	Subscription Information \| About SMI
	Search:

Home : 2010 Back Issues : Article

Earnings Expectations, P/Es, and Assessing Risk

By Austin Pryor

Share |

Because the financial markets constantly speak a language of numbers, we get the impression that investing is like scientific research — when you add up all the relevant facts, you can arrive at the "correct" answer. We assume that investing decisions can be made with some degree of mathematical certainty.

None of this is the case. Investing is more an art than a science. The markets are merely collections of people who act according to their emotions and desires of the moment. We've all observed (and probably contributed to) how people act differently when others are present than when they are acting alone. Gustave Le Bon wrote his classic study of "crowd psychology" in 1895, but peer pressure has always been a part of human nature — you might even say Adam fell victim to it.

In the same way, the herd mentality in the stock market often leads investors to actions they later regret. We saw what happened to investors who threw valuation concerns out the window during the 1999-2000 bubble in tech stocks.

In our last issue, we began discussing a valuation tool that can provide a reality check and help investors keep their bearings while sailing on the sea of investor emotions. Ultimately, stock prices are driven by the amount of profit a company is expected to make. So, investors turn to the "price-to-earnings ratio" (P/E), a measure used to describe the relationship between a company's stock price and its profits. In other words, how expensive is a stock in relation to the earnings the company generates each year?

The P/E is most commonly calculated by dividing a company's price by the earnings the company has reported for the most recent four quarters.

As an example, last month we looked at Merck (ticker: MRK), and the fact that it had reported $5.65 in per share profits over the previous four quarters. At its then-price of $35 per share, it had a P/E of 6.2 ($35 divided by $5.65). Stated another way, investors were willing to pay $6.20 for every $1.00 of MRK earnings. We discussed how to put that information into historical perspective to see whether Merck was bargain-priced, fairly-priced, or over-priced.

We also noted that you can compute P/E ratios for stock indexes and stock funds as well as individual stocks, and we looked at the historical results when buying the blue chip S&P 500 stocks at various P/E levels.

We covered a lot of ground, but there is a conceptual problem with the P/E that we didn't have time to tackle. The kind of P/E calculation that we discussed uses earnings that have already been reported; it's referred to as a P/E that's based on "trailing" earnings. When you see P/E ratios in the media, they typically refer to this kind of "looking back" approach.

The problem is that the market is always looking forward. It really doesn't matter to potential investors what a company's earnings were last year; they want to know what profits are likely to be next year, and the next, and the next. That's where Wall Street analysts and their earnings forecasts come into the picture, and where we pick up our story.

There is another widely used way of computing a P/E ratio. Rather than use the earnings reported over the past year, the calculation is based on what the operating earnings are projected to be in the coming 12 months. This kind of P/E is based on "forward" earnings.

Let's review the math to see why this often makes a stock appear to be a better buy. For our example this time, we'll use another member of the Dow Industrials, Cisco Systems (CSCO). Looking in the rear view mirror, the company reported earning $1.18 per share over the previous four quarters. At Cisco's current price of $23.50, investors are paying $19.92 for each dollar of last year's earnings.

However, Wall Street analysts are currently projecting that CSCO will earn $1.60 per share in the current fiscal year. This gives it a forward P/E of only 14.7 ($23.50 divided by $1.60). Looking ahead, then, investors would only be paying $14.70 per dollar of this year's anticipated earnings. Voilà! — the stock suddenly appears to be more reasonably priced.

Obviously, there's a lot of guesswork involved in a forward-based P/E, but to the extent it is close to being accurate, it does represent a better value benchmark. Unfortunately, it's the "being accurate" part that makes the forward-based P/E problematic.

The table at right shows the initial earnings forecasts (on average) from about 200 Wall Street analysts made each December for the coming year. Their initial projections are then compared to the actual operating earnings of the companies in the S&P 500 index for that year.

It's not a pretty sight. The analysts frequently are wrong by double digit margins, typically erring on the positive side. Without exception in the past 15 years, analysts expected profits to rise in the coming year. The analysts' optimism leads investors to have more confidence over the short-term than is justified.

As can be seen in the table, this pattern appears to have been altered, at least temporarily, following the 2000-2002 bear market when the analysts were caught with their rose-colored glasses on. Chastised, they were actually too conservative; profit gains were higher than projected from 2003-06.

But they swung back to an optimistic perspective just in time to wildly overstate what corporate profits would be in 2008-09. They obviously were unaware of, or ignored, the risks the housing bubble posed for the economy. In any event, the fact that the projections are not 100% accurate is well understood.

Let's now look back at 25 years of history to see how the market has behaved at various P/E levels, both for trailing and forward-looking P/Es. In Table A below, we have shown the present trailing P/E of 18.3, and where it fits on the bargain-to-expensive scale. It's neither too hot nor too cold, but is right in the middle at "fair pricing."

If you look at the "Average 1 Year Later" column, you'll find that, on average, the market (as measured by the S&P 500) is 6.5% higher one year after "fair pricing" readings. In those instances during the 1984-2009 test period, stock prices rose 73% of the time over the following 12-month periods, and fell just 27% of the time.

You can see that risk increases as you move up the table, and by the time you get to trailing P/Es of 30 and higher, the market has dropped one-half of the time in the year following those "expensive" prices. Clearly, when you pay top dollar for stocks in relation to their earnings, your chances of profiting in the near term fall significantly.

Table B (below) is structured similarly, but uses the calculation based on forward earnings. The basic risk/reward relationships remain the same; only the P/E levels differ. Because forward earnings are typically higher, they result in lower P/E numbers.

Currently, with the S&P 500 at 1,115 (as of mid-June) and analysts projecting earnings of $87.60 over the next 12 months (from the second quarter 2010 through the first quarter 2011), the forward P/E is 12.7 (1115 divided by 87.60). By this measure, current prices seem undervalued in relation to profit expectations.

Price-to-earnings ratios can be helpful in understanding how reasonable prices are at a given time. Just be sure you're clear as to whether the commentary you're reading is based on the recent past or analyst projections of what the future holds, and refer to the appropriate historical tables.

Even then, of course, there are no guarantees; it's principally a matter of getting the probabilities on your side.

Austin Pryor has three decades of experience advising investors, and is the founder of the Sound Mind Investing newsletter and website. Austin lives in Louisville, Kentucky, with Susie, his wife of 46 years. Two of his sons, Andrew and Matthew, work with him at Sound Mind Investing. A third son, Tre, is a Realtor® in the area.