Here's Why Stocks a Rallying

Norbert said on February 07, 2012

what Gold futures month are you using now?

TM said on February 04, 2012

We are definitely getting close to a pullback. We are starting to get outside the B Bands.......

TO said on February 03, 2012

Gold Bug, good stuff. I have been in cash waiting for 1340-1345. At about 1343 today, I shorted the FAZ for a swing trade. I did leave some dry powder to add a little more into further weakest or add higher. Using tight stops as we fill gaps at the upper end of the range. Cycles appear to be topping, I just hope it isn't a 2-3 week process.

Gold Bug said on February 03, 2012

I would like to share this Newsletter with members of the board. It's a bit long winded but very informative and interesting. There are no charts to look at but you will get the gist from the explanations. Here we go.

As a financial advisor, I’m often invited to investment conferences. Most are sponsored by fund managers – they’re hoping I’ll recommend their funds. So their presentations contain too much marketing and too little analysis for my taste. But still I go. Because sometimes you get an interesting character with a story to tell. And it can spark off an idea or two that is worth researching.

Three weeks ago, I was at the 2012 Unique Boutiques conference. Every year, five of the best independent managers make their pitches. This year I was treated to a classic bullish presentation. It was called, “Inconvenient Truths for the Bears”. And it laid out the reasons why you should buy stocks now:

Price earnings (PE) ratios are near 15 year lows
Earnings yields (1/PE) are high relative to bond yields
Dividend yields are well above the 15 year average
Dividend yields are high relative to bond yields

After the speeches, I spoke to the other delegates over coffee. Everyone was convinced. These four metrics prove that stocks are good value. It’s true that stocks have done badly for a decade, they conceded. And the economy is weak. But that’s why stocks are cheap! Surely now is a good time to buy?

In this week’s issue, I thought it would be fun to test these ‘truths’. You’ll see that they are really dangerous. So my mission this week is to prove to you that they don’t work. Instead, I’ll show you one unknown signal that has a terrific track record. And it’s screaming sell right now.

Since we’re talking about earnings this week, I also want to update you on fourth quarter 2011 US company results. As you know, I’ve been expecting earnings to peak in Q4 as companies rushed to take advantage of the tax incentive on investment spending before it expired at the end of last year. And there have been some very interesting developments on that front.

But first let’s test the four measures above for their effectiveness as buy/sell signals…

PE ratio? You might as well toss a coin

I’m using US data for my tests because it’s got a longer history than the UK.

First up is the PE ratio – the most common metric used by investors to value shares. As you know, it is the current share price divided by the company’s earnings per share. The PE ratio on the S&P 500 index has averaged 15.5 since the data began in 1871. So if the signal works, you should buy stocks when the PE ratio is well below 15.5. And of course, you sell when the ratio is much higher than 15.5.

So let’s see if it works.

In Figure 1, the blue line represents the S&P 500 Index adjusted for inflation since 1871. It uses the left hand vertical axis. And it’s plotted on a logarithmic scale. That’s so a doubling, say, of the index always shows as the same distance on the vertical axis. But look closely - log scales compress distances.

You can see there have been 6 great buying opportunities – 1877, 1921, 1932, 1949, 1982 and 2009. And there’ve been seven great selling opportunities – 1906, 1917, 1929, 1937, 1969, 2000 and 2007.

Figure 1: Real S&P 500 Index and PE Ratio since 1871

Source: Professor Robert Shiller, Yale University, my calculations

The red line shows the historic PE ratio using the right hand vertical axis. If the theory works, peaks and troughs in the market should correspond to peaks and troughs in the PE ratio. The ticks show you when they do and the crosses when they don’t.

Notice that the PE ratio is volatile. It often gives buy or sell signals only to change its mind soon after. Look at the 1930s for example. In reality, there was one huge buying opportunity and two massive selling ones. But just try making sense of the signals from the PE ratio. You’ll see later that this is a drawback for all four measures.

Overall the PE ratio only works 7 times out of 13. For example, a soaring PE ratio gave you a strong sell signal in 2000. But the PE ratio was even higher in 2009. So the signal was saying sell when you should have been buying.

Here’s one of the problems. At the top of a market cycle, the PE ratio may be high because share prices are high – like in 2000. But at the bottom of a cycle, it’s high because earnings are low – like in 2009.

Today, the PE ratio on the S&P 500 is about 13.5. That’s quite a bit lower than the average 15.5. So should you buy? You might as well toss a coin.

The earnings yield relative to bonds is worse

Many bulls argue that you should buy stocks because they are cheap relative to bond yields. This line of thinking became popular after former Chairman of the Federal Reserve, Alan Greenspan, once said it was how he valued stocks. So it’s sometimes called the Fed model.

First you invert the PE ratio. That gives you a figure called the earnings yield. For example, if a stock has a PE of 15.5, its earnings yield is 6.5% (1 divided by 15.5). Then you subtract the yield on the 10 year government bond. That’s averaged 4.7% since 1871. So the difference, sometimes called the equity premium, has averaged 1.8%. If the signal works, you should buy when it’s a lot higher than 1.8%. That would make equities seem cheap relative to bonds. Sell when it’s much lower than 1.8%.

So let’s see how well it does. In Figure 2, I’ve simply replaced the PE ratio with the equity premium on the right hand vertical axis.

Figure 2: Real S&P 500 Index and Equity Premium

Source: Professor Robert Shiller, Yale University, my calculations

As you see, the equity premium averaged much more than 1.8% until the 1960s. Since then, it’s been much lower. So it’s not easy to use as a signal. That’s partly because bond yields exploded in the 1970s. The US 10 year government bond reached a yield of 16% in 1981 as inflation got out of control. It fell back to average levels by 2001. But the equity premium has not recovered.

This is not what you’d want to see with a signal. But I’ll be generous and ignore that problem. Let’s just check how well the peaks and troughs line up with the S&P 500 index.

The Fed model scores just five out of thirteen. Perhaps we should not be surprised. The PE ratio only worked six out of thirteen. And very high bond yields mean that the equity premium also fails to signal the enormous buying opportunity in 1982.

Today, the PE ratio is 13.5, so the earnings yield is 7.4%. The ten year Treasury bond yields 2%. So the equity premium is 5.4% - way above the long run average of 1.8%. Don’t be fooled. It means nothing – except perhaps bonds are expensive.

The dividend yield is better, but only just

Figure 3 shows the same graph but with the dividend yield (in light blue) on the right hand vertical axis. If the signal works, you should buy when the yield is high because you’re getting a lot of dividend for your money. And sell when it’s low. So how well does it work?

Figure 3: Real S&P 500 Index and Dividend Yield since 1871

Source: Professor Robert Shiller, Yale University, my calculations

Well, like the equity premium chart, it looks a little odd. The average yield since 1871 is 4.5%. But it’s been a lot lower in recent years. And that makes it very difficult to read. Again that’s not what you want for a clear and reliable signal.

Nonetheless, let’s assume we are aware of this change. So we’ll just look for peaks and troughs and see how they match up with the S&P 500 index.

As you see, the dividend yield signals five of the six buying opportunities. At a stretch, you might even say it signals the sixth too – in 2009. But I’ve counted it a miss because the yield is no higher than in 1969 when the right strategy was to sell.

In fact, I think I’m being generous to score it five out of six. Look at 1982. Yes, there’s a good buy signal. But by the early nineties it’s changed to a strong sell right in the middle of the strongest bear market in history. The same thing happens with the 1949 buy signal.

And the dividend yield scores only four out of seven as a sell signal. In two of the misses – 1917 and 1937- there is an initial sell signal. But they’re both quickly followed by an incorrect buy signal. So I feel justified in counting them as errors.

So all told, the dividend yield scores nine out of thirteen. That’s a better than PE, but still only 69%. So you daren’t risk your life savings. And even if you did, the dividend yield today is much lower than average. It’s barely higher than in 2000. So you’d be selling not buying. That’s not what the bulls say.

Dividend yield relative to bonds is hopeless

As you’d expect, measuring the dividend yield relative to bonds makes things worse.

Figure 4: Real S&P 500 Index and Dividend Yields relative to Bonds

Source: Professor Robert Shiller, Yale University, my calculations

Dividend yield relative to bond yield is also hard to interpret from the 1970s. And it’s only right six times out of thirteen. Look at the terrible error in 1982. The dividend yield is ten percentage points below the bond yield. It’s a screaming sell at the start of the biggest bull market in history.

It’s astonishing to me that these four signals remain so popular. Yes, if you pick and choose your examples carefully, you can find instances where they have worked. But over time, they work only about half the time. So, if you follow them, you’ll make some very big mistakes. And you’ll suffer big losses. So next time you hear a bull extolling one or more of these measures to persuade you to buy stocks, you can say its bull****.

Thankfully, there is one signal that has a near-perfect track-record.

Real earnings – the buy/sell signal that actually works

Figure 5 is like all the other charts I’ve shown you. But this time, I’m showing you earnings after inflation on the right hand axis. I’m using a log scale again. And I’ve included a dotted trend line to help you see the peaks and troughs. The only other change is an adjustment to the scale of the left hand axis. That’s simply so I can overlay the two graphs on top of each other.

Figure 5: Real S&P 500 Index and Real Earnings per Share

Source: Professor Robert Shiller, Yale University, my calculations

Sadly the charting package in Microsoft Excel does not have the flexibility to allow me to overlay the two graphs perfectly. But you can see that the fit is astonishingly close. The real earnings signal flashes correctly all thirteen times. And it also picks up other opportunities that the other signals miss completely. For example, it gives you great buying signals in 2003, 1992 and 1895.

But have you noticed something? Conventional wisdom is that you buy stocks when earnings are high. This signal says the exact opposite. Sell stocks when earnings are high; buy when earnings are low.

Does this seem counter-intuitive? The bulls certainly think so. What’s going on? Well, the market is smarter than the bulls realise. It knows that there is a business cycle. Earnings hit their peak at the top of the cycle. And then they’ve only one way to go – taking stocks down with them. Similarly, earnings bottom out at the trough of the business cycle. As they rise, so do the stocks.

This explains why markets often trade at low PE ratios when earnings are peaking – like today. And they’ll trade at high PEs when earnings are bottoming out. Of course, the dot.com boom was an exception to this rule. Investors went mad and valued bubble-like earnings at bubble-like PE ratios.

So what does the real-earnings signal tell you to do right now? Well earnings recovered incredibly quickly after the 2009 trough. They’re now higher than in 2000 and 2007 – when the last two savage bear markets started. In fact, real earnings relative to trend have only been higher once before – in 1916. So we can’t be very far from peak earnings and peak stock markets.

And that brings me nicely to the Q4 US earnings season. Is there any evidence that earnings are peaking?

US company profits have already started down

At the time of writing, we’ve got consolidated results for 47.6% of the S&P 500 which have reported for Q4 2011. And S&P estimate that the final result will be $24.4 per share. That means profits are falling for the first time since 2008.

Figure 6: S&P 500 Companies Quarterly Earnings per Share

Source: Standard & Poors

This is one quarter earlier than I had expected. As best as I can tell, there are three reasons why. First is the slowdown in Europe. Second is a stronger dollar cutting US companies’ overseas earnings. And finally, the expiry of the 100% write down incentive on investment spending looks like it kicked in before the end of the year. To qualify, companies not only had to purchase the gear, they had to install it. That meant the effective deadline for complex software and machinery was well before the year-end.

The stock market hasn’t reacted much to this news. There may be two reasons for that. Few companies gave earnings guidance for quarter 1. Most preferred to forecast results for the full year which were ok. That seems very dodgy to me. And perhaps it’s why analysts have not adjusted their 2012 estimates. They’re still forecasting a 9% increase over 2011. Of course, they’ve never forecast a reduction in full year earnings in 26 years, despite that fact that it happens about one year in three.

I believe profits will fall further in 2012. Take last Friday’s US GDP figures for Q4. The headline figure was 2.8%. But 1.9% came from rebuilding stocks – which won’t be repeated in Q1 2012. It might even reverse. 0.8% came from car sales and 0.6% came from business investment. Both of these are likely to have been boosted by the tax incentive. And both could easily be negative in Q1. And government spending fell by 0.9%. But this will persist into 2012. The government is cutting spending in part caused by the automatic spending cuts triggered by the deal over the raising the debt ceiling in August. Finally, European recession will continue to hurt US exporters.

I think first quarter GDP will be at best flat. It could easily be worse. And second quarter will only be a little better. This will cut into US company earnings.

And the longer term outlook is bleak too. The huge run-up in private and government debts over recent years represents spending that is borrowed from the future. So it’s given a massive boost to profits. Paying down debts will reverse this effect and cut profits hard.

None of this is reflected in stock prices. That’s why it’s time to top up our short positions.

Gold Bug said on February 03, 2012

Back in Dec Tom Demark called 1338 to 1342 on the S&P. His timing is out a month or so but could his call be prescient.

I have set my stop at 1346.

Jay said on February 03, 2012

Happy Friday Everyone!
The regulars know I do not post during the work day because of my employer's pesky Computer Use Policy that would characterize this as "blogging" - hate that word. But I am home today and taking in the market in all it's irrational bullish glory.

I am certainly not fighting it. My long holdings are on a roll. I am still buying beach chairs off season with the VIX.

Speaking of rolls: TM, you really have been prescient in your recent calls. If I may embarass you kind heartedly for a moment our newer friends may not know your original log on was "Trend Master" which you subsequently shortened. I thought it boastful at the time but you have certainly proven worthy since!

My 2x long on the S&P at the Tuesday close has worked out splendidly. Thank you Simon for pointing out the early February effect in your TF.

I know you are skeptical about this market as are many. And I know you do not advise going long as often as I wish you would - hell, Ron just did in his sector picks! But you have a delightful code I have deciphered: when you say "we must allow for higher prices" I go long:).

A peaceful and fun Super Bowl weekend to all. Hari, good luck to your Patriots. mpman, good luck to your Giants - hope the move to North Carolina is going well. -jay

TM said on February 03, 2012

Wow...the fundamentals continue to improve. Roy, I said the best chance for a pullback was next week due to the fact that we have had 31 trading sessions without pressure, the improvement in reports over the past 3 or 4 months have been spectacular, and I think are due to decellerate some in the next month or 2. NFP hit a homerun today. It will be difficult to keep this kind of momentum moving forward. I think we will start getting some profit taking soon. New highs on the NASDAQ today, the rest of the indexes are not far behind, and this has caught a lot of people by surprise. It will be difficult to sell QE3 at these levels. I would not short this market, but a fade at the 1342 might prove timely....

Andrew said on February 03, 2012

Correction

Andrew's Portfolio Update: 02-02-12:
Current Value: $96,585
2 days change -$3415
(all are SPXUs)

Trades:
SPXU @ 11.15
Stopped out @ 10.77

Cash only

Andrew said on February 03, 2012

Andrew's Portfolio Update: 02-02-12:
Current Value: $96,585
2 days change -$3415

Trades:
UVXY @ 11.15
Stopped out @ 10.77

Cash only

Gold Bug said on February 03, 2012

Just shorted S&P at 1340 in the futures market.

*** FUND MANAGER : Gold Bug ****

Portfolio Update: 02-04-12: 16:03 EST
Current Value: $100,000
Trades: NONE
Previous Holdings:
CASH: $100,000
Current Holdings:
CASH: $100,000

More Comments...