I don't mean high multiplier as in 40 and up. I mean in the high teens, maybe 20 or so.If you look at stocks that have great business futures like Coca-Cola and Yum! Brands and use the current purchase prices and a conservative projection of future earning power you will find that you will make more money in a 10 year period just by buying a AA corporate bond.However, sometimes many investors are able to buy Coke and Yum or something along those lines for as much as 25 times earnings. And the general consensus here is that these companies are some of the best to invest in.How do you justify a purchase like this if the bond will earn more money?
However, sometimes many investors are able to buy Coke and Yum or something along those lines for as much as 25 times earnings. And the general consensus here is that these companies are some of the best to invest in.How do you justify a purchase like this if the bond will earn more money?Equities have several (dozen?) things that might make them more attractive to some investmentors than bonds at the same earnings yield.1. If inflation rises my bond will suffer. But the company may be able to raise prices, thus preserving margins and real return to the equity.2. The company may improve their operations and the fruits of that improvement will accrue to the equityholder while the bondholder gets the same old coupon.3. I'm just starting out and I can't achieve a reasonable level of diversification buying bonds at $10K each.4. This company's going to the moon! I expect this stock to double over the next 3-4 weeks and those poor bondholders get nothing but their measly 5.5% per annum. What dolts!Regards,Tom
GE and Berkshire Hathaway (BRK) are kind of strange critters. They are real, publicly traded companies. However, the depth and breadth of their holdings makes them much different from your typical company. Both perform a lot less like stocks and more like mutual funds. Their senior management actually acts a lot more like fund managers. When it comes to investing in GE and Berkshire Hathaway, it's best to think of them in those terms. They have consistently produced mutual-fund like results that beat the pants off any high quality bond investment. So if you goal is long term buy-and-hold, then it doesn't matter at all what the price is. Just buy any time and hold. Even if you bought when the price was a bit high or low, within a couple years, this difference will be quickly erased. The downside of these kinds of companies is that there is very little chance of getting multibagger results. The upside is that they are going to plod along at the same pace almost without regard to the market around them over the long term.Another example is if you are doing dollar-cost-averaging. This is a type of long term buy-and-hold strategy where you invest small amounts of money on a regular basis to smooth out the "bumps". Instead of predicting when the stock price is excessively high or low, the goal is to simply buy the stock at any price at regular intervals. Sometimes it will be high and sometimes it will be low, but with enough intervals, you will be buying at an average price.
So if you goal is long term buy-and-hold, then it doesn't matter at all what the price is. Just buy any time and hold. Even if you bought when the price was a bit high or low, within a couple years, this difference will be quickly erased.Data will help here:http://invest.kleinnet.com/bmw1/statsdjia/GE.htmlHaving bought GE in 1973 would have shown a loss of 50% within a year, and no gain until 1980. By 1990, after 17 years, you would have been at about a 12% CAGR, finally acceptable. Note that these returns include dividends.Having bought GE in 2000, you would be sitting on over a 40% loss today. That would be pretty disheartening after 8 years, right?On the other hand, if you had bought GE in 1974, by 1987 when it had completed a long rising run and was above its average growth rate, you would have achieved a gain of 12x, or 1100%, a CAGR of 21% per year. That is pretty impressive.In other words, I would never ignore valuation, if at all possible. Buy great companies when they are unusually low priced.Dollar-cost-averaging is not a bad approach as a default, leave-it-alone strategy. But if you follow that strategy, you must do two things absolutely:1) you must be buying into great companies that have decades of good performance ahead of them, not a biotech or solar energy startup2) you must dollar cost average regularly, on a set schedule, over a very long time period-Mike
Data will help here:Thanks.My response is that there are some investment strategies that will consistently beat bonds and have almost zero required brain power to apply that do not involve mutual funds (the criteria). These strategies will beat out bonds very trivially. Buying a small amount of a "mega-cap" such as GE on a set schedule over a long period of time is one such strategy.This particular one exposes you to company-specific risk. In other words, GE does have it's ups and downs. A better strategy would be to apply the same one except to rotate among 5-10 different stocks. Say you buy GE in January, BRK in February, MMM (3M) in March, etc., etc.It is the same thing that happens if you buy a single bond issuance, except that typically bond returns do not vary all that much...you are paying for an almost fixed return in the end after all. So if you buy a half dozen different issuances from different companies or agencies, even if one defaults, the others will compensate.This strategy pretty much eliminates company-specific risk. A simpler strategy which may or may not have lower costs (depends on how low you can make your costs go) is to buy an index fund or an index ETF such as SPY and let someone else do all the trading to average everything out for you. This obscures the "stock-only" strategy because there's now a mutual fund company doing the actual individual buying and selling but saves you the trouble of keeping track.If you apply some amount of investment strategy to pick when companies are performing poorly, even if it's just to look at long term performance (over say 20-30 years), you can capture some additional returns by catching stocks when they are available at bargain prices. The BMW method is one such approach. Mike Klein's chart is a graphical interpretation of that method. You simply chart what a stock's average return is over a very long period of time (20-30 years) and buy the stock when it is well below it's average return, selling when it is at or above the average return.By the way, I consider "low PE" to be less than about 10-15, not 15-23. Historically over the past 20-30 years stocks have been averaging about 20-25 for PE's (depending on who crunches the numbers). When stocks dip well below this (with a safety margin), say under 15, two things are at work. The company is obviously performing poorly. The question is whether the market is punishing them for a short term problem or if it's a long term issue.For example, Sun Hydraulics is a small cap but has been producing market beating results for a few years, netting them a PE in the high 20's. They returns have been fairly steady as well. I bailed out when the PE got just plain too high and shortly after, the stock price dropped quite a bit. More recently in the past few days, the stock failed to meet analyst expectations by about 5-6%. The stock price dropped substantially again. The reason for the drop was a one-time tax hit of about 5%. So right now the company is a high profit company (a typical growth stock) and the stock is cheap. I'm getting back in for another ride from undervalued to overvalued territory.
By the way, I consider "low PE" to be less than about 10-15, not 15-23. Historically over the past 20-30 years stocks have been averaging about 20-25 for PE's (depending on who crunches the numbers).I still don't understand where you get this number. Could you provide a reference? One major issue is that the last 20-30 years include a very large bull market that persistently showed very high P/E ratios that will skew this number upwards. I think you have to take a much longer historical view to include at least the bear market preceding it to get a useful average.The long term P/E averages I've seen are generally calculated to be around 14. There are many sources for this. In this article, see the first chart, from Vitaliy Katsenelson, who is pretty good at getting his facts right: http://www.frontlinethoughts.com/printarticle.asp?id=mwo0808... (you may need to type in your email but can use a fake one if you want). He gets 15.2.Low P/E, historically, is around 8-10 for an entire average like S&P 500. In 1982 the S&P 500's P/E was 7.3. For an excellent source on this, see Crestmont Research's outstanding Stock Market Matrix (http://www.crestmontresearch.com/content/Matrix%20Options.ht... ). Crestmont gets a long-term P/E average of around 14.That's a big difference vs. 20-25. And especially it implies that P/Es in the low teens range (13, 14, ...) are *average*, not low. Caveat emptor.-Mike
I'm not convinced that historical PE ratios have much meaning.When interest rates are high, investors can easily get nice returns on bonds and fixed income investments. Why invest in stocks? Similarly businesses have that more trouble borrowing funds to grow their businesses, finance expansions, etc, etc. It follows that earnings will be constrained. Growth rates are lower. Stocks are less attractive. So PE ratio will be lower.We saw record low interest rates under Alan Greenspan, the lowest since the 1930s they say. We saw record high interest rates (CDs paying 18 to 20%) under Paul Volker during the Carter Administration in 1979. Comparing PEs between these two periods is apples and oranges. Averaging one into the other can give numbers all over the map.Interest rates tend to track inflation. We have had very low inflation, but for now oil prices seem likely to be high for the forseeable future. As soon as the economy stabilizes, interest rates are going to rise.People have been worrying about high PEs compared to historical values for quite a while. But making decisions based on that is so complicated. You may as well buy gold and put it under your mattress. (I wouldn't stock up on paper money but grocery store commodities that keep might be a great tax free investment.)
I still don't understand where you get this number. Could you provide a reference? One major issue is that the last 20-30 years include a very large bull market that persistently showed very high P/E ratios that will skew this number upwards. I think you have to take a much longer historical view to include at least the bear market preceding it to get a useful average.I quote a number in the teens, then I'm going to get beat up for throwing out a low ball by those who's expectations are much more myopic. If I use a number closer to the last couple decades, I get responses like yours. This may not be the best source but here's a graphical chart of it over the last century:http://bespokeinvest.typepad.com/bespoke/2007/09/historical-...Clearly, anyone drawing a line through that data and calling it an average is probably grasping at straws. If we went back 10 years, you could claim that there's been a historical increase in PE's starting sometime in the early 70's. If you look at that chart and what's going on currently, you could just as easily claim that we are now in the midst of reverting to the mean. Either way, "average" PE is clearly a subjective term, no matter how many people have researched the topic.If you are a believer in Kondratiev cycle theory, then the last big peak corresponds roughly to the length of the last cycle. So we are now in a "trough" of sorts where your teens quotes are accurate and we can expect this to last about 10 years before things start rising again based on the new economic driver because Kondratiev cycles are roughly 55 years in length. K-wave theory suggests that using anything more than an average (or even fitting the data to a logarithmic curve) is invalid at best because the historical pattern is cycle, not stagnant. Again, probably not the best but here's a jumping off point:http://en.wikipedia.org/wiki/Kondratiev_waveIf you're an "efficiency" or paradigm theorist, then we've been riding on the last sigmoidal curve of the last major paradigm shift. We're going to fumble around in the dark here for a little while before the next paradigm shift happens. The reason for the increase over time in historical PE's is simply that we've gradually improved socially or technologically, or both. Once the next paradigm shift kicks in, we can expect another boost and yet another "long term bull market" as you put it. However, it may be a while before that happens.In addition, there is also evidence that the average PE for various sectors is significantly different. More than one economist has tried to claim that the correlation is based on the degree of market inefficiency due to barriers to entry such as government regulation...pharmaceuticals tend to have very high PE's while construction tends to be very low. If you subscribe to this theory, then any discussion about "average" PE's is almost meaningless. Sorry that I can't give you an exact reference on this one but it's late and my notes are buried in paper documents that wouldn't do any good to reference.Anyways, my original point was to buy stocks with historically depressed PE's and sell stocks with historically over-average PE's. Here, I've got two choices. I can either bias the results based on very long term averages or try to adjust to some sort of approximation current market conditions (current conditions being those of the last decade or so, or based on averages for an industry). I choose to use an average of the last 10 years or so simply because I may not subscribe to K-wave theory or some such directly, I do believe that there are naturally occurring economic cycles and that trying to claim a flat, stagnant average for a metric like historical market average PE is foolhardy at best, and that I might tend to slightly bias my results towards industries with higher barriers to entry but that's OK since I am specifically seeking out those industries anyways as part of my other financial checks on a particular candidate to buy.
I quote a number in the teens, then I'm going to get beat up for throwing out a low ball by those who's expectations are much more myopic.Hey, don't worry about who's going to think what about your numbers... if they're right, they're right, and anybody who takes issues with you is wrong. Problem in this case for me is, except for within boom cycles, I have *never* heard of anyone quoting average P/Es in the 20-25 range.The Bespoke chart is useful to look at. One way to interpret it is that until about 1998, we had a century of pretty consistent P/E ranges bottom at about 7 and topping out at about 22. Only in 1998 and beyond have we gone above 22, so it really begs the question why that happened, and whether it's a fundamental shift or not, which leads to this topic:The reason for the increase over time in historical PE's is simply that we've gradually improved socially or technologically, or both. Once the next paradigm shift kicks in, we can expect another boost and yet another "long term bull market" as you put it. However, it may be a while before that happens.I personally don't believe "this time it's different" because in any competitive market system, excess valuation (high P/E) is punished due to its inherent risk compared to alternatives, in other words, bubbles popping. We've had a number of those bubbles pop in the last 8 years, and I sure hope we are learning our lessons.I guess, then, that this implicit assumption -- that you believe there is an inherent structural change that explains and supports higher P/Es going forward -- should be made clear when you quote average P/E numbers like 20-25, IMHO. Maybe you will be proven right, but by no means is that a generally accepted idea today.If you haven't looked at Crestmont Research's Stock Matrix, it's really quite fascinating. Easterling's book covers the same material in maybe a more easily understandable manner with good explanations. There is likely a danger in assuming permanently high valuations, as many people did with houses, structured investment vehicles, tech stocks of companies with no profits, etc. Eventually, in comparison against other alternatives and when sanity is regained in assessing risk, high valuations falter. By how much is not clear. But stocks at average P/Es of 20-25 are, IMHO, firmly in the camp of overvalued.That is not to say that an individual stock with a P/E of 25 is overvalued, and you gave some good reasons why. High P/Es are justified if the earnings can be predicted with high certainty to grow at modest rates for a very long time. Coke is a great example, in fact most of Buffett's purchases fall under this category. Those high multipliers come from plugging long growth periods into a DCF or some similar valuation approach. Long periods of compounding growth create tremendous value and high multipliers are justified. But that is the exceptional company, as you noted. Most companies will not pass the sniff test on that kind of valuation.I would also note on the Bespoke chart, and even more clearly on Crestmont Research's matrix, that once a high average P/E is reached and it starts to fall, it usually falls a long long way, down into the 10 range, through some combination of falling share prices and increasing earnings over time. We are almost certainly in such a period now. It probably has 5-10 more years to play out.-Mike
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |