If you believe you can make good money by selecting stocks with low P/Es solely, dream on. If it is that easy, there will be no poor folks. However, buying fundamentally sound companies would reduce risk and improve the chance of its appreciation.
P/E is the most misunderstood indicator. To me, it is the most useful one among all metrics if it is properly used. Earnings are the key to stock appreciation and P/E measures its value. To illustrate on P/E, you pay a million for a hot-dog cart in NYC. Even if its earnings increase year after year, you will never recoup your investment as you have paid too much even for a good business.
P/E should be inverted as E/P, which is termed as Earning Yield. Earning Yield is easy to be compared and understood. It takes care of negative earnings for screening stocks and ranking (comparing stocks with better P/E first). If you sort P/E in ascending order, your order is wrong with negative earnings but right with E/P.
It is usually compared to a 10-year Treasury bill yield (or 30 years) or a CD rate. If the stock has a 5% earning yield and your one-year CD is 1%, then it beats the CD by 4% in absolute number and four times better. However, the CD is virtually risk free (with deposit amount limit in most banks). Earning yield is an estimated guess and it may not materialize.
Many ways to predict E/P
- Based on the last 12 months. Project it to future E/P. It is also called the last twelve month E/P.
- Based on analysts' educated guesses. Guesses may not materialize. Based on my experience, the expected usually predicts better than the one based on the last 12 months. This is the one I use most and many investing subscriptions provide this expected P/E (same as Forward P/E) or expected E/P.
Usually, I do not trust analysts' opinions due to conflict of interest. However, the earnings estimate is my exception.
- Based on the last month or the last quarter. Latest information could be better for prediction. However, they are not good for seasonal businesses such as retail where most sales are done during the Christmas season.
- Besides the Pow PE described later, I take the average of the earning yield (EY) as:
The Avg. EY = (EY from the last twelve month + Expected EY + EY from the current month of prior year) / 3
It averages out using figures from the past, the present and the future. If no one has used it, I claim shamelessly it is my original idea.
Best E/P could not be the best
A very high E/P could be a sign of trouble ahead such as a lawsuit pending, fraud, etc. If you find companies with E/P over 50%, it means two years' profits could equal to the entire cost of the company! I can tell you right away they smell fishy unless you believe there is free lunch in life.
However, from time to time, some bargains do exist due to certain conditions or Wall Street is just wrong about the company. You need to find out whether they are bargains or traps. When the E/P is low (sometimes even negative) but is improving fast, it could mean a big profit for you. Fundamentalists may miss this opportunity in the early stage, which is also the most profitable time to buy. This could be a turnaround.
During a recession, a good company will have a hard time to promote a new product as the consumers are thrifty. At the same time, it usually is the best time to develop products if the company has enough cash to finance it. In this case, there is no alarm even with negative earnings. The only alarm is when the company cannot meet the debt obligations.
Some companies can manipulate earnings via dirty tricks in accounting. It could make this year look really good, but it is hard to continue the same trick for many years. Check out the footnotes in the financial statement.
E/P and PEG
For value investing, E/P is usually used and the higher the better. But watch out when it is extraordinarily high.
PEG (P/E growth) measures the rate of improving P/E.
PEG = (P/E) / Earnings Growth Rate
As described, earnings can be based on the last 12 months, expected or the average. I prefer expected earnings.
Which of the following two stocks do you want to buy based on their historical earning yields and earnings growth?
- A stock has a 10% earning yield with no earnings growth.
- A stock has a 8% earning yield with 50% earnings growth.
If the earnings growth continues, in next year the second stock should pay 12%, substantially better than the first stock. This is another reason we should use expected earnings rather than historical earnings.
PEG may give a low value for companies that pay high dividends. To correct it,
PEG = (P/E)/ (Earning Growth Rate + Dividend Yield)
When the general market favors growth stocks, weigh more on growth metrics, including PEG. I claim no credit on the adjusted PEG.
I modified P/E to take care of cash and debts.
Pow P/E = (P - Cash per Share + Debt per Share) / (Earning - Interest gained per share - Interest paid per share)
To illustrate, the stock price is $10 and it has $2 in cash (actually cash and securities). The real price of the stock is $8. When we ignore the cash and debt, we have to ignore the interests gained and paid.
Many companies park their cash in bank accounts that do not generate much interest today. Some cash-rich companies such as Microsoft (NASDAQ:MSFT) and Cisco (NASDAQ:CSCO) have better P/Es after excluding the cash per share.
My official definition:
The above is for simple illustration. I need to expand it here.
Pow P/E = (P - net short-term asset per share + liability per share) / (Earning - Gain from short-term asset per share - interest paid per share)
You can get the short-term asset and liability from the financial statement and divide it by the number of shares. In addition, pension liability should be included in the liability. As before, I prefer the expected earnings.
In this calculation, P could be a negative and so is E. It is misleading that when both are negative to generate a positive Pow P/E. Interpret the number accordingly.
When either one is negative, most likely the fundamental is not good. There are many examples of bankruptcy when both of them are negative. Avoid companies with negative earnings and high debt unless there is a good reason such as a turnaround. The previous GM was one of the examples in this situation.
Stock buyout could reduce the outstanding shares. It would improve most metrics that use earnings per share such as P/E. However, if the company borrows money to buy stocks, it would deteriorate the debt metrics or its cash position. Pow P/E handles this situation well if the financial statement is up-to-date.
Most likely, I will not get a Nobel Prize like Shiller on a new version of P/E. It is common sense to me. I have not heard others using this same concept, so as before I shamelessly claim it is mine. Ignore my ignorance if someone already uses this simple concept.
Using IBM as an example to calculate Pow P/E based on 12/31/2013 financial statements.
Pow P/E = (187.57 - 10.62 + 98.00) / 15.06 = 18.26
Take out the gain from short-term asset and the interest paid for simplicity.
The historical P/E was 10.41 on 12/31/2013 and it looked better than my 18.26. Mine is a better representation reflecting the high debt of the company. Hence, do not rush to buy IBM only solely considering the low P/E. However, even for a P/E of 18, it could still be a good buy.
Better representation does not mean better prediction. I do not have a historical database with all the information available. Vendors, if you have one, would you let me use yours to verify any improvement on predictability on this one and the Avg. P/E described before.
E/P is one of the metrics you should use but not exclusively. If the earning yield is high but the % of debt is high too, then a good bargain may not be as good as it appears to be. Pow P/E considers this effect.
Some other metrics may not be easily found in the financial statements such as the intangibles, insider buying, pension obligation, trade secrets, losing market share, brand name, customers' loyalty, etc. It is interesting that most metrics change its ability to predict from time to time as illustrated in my book Scoring Stocks.
There are other P/E variations like Shiller P/E (same as CAPE and PE10). Shiller P/E can also be used to track the current market valuation. It is controversial and its value is easily misinterpreted. Hence, use it as a reference only unless you understand all its issues. I prefer to use two-year average of the P/E instead of 10, as I believe the market changes too much for a ten-year span. Currently, Shiller P/E does not work that well as before. It is due to the excessive printing of money.
Personally, I prefer to compare a company's current P/E to its average P/E in the last 5 years. Also, compare it to the average value of the companies in the same industry. The average P/E for high-tech companies is different from supermarkets for example.
P/E is more reliable for a group of stocks (SPY for example) instead of individual stocks which have too many other metrics and intangibles to deal with.
When you compare the total return of an ETF to a corresponding index, you need to add the respective dividends to the index to ensure a fair comparison of total returns. Currently, S&P 500 is paying about 2% dividend.
Garbage in, garbage out
I do not trust most financial statements of emerging countries, especially the smaller companies. Watch out for fraudulent data. Most metrics can be manipulated. Recently, I had a US stock that lost 18% in one day due to the SEC's investigation of its financial data.
The announced earnings may not be reflected in the financial statements you use from the web. Ensure your data is up-to-date by checking the date of the financial statements. Seeking Alpha has transcripts for the earnings announcements that would save you a trip to attend the companies' quarterly meetings.
Sector and entire market
You can find the value of a sector using the P/E of an ETF for that sector. It is similar for the market. For example, use SPY (an ETF simulating the S&P 500 index). If it is lower than the average (15 to me), then most likely the market is good value and a buy signal. It is one of the many hints for market timing.
Where to use P/E
Each highlight of the following corresponds to one of my books. Click it for the description of the strategy.
My book The Art of Investing describes how to evaluate stocks and time the market. It has most ideas of most of my books. The top-down approach starts with a safe market, then sector analysis, fundamental analysis, intangible analysis and optionally technical analysis. P/E is one of the many metrics in fundamental analysis.
There are many styles of investing. In general, fundamental analysis is important when you hold the stock longer.
- P/E is important in Long-Term Swing, Dividend Investing, Retirees and Conservative Strategies.
- P/E is moderately important in Short-Term Swing and Sector Rotation.
- P/E is least important in Momentum Strategy and Day Trading.
Again, one metric should not dictate the reason to trade a stock. Compare the company P/E to its industry average and its own five-year average. In addition, many industries have cycles. If you buy it at the peak of the industry, the P/E may mislead you. Besides fundamental analysis, you need to consider intangible analysis and time the entry / exit point by using technical analysis.
- Industry is a subsector. For example, Computer is a sector (for most) and Computer Software is a subsector.
- Top-down Investing. If the market is not risky, check out the industry and then find the best stock within the industry.
- Currently, P/E is widely used compared to E/P (Earning Yield).
- This is a continuation of my last article, A Tale of Two Portfolios.
Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.