Comments
"If future market level is high, like the red line, the expected return from this point is high." If you look at the peaks and valleys of your second chart, The Ratio of Total Market Cap to US GDP, from the high of 150% to the 1980 price as well as the 2009 price, wouldn't the green line have the higher
expected return? I'm still struggling with that.
I have a little trouble understanding the formula. What is the expected return of TMC/GDP=40 over the next 8 years as well as TMC/GDP =120 over the next 8 years? As I understand it the lower market cap to GDP would produce higher returns. Thanks for the timely replies.
Red, brown, and green lines are not actual returns. They are expected returns if future market is high, median, and low, respectively. If future market level is high, like the red line, the expected return from this point is high.
See this part of the equation in the the text: (Re/Rb)(1/T)-1
to add to my question, why does TMC/GDP, of 40 , the undervalued line "generates low returns" as you said?
on that chart why does the overvalued line, the red line, perform so well in the next eight years. I know it doesn't on the rest of the page but the chart confuses the data with me. Why does the red line perform 30%?
The 40%, 80%, or 120% lines in the third chart indicate where the return would be if the market-cap/GDP ratio is 40%, 80%, or 120% 8 years from now. A lower future ratio generates low returns.
I may be reading the chart wrong, but the last chart on this page the "Predicted and actual stock market returns" with the TMC/GDP ratio seems a little off to me. With the ratio at 40%, the green line, the overall returns are negative. With the ratio of 120 and the market of overvalued and the red line the returns went as high as 30% and finished at 12%. I really like this ratio as well as the fact that Warren Buffet uses this as Top analysis. Can you see if I'm reading the chart wrong? Thanks.
Great website as well!
Wilshire 5000 total market capitalization is wrong as of 3/2/12. Correct value is $14,192.82
Daisy42,
We can make these data available for download. But like any other downloads, it will be for Premium Members only.
Thanks!
GuruFocus.
I would like to see month-by-month market cap to GDP figures going back to 1971 as shown in the graph, because I would like to calculate the percentile values of each month (or quarter) on a rolling basis. You can see the individual values by mousing over the chart, but copying these down and inputting them to a spreadsheet is laborious and not particularly reliable. I have been able to do this with Shiller's PE10 data. I think it would give a more finely graded asset allocation suggestion than those derived from th broad bands you give. Are you able to / prepared to make the data avilable, or point me to where I can download the historic data myself?
Thanks.
My mistake, the comment is on the top of the page rather than the bottom.
Jhodges72,
We are aware that
Warren Buffett article had a different total market cap from Wilshire 5000; we referred that article many times in the text. We said "not available" meaning that it is not available to us.
We do have our limitations, and we make mistakes. The good thing is that we are always learning and getting better.
Thank you for pointing it out (again).
I'm well aware of what the ratio represents as I've had in depth discussions via email regarding it with Carol Loomis, who was the first person to write a national article concerning it and is one of
Warren Buffetts' closest friends. Its purpose is a bit more detailed than you've described. It is to measure, in actual dollars, how much value Is being created in the U.S. for a given year (GNP) and comparing that information with how much speculated market value the public is pricing a "portion" of the businesses that are represented in the overall GNP number. If the total "public market" is being valued equal to the GNP (actual total market value) then we know the public markets are overvalued because the public market is just a portion that is accounted for in the GNP. the private market is quite relative to the public market but is not accounted for, obviously, in the public markets quotation. All the historical data required to accurately accomplish the "proper" measurement is available. Not for "free" it isn't and is maybe why you're under the impression that it isn't available at all, but it's certainly available. The
Warren Buffett article that was written by Carol Loomis used the "entire" historical market, not the Wilshire 5000. Although using it is better than having nothing to go on. I'm just surprised that on a partially paid for and partially free site (GuruFocus) that you didn't know that type of historical data is available. I'm actually blown away by that. The only limitations you are experiencing is do to your own accord, not because the information isn't available.
Also we are aware that Wilshire 5000 does not cover all markets. But the index that cover all markets and have long enough history just does not exist. We do have limitations with data.
On the other hand, what is important here is the current value of the ratio relative to its historical values. The assumption here is that Wilshire 5000 changes in the same way as the whole market. Although it might not be exactly the case, but it is quite close.
Again please understand we are limited by the availability of data.
Jhodges72,
We are aware that GDP and GNP are different. We explained in the text why we use GDP.
First of all, historically they are always within 2% of each other. Also GDP data updates much earlier than GNP from BEA website, by using GDP we can update the information earlier.
You use GDP & GNP in the same statement almost as if they were synonymous. They are not. Also, if you are using the Wiltshire to compare, you do know that you're several billion dollars of capital off because the Wiltshire doesn't account for stocks traded on the OTC markets. That's a huge market place. Nor does it account for private investments, also a huge market place.
Instead of just Market Cap, why aren't you considering 'enterprise value' of the S&P 500. Isn't that more realistic??
We will add international valuations soon. No. We don't have Tobin Q chart now, but we will look into it.
Thanks!
GuruFocus.
Do you have a Tobin Q chart
Please add international markets to this site. Please!
On market valuation. Long term analysis with adjustment. 10.2.11
[
seekingalpha.com]
On market valuation. An interesting long term PE adjsutment & analysis.
[
seekingalpha.com]
Why does your analysis not take into consideration interest rates. In determining whether the market is undervalued, shouldn't the TMC/GDP ratio be influenced by interest rates?
there was a bug from the Data source. It has been fixed.
thanks!
In your DCF calculator, the benchmark is 11%, i.e., for annual return of S&P. The annual market return, as stated above, is 5.7%. Would it be correct then, to change the benchmark to 5.7% vs 10 or 11%?
In March 2009 you could actually expect 12% a year.
Question. Using your methodology it looks like the highest starting expected return over the past ten years would have been around 8% annually if you look at the 3/1/2009 figures. Is that accurate or am I missing something in the calculation?
Thanks!
very helpful, glad to see we're getting closer to normal valuation. I understand that is easier to use GDP but if we look at GNP for q1 2011 it is 15.094 trillion. If the total market is 12.617 trillion then the ratio would be about 83.6 (versus 86.4 using GDP numbers)
Total Market Value (TMC) per Wilshire on 07/20/2011 : 14091
GDP as of the 1st quarter 2011 : 13,444.3
GNP as of the 1st quarter 2011 : 13,655.8
TMC/GDP = 104.81
TMC/GNP = 103.19
Either way (GDP or GNP), market seems to be modestly overvalued.
Advance GDP and GNP as of the 2nd quarter 2011 will be released only towards the end of July. Gurufocus, please correct me if I'm wrong.
Valuable metrics...thanks gurufocus for making it available. I monitor this page frequently and the ratio seems to have dramatically reduced. Used to be 'modestly overvalued' not long ago (over 95%), now its 'fairly valued (less than 85%). Has the change in GDP and/or Total Stk Mkt been this dramatic to cause over 10% change? I find it puzzling indeed.
Valuable metrics...thanks gurufocus for making it available. I monitor this page frequently and the ratio seems to have dramatically reduced. Used to be 'modestly undervalued' not long ago (over 95%), now its 'fairly valued (less than 85%). Has the change in GDP and/or Total Stk Mkt been this dramatic to cause over 10% change? I find it puzzling indeed.
Very insightful question. The answer is complex but I will attempt to give a simple one. If government lead the GDP expansion, then it's bound to result in lower market multiple vs. Private cpital led expansion.
Does anyone know how to find out what this ratio is for different countries around the world?
I don't understand how earning yield could be 7% for sp500 and yet we expect net 4% returns... I wonder now with more global nature of companies this metric is not really effective.
How does the (changing) mix of industries and government spending that contribute to GDP affect this ratio? If government spending or non-productive consumerism make up an even larger slice of GDP, does that argue for a ratio that should be (rationally) lower than the long-term norm?
How does the (changing) mix of industries and government
No, it includes inflation.
Is the expected average annual expected return of 3.8% calculated net of inflation?
Everything we use here is from GDP. Therefore, GDP growth is the business growth. We do allow dividend distribution. It is a part of investment return.
Hussman uses PPE reverse to the mean, we use TMC/GDP reverse to the mean. In our calculation, profit margin does not play a role, therefore we don't need to do what Hussman and Shiller did to flat the profit margins.
We don't have data for earlier times.
Yes, intersting to compare this vs Hussman.
Curious to know how you guess at "business growth"; I think Husmanviews this along the same lines as long-run real earnigns growth. Implicitly you seem to allow the reinvestment of earnings not distributed as dividends, so this will give a higher rate than he uses. Is that right?
On the other hand he effectiviely annuitizes the change in the valuation ratio (P/Normalized E) to some long run average (which is why I think he now forecasts barely any return-after inflation). But what do you do. Please clarify.
Thanks for your attention
PS. Do you have data before the 70's?
Sorry, but you can not sell me nonsense, claiming to be fresh out of sense. A flaw in your argument is that the future net cash flows are difficult to predict accurately ALSO interest rates are high or low. The definition of intrinsic value is mathematically correct, but not computable. All we can do is our best estimate is to calculate the present value. Unless you think there will be a negative correlation between interest rates and the level of corporate profits, your argument is invalid. Historically, no such inverse correlation. In fact, I read (forgot the source) that profits and interest rates tend to be directly correlated, as long as interest rates are not at extreme levels or lower.
[
www.whatisguide.net]
I would be interested in anyone who could tell me about the Case Schiller p/e index and how it relates to fair value for the market. Particularly, I would like to know how dividends are calculated into his formula. I noticed that dividends at many stretches of the history of the index averaged between 4 to 7 percent with a 10 year trailing p/e of 15. Today the Case index is at roughly 22 x earnings with a paltry yield of less than 2%.
Any info would be helpful.
Thank you for the question, Aagold. It is a great question!
To include the revenue of US companies from international, one needs to use GNP (Gross National Product) instead of GDP, as we explained in the article. But we also explained in the article "The size of the US economy is measured by Gross National Product (GNP). Although GNP is different from GDP (gross domestic product), the two numbers have always been within 1% of each other. For the purpose of calculation, GDP is used here."
Another reason we use GDP here is because GDP data comes out much earlier than GNP data, as you can imagine why.
Thanks!
GuruFocus.
Does anybody know what percent of the total US stock market (Wilshire total market) income is either earned abroad or is due to exports? I've heard a few times that the S&P 500 companies this number is about 50% (probably much less for small and midcaps). So my question is, if 50% is anywhere near accurage, doesn't that make comparing the US stock market valuation to US GDP less relevant? It seems like the US stock market valuation should be compared to a weighted average of US GDP and non-US GDP, where the weights are proportional to US-based income version non-US-based income.
I don't know how anyone can think the market will average more in the next 50 years (10%) than it did in the last 50 years (6.03%). If the Dow Jones averages just what it did since 1960, by 2060 the market will be priced over 214,000. Is this realistic? I don't know. But, it's far more realistic than thinking it will be priced above 1,300,000 as it would if it averaged 10% a year. Guys, it's simple math here.
www.thepolandreport.com
The average rate of return that you can expect long term from the stock market [long term meaning 30+ years] is 10%. 10% in one year is not much and in two or three years 10% a year will not do anything amazing as far as growing money is concerned. However after 10, 20, and even 50 years 10% a year compounded will do unbelievable things to money, even in small amounts.
[
www.financialculture.com]
This is an excellent page that I just discovered. I applaud you providing this to subscribers.
Despite the relative basis, emerging market economies, looks much better than in industrialized countries. I believe that such a fee, which is currently located in emerging markets, valuations are justified to some extent. Unless some of these tightening of monetary policy in emerging economies, I believe the current valuation level is sustainable. We must remember that most developing economies are generally much better than the developed economies expectations of earnings growth.
[
www.financemetrics.com]
guru--
this site would be a more accurate place to get TMV. See the top line. The value is updated monthly
<[
www.wilshire.com]>
guru
apparently your are using ^dwc for tmv; perhaps you should be using ^w5000
Thanks. I was confused because you actually list business growth % instead of GDP % in the formula. You may want to change this to avoid confusing other readers also.
You are using 6% as the business growth rate. In our calculation we use the 10-year average GDP growth rate, which is around 4%. That is the difference.
The market index we use is Wilshire5000.
Can someone please walk me through the expected return formula. I can't figure out how they are arriving at a current 6.2% expected annual return for the coming 8 years. From the above description, here is what I understand the formula should currently be;
Investment Return % = 1.9% + 6% + (((80/80.5)^1/8) - 1)
This ends up with an expected return of 7.9% not the currently listed 6.2%. What am I missing or doing wrong?
where does he get total market index? is it dow jones u.s.? or wilshire 5000?
Hi,
Can you tell what was the TMC/GNP ratio in 2007 2008 and 2009.
All we can do is make our best estimates to actually calculate the present value. Unless you think there will be an inverse correlation between the level of interest rates and the level of corporate profits, your argument is not valid. Historically there is no such inverse correlation. In fact, I have read (forget the source) that profits and interest rates tend to be directly correlated as long as interest rates are not at extreme highs or lows.
[
www.financeandmarkets.net]
Do you decide your asset allocation based on market valuation?
Just like Buffett - He was piling up cash when markets were overvalued and as soon as market started going into nosedive he started putting cash to work.
Not sure if he was watching market valuation or he just was not finding anything worth buying at the price offered to him?
Do market valuation matter?
Ok, I actually see where the Predicted Return graph is going. It would be the returns if today's market moved to 40% or 120%. If it moved downward to 40%, it would produce negative returns from today.
"The definition of intrinsic value is mathematically precise, just not calculable." - Buffetteer17
that is what make investing so much fun, hard and challenging...
It is like working with humans... so unpredictable....as opposed to say working with computers or machines.
But great leaders know of how to maximize returns from people..and great investors know how to maximize returns from assets like stocks and bonds without taking too much risk.
Not sure if I am reading the predicted return graph correctly. The green line is for the 40% TMC/GDP, but yet it produces negative returns. Is there a possibility that the red and green lines are each mislabeled as the other? The brown line or 80% TMC/GDP looks fine.
As of 3/22/10, the expected returns from the stock market is calculated to be 5.4% per year. Something doesn't seem right.
The first two terms in the calculation are div_yield(%) + business_growth(%), which comes to 1.97% + 6% = 7.97%. The final valuation_change(%) term, if we use fair value = 80% of GDP, comes out to (80/83.9)^(1/8) - 1 = -0.6%. So the average 8-year average return comes out to 7.37%.
Am I making a mistake in this calculation?
Gurufocus,
The total market value divisor is 1059.32 (in millions) and that puts the
total market at 12.293 (in millions) and the full-cap total market at
13.453 (in millions). The percentages are then 86% and 94%. Why don't
you use the US Total Market full cap? The data can be found at
[
www.djindexes.com].
Thanks for posting and updating this information.
Buffetteer17, thanks for your answer. I look forward to seeing your investment returns. By the way, where can I see your returns? Is there a link/website? Best regards
Latc, what does my track record have to do with this argument? Nothing. That question is implicitly an ad hominem attack on me. You accuse me of attacking you but I have not, at least not intentionally. I only disagreed with your logic and called it nonsense. That's strong language, but I had thought you could tell from the context that the sentence was intended to be humorous rather than vindictive. I guess I just do not have a good sense of humor. That statement is intended as an attack of your argument, not against you. I am sorry if you took it as a personal attack.
Next you make an appeal to authority, by referring to Buffett. I am unsure of just what relationship you are talking about. Is it the fact that stocks are sometimes mis-priced? Buffett has consistently maintained that the intrinsic value of a company is the discounted present value of future "owner earnings," which, while not exactly the same as net cash flows, is pretty close. He apparently believes this strongly, as he put this statement in his Owner's Manual [http://www.berkshirehathaway.com/ownman.pdf]: "Intrinsic value is an all-important concept that offers the only logical approach to evaluating the relative attractiveness of investments and businesses. Intrinsic value can be defined simply: It is the discounted value of the cash that can be taken out of a business during its remaining life."
I will shortly be posting my usual "Quarterly Report on The Porfolio Q4 2009," which will answer your question about my investing returns. You can have the last word in this argument. I am done.
Buffetteer17, what is your track record on investments, let’s say in the last 10 years? Even
Warren Buffett has talked about the relationship that you are dismissing, so instead of attacking show me your performance track record based on your theoretical assumptions.
I'm sorry, but you cannot sell me nonsense by claiming to be fresh out of sense. The flaw in your argument is that future net cash flows are hard to predict accurately WHETHER OR NOT interest rates are high or low. The definition of intrinsic value is mathematically precise, just not calculable. All we can do is make our best estimates to actually calculate the present value. Unless you think there will be an inverse correlation between the level of interest rates and the level of corporate profits, your argument is not valid. Historically there is no such inverse correlation. In fact, I have read (forget the source) that profits and interest rates tend to be directly correlated as long as interest rates are not at extreme highs or lows.
Buffetteer17, I know the theory but theory normally does not work, I can not precisely predict net cash flows many years out, can you? Therefore, I just go one year out, in this regard, the current 10 year-Treasury yield is only 3.8% while the S&P 500 earning yield (using the expected earnings for 2010) is around 6.9%, so clearly stocks are undervalued vis a vis bonds right now. I do not have the TM expected earnings for 2010 but I am sure the result will show the same conclusion.
If you believe that current low interest rates will persist for a decade or more, your reasoning is correct. Reminder: the discounted value of net cash flows is the correct theoretical measure of the value of a stock. However, remember that this formula refers to net cash flows far in the future as well as near term. If you believe, as I do, that interest rates will rise significantly during the next decade, you must use a much higher discount rate for all but the next 2-3 years of net cash flows. This adjustment causes the value of stocks to be much less.
This stock market valuation TMC vis a vis GDP gives a partial view on the over/under/fair valuation of the stock market. The level of interest rates needs to be incorporated into the equation to provide a more complete view. So, the current low level of interest rates justifies a higher TMC/GDP ratio, hence I would argue that stocks are currently undervalued.
Alahendrix,
We use Wilshire 5000 for the measurement of total market cap, which is different from what Buffett and Van Den Berg use. They use the total of all stocks traded. We do not have the daily data for that. Their numbers are higher than Wilshire 5000.
But relatively speaking, the overall picture is the same.
GuruFocus.
You guys are calculating TMC/GDP differently than most. For i.e, Arnold Van DeBerg, whom you site, acutally has today's TMC/GDP well over 100%. He pegged it at 108% back in Oct!
So, while using the Wishire 5000 is interesting for your Market Cap input, it is somehow very off from Buffett's data points and Van DeBerg's as well. If you're going to use this metric, you should at least make sure that it is accurate. And clearly it is not!
Can anybody clarify if the TMC number has been adjusted to exclude foreign ADRs traded in the US? If this is not the case, then stocks might be undervalued.
The first graph - "Total Market Cap and US GDP" would serve better if it were in log scale. The graph gives the false impression that valuations were at record low levels in early 2009. The "The Ratio of Total Market Cap to US GDP" graph gives a much more realistic representation of the relative valuations over time.
Does anybody have a website with a similar graph for China's equity markets? In addition, anyone have a graph for the U.S. going back further than 1970? It could be interesting to check out what the valuations were like during the Ben Graham days.
The 12/10/2001 Fortune article covering W.B. on the stock market clearly states that TMC/GNP peaked at 190% in 3/2000, not the 148% stated in the GF article. The actual TMC/GNP valuation on 10/9/09 approximates 97%. WB states "... if the percentage relationship falls to 70%-80%, buying stocks is likely to work very well..."
That looks really cool. This is simple enough to understand. No wonder Buffett is so beloved.
This site just published an article to take this a step further: they backtested this metric as a timing indicator. Interesting is if you only buy at the significantly undervalued and sell at the significantly overvalued, you would have avoided the last two bubbles and made 9.7% (?) per year since 1980. See this link: [
www.validfi.com].
Sorry, how do you put an embedded link on gurufocus.com?
JC.
Any idea what the difference between Total Market Value and Total Market Value - Full Cap on the Wilshire website is? The latter is about 11% higher than the former.
[
www.wilshire.com]
Those charts cannot be copied directly. But you can print the screen to PDF file or other image file and crop the charts out.
Sorry about that.
Dumb question - but how can I copy the above graphs?
Thank you for this very useful tool and article.
Good work here. You can get the raw data here and see for yourself:
[
www.wilshire.com]
and here:
[
www.bea.gov]
For GNP (which runs approximately close to GDP), click on "Full Release and Tables" in right column, and go to Table 3. GNP appears as the second line from the bottom, "Equals: Gross National Product."
This is really helpful. In fact, I was just searching on line yesterday for this exact info and getting frustrated because it wasn't easy to find the data. Definitely makes my gurufocus.com fees worthwhile.
