Stock Buybacks and the Equity Premium
Update 1 pulled up here (Dean Baker says current dividends plus share buybacks are unsustainable)
Update 2 Nick Bunker considers the dramatic example of Apple and the issue of unsustainable payouts to shareholders.
The point of this post is to try to estimate the sustainable long run expected real return on US common stock. Historically, the long run average return has been roughly 7% per year vastly greater than the safe real interest rate. This means that over all historical 30 year periods, US common stock has yielded a higher return than bonds. However, price earnings ratios are much higher than they were for much of the 20th century, so it doesn’t seem plausible that common stock could pay 7% real per year over long periods starting now.
Dean Baker challenged economists to come up with a story for how they could forecast 7% real returns (I can’t find the link). OK so I will.
First I will assume that the ratio of market capitalization to GDP won’t have a long term trend. US Real GDP growth has averaged about 3% per year since 973, so I assume that the value of the stock market will grow about 3% per year. Currently the S&P 500 dividend yield is about 1.9%. I guess (but don’t find the numbers) that the overall market dividend yield is higher, so I use 1.9%. So far I have 4.9% real return, well below 7%.
However, holding the market also requires buying newly issued shares and selling shares when corporations buy them back. In the USA corporations buy more shares from households than they sell — share buybacks are greater than public offerings. Net selling to the corporate sector is an additional source of returns. Share buybacks have recently been massive involving more dollars than were paid in dividends.
stock buybacks—more than $6.9 trillion of them since 2004, according to data compiled by Mustafa Erdem Sakinç of The Academic-Industry Research Network. Over the past decade, the companies that make up the S&P 500 have spent an astounding 54 percent of profits on stock buybacks. Last year alone, U.S. corporations spent about $700 billion, or roughly 4 percent of GDP, to prop up their share prices by repurchasing their own stock.
Last year’s buybacks were about 3.3% of market capitalization.
The reader may have noticed that I wrote “public offerings” not “initial public offerings”. It is possible for an existing joint stock corporation to issue new shares diluting the old shares. This is very tightly regulated and is very rare in the USA. Google keeps sending me to IPO when I ask for public offerings, so I will guess that recently all US public offerings were initial public offerings. IPO volume is much higher when share prices are high. Also 2014 had an extraordinarily high level compared to the rest of the past 10 years. That level was $ 85.2 billion less than 13% of the volume of share buybacks. I will assume that 2014 is the new normal and subtract 0.4% per year to pay for newly issued shares.
So my final back of the envelope guess of likely long term real returns is 7.8% per year*.
It seems to me that the equity premium is alive and well.
*arithmetic corrected.
update: Dean Baker responds. I like his post except that the title includes “Robert (not Paul) Waldman” which makes me think of “Robert (not Paul) Samuelson” and I sure hope Robert Waldmann isn’t to Paul Waldman as Robert Samuelson is to Paul Samuelson.
He pointed out that stock buybacks plus dividends of 5.2% of market capitalization are more than 100 % of profits (currently almost exactly 5% of market capitalization). To invest in physical capital (as they are) corporations have to be spending their financial assets or borrowing. This isn’t sustainable.
I am convinced. With a price earnings ratio of 20, there can’t be a dividend plus share buyback yield of 5.2% *and* growth of 3% without increased leverage and leverage can’t increase forever.
My tired front of the envelope guess is that with corporate debt equal to book equity (about right) both real fixed capital (K) and real debt growing at 3% per year, and no trends in relative prices, the ratio of reinvested profits to fixed capital has to be 1.5%. Here the calculation is increased real debt for a fixed debt/K ratio pays for half of the growth of K.
Market capitalization is somewhat greater than but roughly equal to nonfinancial assets so I think balanced growth of 3% per year requires a ratio of reinvested profits to market capitalization of somewhat less than 1.5%. So with a earnings/price ratio of 5% profits available for buybacks and dividends are about 3.5% and the sustainable return for everything balanced is about 6.5%.
For growth = 2% not 3% the sustainable return would be about 6%.
Interesting article on Mahablog on stock buy backs. In an article by Nick Hanauer:
Interesting? Mahablog asks where did the money go?
What used to ne a $20 – $40 share now goes for well over $100 – $200 per share. Free markets in the form of capital gains in place of trickle down economy.
Dean called it the No Economist Left Behind Challenge:
http://www.cepr.net/documents/publications/ss_economist_test.htm
Also of interest here is the paper he authored with a couple other guys nobody ever heard of (well THEN) which is popularly called ‘BDK’.
For Baker-Delong-Krugman That paper can be found here under the title “Asset Returns and Economic Growth”
link and commentary:http://delong.typepad.com/sdj/2005/03/asset_returns_a.html
PDF of paper: http://delong.typepad.com/files/bdk-bpea_20050629.pdf
The paper rapidly departs from English into Algebra and is not for the faint of heart (i.e. me). But 50 pages of presumedly good stuff for those who have the chops.
It is also worth noting that NELB (No Economist Left Behind) was explicitly issued to challenge the idea of private accounts for Social Security. The idea wasn’t to prove that NOBODY could get traditional returns on equity given the economic conditions of the time, but that it would be impossible for EVERYBODY to do it at the same time. Which would be a requirement it a private/personal account largely based in equities was to be set up for every worker in America.
I don’t know whether or how this extra condition would effect your calculations but it was I think fully implicit (and perhaps explicit – I would have to re-read those portions of BDK that I actually understood)
Clearly if people choose the their privatized personal portfolio many will get returns lower than the market return. Given management fees, I’d guess a substantial majority would get lower returns.
My interest is not personal and partially privatized porfolios. Rather, I think the US Federal Government should buy 10% of the shares (of all listed firms). So this is partial nationalization of the corporate sector not partial privatization of social security.
It is absolutely key that I am talking about an infinitely lived deeper than deep pocketed agent. Returns on stock from when I invest until when I need the money can be risky (I personally have never ever invested in stock).
So I agree with Baker that partially personal privatization of social security is a terrible proposal. But I am interested in the equity premium for a very different reason.
Run75441 the post was a reaction to something I read at Dailykos.com. In fact it was a cut and paste (with a link and proper citation) of the mahablog post.
Fair enough. On your own terms at least. But you led with this:
“Dean Baker challenged economists to come up with a story for how they could forecast 7% real returns (I can’t find the link). OK so I will.”
I am just pointing out that you are not actually responding to Dean and still less undercutting the argument underlying NELB and BDK. Neither he nor he and them were addressing anything like a “infinitely lived deeper than deep pocketed agent”.
There is a joke about the Wright Brother’s at Kittyhawk that has the punchline: “Well yeah, if you do it THAT way” in response to the plane taking off the ground. I have somewhat the same reaction to this implied rebuttal of NELB.
Robert,
I remember reading from a Morgan Stanley analyst who said that historical stock returns were all inflated by what he called “survivor bias”. That is firms move into and out of indexes, so unsuccessful firms with their losses disappear from the record and are replaced by new firms on the way up. I’m not sure how immune from this effect your statistics are.
I’m not sure of this, but would stock buy back instead of dividend distribution of some portion of a corporation’s profits result in higher CEO,
et al, compensation given that such compensation is often tied to stock price performance? Or, do such compensation bonus contracts take into account changes in stock price resulting from reductions of shares outstanding?
The computation of long term equity returns also has to take into account corporate mortality. If you own a piece of the market through a diversified portfolio or an index fund, some fraction of your holdings will croak from time to time. In his paper “The 2% Dilution”
http://www.efficientfrontier.com/ef/702/2percent.htm
William Bernstein calculated this drag at 2%. He also notes that this drag and the emergence of new issues accounts for the variance between the changes in value of the total stock market (including new issues) and the performance of a market-tracking index fund. As new issues, Google, Facebook, etc., emerge existing holdings need to be reduced to provide the funds necessary to maintain a market-tracking position.
Re: Reasons comment. Assuming the investment vehicle is an index fund, then survivor bias should not exist. Changes in the composition of the index are reflected in the the index fund in a short time. Usually the bad news issues are reflected in the prices before the component is removed. So if you use an index fund or invest like one (if one is rich enough), the survivor bias should wash out.
Stock buyback tends to be preferable from a management perspective for a few reasons – (1) dividend cuts are taken really poorly by market participants, whereas turning off repurchase isn’t taken as negatively, (2) it’s preferable from a tax perspective for shareholders – giving them more options to manage their taxes, (3) while meaningless from a pure academic perspective, repurchases accelerate EPS growth and dividends don’t, (4) whether they’re good at it or not, management teams like to be opportunistic about taking advantage of their perception of value in the stock and repurchasing when they believe their stock is below intrinsic value, (5) shareholders like to know there’s a persistent buyer (the company itself) in the market.
Stock buybacks are how capital invests.
M.Jed thanks. I naively thought that taxes were the only issue.
@Reason that’s not a problem for me, because I assume that market capitalization grows proportional to GDP. So I didn’t use the change in stock market indices at all. I think it is very hard to tell anything useful about the long term using the indices because I think there are longish waves of over and under pricing which last about 5 years to a decade.
Plotting S&P data (from Shiller, http://www.econ.yale.edu/~shiller/data.htm) is interesting. from 11969 to 1994 it increases regularly somewhat more slowly than GDP. After that it becomes wildly erratic, possibly growing faster than GDP, possibly just showing we are currently riding an asset bubble.
I will address M.Jed’s list of reasons for stock buybacks one by one.
1. “Dividend cuts are taken poorly.” This implies that investors are irrational actors who are influenced more by appearances than substance. It claims that managers attempt to conceal the fact of earnings volatility from investors by hiding them as buybacks rather than more visible dividends. This implies that they think investors are too stupid to look at earnings.
2. “Taxes.” This ignores that fact that 70% of shareholders do not pay dividend taxes — IRAs, 401(k)s, pension funds, endowments, foreigners — which means that managers are minimizing dividends for the advantage of a small minority of high-income shareholders, a class that not coincidentally the managers themselves belong to. Note also that managers with stock options do not collect dividends but do benefit from a higher stock price.
3. “Buybacks are meaningless financially.” This presumably refers to the Modigliani-Miller hypothesis which says that investors are indifferent to dividends vs buybacks. This is because dividends and buybacks do not change the value of an investment. A buyback is simply an exchange of one asset, cash, for another asset, stock, of exactly the same value. This assertion of indifference is in direct conflict with point number 1 which claims that investors prefer buybacks. At least one of these two points, if not both, must be false.
The more likely explanation is that buybacks are not meaningless to managers because bonuses are rewarded on the basis of earnings per share and by manipulating this metric with buybacks, they increase their own compensation at no gain to shareholders.
4. “Managers believe their current stock price is below fair market value.” Yet there is no evidence that corporate managers are any more skilled in market timing than any other investment manager, which is to say, not at all. They are simply making a speculative, all-in bet on a single stock after evaluating the entire universe of possible investments and finding nothing else of value, rather than returning excess earnings to shareholders for more efficient re-investment.
5. “Shareholders like to know there is a persistent buyer in the market.” This means that managers are artificially propping up the price of shares which also implies that they are paying more than fair market value for their buybacks which reduces shareholder investment return.
None of these items seem to justify a preference for stock buybacks over dividends except for manager self-dealing. If managers have no good ideas for reinvestment of earnings, they should give them to shareholders to reinvest more efficiently.
Bill all excellent points but are swamped in importance by your parenthetical/’note’:
” Note also that managers with stock options do not collect dividends but do benefit from a higher stock price.”
Boy Howdy! If there is one thing clear about modern corporate governance is that most Directors and certainly most ‘independent’ Directors consider themselves fiduciary agents to Management which all too often is lead by a joint Chairman/CEO. And that CEO/Chairman sees himself as a fiduciary agent to the principal he (and it is almost always he) sees in the mirror each morning.
At least for large cap public corporations (and not still led by actual stock holding founders).
not going to bother with a one by one counter of BillB, just this:
“If managers have no good ideas for reinvestment of earnings, they should give them to shareholders to reinvest more efficiently.”
buybacks and dividends are each ways of returning capital to shareholders so that the shareholders can invest more efficiently.
M.jed said: “buybacks and dividends are each ways of returning capital to shareholders so that the shareholders can invest more efficiently.”
That is only true if buybacks occur at fair market value. This is almost guaranteed not to be the case. The mere announcement of buybacks tends to push up prices so that buybacks occur at a premium.
Further, managers have incentives to buy back shares with timing to improve quarterly reports and boost their bonuses, not necessarily at lowest prices. Pushing up prices improves their stock options.
As you yourself stated, “shareholders like to know there is a persistent buyer in the market” which props up prices, guaranteeing dividends are squandered on over-priced shares.
People make the mistake of thinking that dividends and buybacks are exactly equivalent. As with anything a corporation invests money in, prices matter.
If the objective is to disgorge excess capital to shareholders, whether the stock price is under or over-valued relative to “fair market value” at the time of executing the repurchase is not relevant as long as the repurchase is being done in a fixed dollar amount rather than fixed share quantity.
The mere announcement of buybacks tends to push up prices so that buybacks occur at a premium.
This implies that investors are irrational actors who are influenced more by appearances than substance.
I’d also note that dividend increases or initiations also tend to push up prices, which is why a change in dividend is considered material information.
Dividends constrain management teams in a way that repurchases don’t. I concede that some shareholders will prefer that the management teams of the stocks they own are constrained.
If you believe this: [As you yourself stated, “shareholders like to know there is a persistent buyer in the market” which props up prices], then companies have succeeded in achieving a structurally higher valuation from shareholders, so who’s to say shares are “over-priced”
This sounds like an excellent game. The companies disinvest by returning capital to the shareholders. The shareholders now have more cash to invest, but even fewer shares to buy. What better formula for driving up stock prices.
Surely, this can’t be a perpetual motion machine. Something has to be getting used up. Is the labor share of the economy? The declining labor share of the economy has been helping by keep earnings high. Is it the number of shares? Something has to happen when each company has only a handful of shares outstanding. Is it actual productive capital? A company can lose money and pay dividends for years by extending its cost recovery periods. Is it TFP? TFP is what is left when you take out labor and capital.
Of course, I am probably wrong. The sheer lack of investment opportunities has made the stock market an excellent game for the last 30 years. It would be great if I could ride it out for another 30 years. Apres moi, le deluge.
http://www.multpl.com/s-p-500-dividend-yield/
Dividend Yield, 1881-2015
(Standard and Poors Composite Stock Index)
February 13, 2015 Div Yield ( 1.88)
Annual Mean ( 4.41)
Annual Median ( 4.35)
— Robert Shiller
http://www.multpl.com/shiller-pe/
Ten Year Cyclically Adjusted Price Earnings Ratio, 1881-2015
(Standard and Poors Composite Stock Index)
February 13, 2015 PE Ratio ( 27.54)
Annual Mean ( 16.58)
Annual Median ( 15.95)
— Robert Shiller
http://www.econ.yale.edu/~shiller/data.htm
January 15, 2015
Ten Year Mean Price Earnings Ratio, 1900-2015
(Standard and Poors Composite Stock Index)
1900 ( 18.3)
1901 ( 20.5) Roosevelt
1902 ( 22.2)
1903 ( 19.9)
1904 ( 15.6)
1905 ( 18.1)
1906 ( 19.7)
1907 ( 16.8)
1908 ( 11.7)
1909 ( 14.7) Taft
1910 ( 14.5)
1911 ( 14.0)
1912 ( 13.8)
1913 ( 13.2) Wilson
1914 ( 11.6)
1915 ( 10.4)
1916 ( 12.5)
1917 ( 10.7)
1918 ( 6.6)
1919 ( 6.2)
1920 ( 6.1)
1921 ( 5.3) Harding
1922 ( 6.5)
1923 ( 8.4) Coolidge
1924 ( 8.2)
1925 ( 9.8)
1926 ( 11.6)
1927 ( 13.9)
1928 ( 19.3)
1929 ( 26.7) Hoover
1930 ( 21.5)
1931 ( 16.5)
1932 ( 9.1)
1933 ( 8.8) Roosevelt
1934 ( 13.3)
1935 ( 11.6)
1936 ( 17.5)
1937 ( 21.8)
1938 ( 13.6)
1939 ( 16.2)
1940 ( 16.9)
1941 ( 14.0)
1942 ( 10.0)
1943 ( 10.0)
1944 ( 11.0)
1945 ( 12.0) Truman
1946 ( 15.7)
1947 ( 11.6)
1948 ( 10.5)
1949 ( 9.9)
1950 ( 10.6)
1951 ( 11.6)
1952 ( 12.4)
1953 ( 12.9) Eisenhower
1954 ( 11.8)
1955 ( 15.6)
1956 ( 17.6)
1957 ( 16.3)
1958 ( 13.5)
1959 ( 18.0)
1960 ( 18.3)
1961 ( 18.5) Kennedy
1962 ( 21.0)
1963 ( 19.0) Johnson
1964 ( 21.3)
1965 ( 22.9)
1966 ( 23.8)
1967 ( 20.1)
1968 ( 21.2)
1969 ( 20.8) Nixon
1970 ( 16.9)
1971 ( 16.4)
1972 ( 17.1)
1973 ( 18.6)
1974 ( 13.4) Ford
1975 ( 8.9)
1976 ( 11.3)
1977 ( 11.5) Carter
1978 ( 9.2)
1979 ( 9.2)
1980 ( 8.8)
1981 ( 8.5) Reagan
1982 ( 7.4)
1983 ( 9.6)
1984 ( 9.4)
1985 ( 10.7)
1986 ( 13.4)
1987 ( 16.0)
1988 ( 14.4)
1989 ( 16.6) Bush
1990 ( 16.5)
1991 ( 17.9)
1992 ( 19.5)
1993 ( 20.8) Clinton
1994 ( 20.5)
1995 ( 22.7)
1996 ( 25.9)
1997 ( 31.0)
1998 ( 36.0)
1999 ( 42.1)
2000 ( 41.7)
2001 ( 32.1) Bush
2002 ( 25.9)
2003 ( 24.1)
2004 ( 26.4)
2005 ( 26.0)
2006 ( 26.0)
2007 ( 26.8)
2008 ( 20.8)
2009 ( 16.9) Obama
2010 ( 20.7)
2011 ( 21.8)
2012 ( 21.4)
2013 ( 23.2)
2014 ( 25.5)
January
2015 ( 26.7)
— Robert Shiller