall getting any calls asking why they did so much better than “the market” back then. I can show you the many who are buying in today. Makes no sense, but it might be best for me to jump off my soap box and get back to valuations.
Now, if you invested at the market top in the fall of 2007, the median P/E was at 19 and the S&P 500 peaked at 1549.38. The S&P 500 closed at 2065.30 at the end of April 2016. That’s about a 33% gain over eight years and seven months or an annualized total return of 5.86% per year.
Ok, so where are we today?
Median P/E reached 22.7 at the end of April. That is higher than any point looking at median P/E data from 1964 to present with the exception of the crazed pre- and post-tech bubble period.
The next chart is courtesy of Ned Davis Research. The traffic light and arrows are my notations as I attempt to simplify the chart. What I like about this chart is that it does a good job estimating overvalued, fair value and undervalued levels on the S&P 500 Index. Kind of an investor reality road map.
With the S&P 500 Index at 2065.30, it (by this measure) means that the market is overvalued by 3% (red light) by historical measures. In the chart, NDR uses a 1SD (standard deviation) move above fair value (it uses the 52.2 year median P/E of 16.9 to determine fair value) to identify the market as overvalued at 2003.69.
In English, a one standard deviation move is a movement away from a historical trend – it is something that doesn’t happen very often. In the case of median P/E, a 1SD move has happened about 10%