Monday, September 22, 2014

Stock Buybacks: They are big, they are back and they scare some people!

This has been a big year for stock buybacks, continuing a return to a trend that started more than two decades ago and was broken only briefly by the crisis in 2008. Focusing just on the S&P 500 companies, buybacks in the 2013 amounted to $475.6 billion, not quite as substantial as the best buyback year in history (2007, with $589.1 billion), but still significantly up since 2009. As stock prices rise and anxiety about bubbles and real economic growth also come to the surface, it is not surprising that some of those looking at rising prices are trying to make a connection, rightly or wrongly, to the buyout numbers. As a general rule, even insightful stories about buybacks tend to focus on one cause or effect of the buyback phenomenon but miss the big picture. In particular, there have two news stories about buybacks, one in the Economist and one in the Wall Street Journal. Since I talked to both journalists as they wrote these stories, and I am quoted in one of them, I should disclose that I like both writers and think they did their research, but their particular perspectives (that stock buybacks can be value destructive in the Economist and that they affect liquidity in the WSJ) may be blurring the big picture of buybacks. In fact, I think that the Economist overplayed their hand by calling buybacks “corporate cocaine”, a loaded header that treats buybacks as a destructive addiction (for which the cure, as with any other addiction, is abstinence). This post is not aimed at the vast majority of investors who sensibly view buybacks as good or bad on a company-by-company basis but at the shameless boosters of buybacks, who treat it as a magic bullet, at one extreme, and the equally clueless Cassandra chorus, who view it as the market equivalent of the Ebola virus, signaling the end of Western civilization as we know it, at the other.

Laying the Groundwork: Trends and History
For much of the last century, companies were not allowed to buy back stock, except in exceptional circumstances. In the United States, companies have been allowed to buy back stock for most of their existence, but the pace of buybacks did not really start picking up until the early 1980s, which some attribute to a SEC rule (10b-18) passed in 1982, providing safe harbor (protection from certain lawsuits) for companies doing repurchases. The legal rules governing buybacks in the US today are captured nicely in this Harvard Law School summary. In the graph below, I show aggregate stock buybacks and dividends at US companies going back to 1980.
Dividends & Buybacks at all US firms (Source: Compustat)
This graph backs up the oft-told story of the shift to buybacks occurring at US companies. While dividends represented the preponderance of cash returned to investors in the early 1980s, the move towards buybacks is clear in the 1990s, and the aggregate amount in buybacks has exceeded the aggregate dividends paid over the last ten years. In 2007, the aggregate amount in buybacks was 32% higher than the dividends paid in that year. The market crisis of 2008 did result in a sharp pullback in buybacks in 2009, and while dividends also fell, they did not fall by as much. While some analysts considered this the end of the buyback era, companies clearly are showing them otherwise, as they return with a vengeance to buy backs.

As some of those who have commented on my use of the total cash yield (where I add buybacks to dividends) in my equity risk premium posts have noted (with a special thank you to Michael Green of Ice Farm Capital, who has been gently persistent on this issue), the jump in cash returned may be exaggerated in this graph, because we are not netting out stock issues made by US companies in each year. This is a reasonable point, and I have brought in the stock issuances each year, to compute a net cash return each year (dividends + buybacks - stock issues) to contrast with the gross cash return (dividends + buybacks).
Gross cash (Dividends+Buybacks) and Net Cash (Dividends+Buybacks-Stock Issues) as % of Market Cap
Note that I have converted all these numbers into yields, by dividing them by the aggregate market capitalization at the end of each year. Both the gross cash yield (5.53%) and net cash yield (3.89%) peaked in 2007, and the lowest values for these numbers were in 1999 and 2000, when the gross cash yield was 2.17% (1999) and the net cash yield was 0.67% (2000). At the end of 2013, the gross cash yield stood at 4.49% and the net cash yield at 3.16%, both slightly higher than the aggregate values of  4.24% for the gross yield and 2.46% for the net yield over the 1980-2013 time periods; the simple averages yield 4.65% for the gross yield and 2.60% for the net yield over the entire time period.

Since the aggregate values gloss over details, it is also worth noting who does the buybacks. It goes without saying that the largest buybacks (in dollar terms) are at the largest market cap companies, and the following is a list of the top fifteen companies buying back stock in 2013:
Companies buying back the most stock in 2013 (in millions)
Not only is more money being returned in the form of buybacks, but the practice of buybacks has also now spread far and wide across the corporate spectrum, with small and large companies, as well as across different sectors, partaking in the phenomenon:
Dividends and Buybacks in 2013: Data from S&P Capital IQ
Other than utilities, the shift to dividends is clear in every other sector, with technology companies leading with almost 76% of cash returned taking the form of buybacks. 

Keep it simple: Buybacks are a return of cash to stockholders
To understand buybacks, it is best to start simple. Publicly traded companies that generate excess cash often want to return that cash to stockholders and stockholders want them to do that. There are only two ways you can return cash to stockholders. One is to pay dividends, either regularly every period (quarter, semiannual or year) or as special dividends. The other is to buy back stock. From the company’s perspective, the aggregate effect is exactly the same, as cash leaves the company and goes to stockholders. There are four differences, though, between the two modes of returning cash. 
  1. Dividends are sticky, buybacks are not: With regular dividends, there is a tradition of maintaining or increasing dividends, a phenomenon referred to as sticky dividends. Thus, if you initiate or increase dividends, you are expected to continue to pay those dividends over time or face a market backlash. Stock buybacks don’t carry this legacy and companies can go from buying back billions of dollars worth of stock in one year to not buying back stock the next, without facing the same market reaction.
  2. Buybacks affect share count, dividends do not: When a company pays dividends, the share count is unaffected, but when it buys back shares, the share count decreases by the number of shares bought back. Consequently, share buybacks do alter the ownership structure of the firm, leaving those who do not sell their shares back with a larger share in a smaller company.
  3. Dividends return cash to all stockholders, buybacks only to the self-selected: When companies pay dividends, all stockholders get paid those dividends, whether they need or want the cash. Thus, it is a return of cash that all stockholders partake in, in proportion to their stockholding. In a stock buyback, only those stockholders who tender their shares back to the company get cash and the remaining stockholders get a larger proportional stake in the remaining firm. As we will see in the next section, this creates the possibility of wealth transfers from one group to the other, depending on the price paid on the buyback.
  4. Dividends and buybacks create different tax consequences: The tax laws may treat dividends and capital gains differently at the investor level. Since dividends are paid out to all stockholders, it will be treated as income in the year in which it is paid out and taxed accordingly; for instance, the US tax code treated it as ordinary income for much of the last century and it has been taxed at a dividend tax rate since 2003. A stock buyback has more subtle tax effects, since investors who tender their shares back in the buyback generally have to pay capital gains taxes on the transaction, but only if the buyback price exceeds the price they paid to acquire the shares. If the remaining shares go up in price, stockholders who do not tender their shares can defer their capital gains taxes until they do sell the shares.
Buybacks: The Value Effect
Buybacks can have no effect, a positive effect or negative effect on equity value per share, depending on where the cash from the buyback is coming from and how it affects the firm’s investment decisions. To illustrate the effects, let’s start with a simple financial balance sheet (not an accounting one), where we estimate the intrinsic values of operating assets and equity and illustrate the effects of a stock buyback on the balance sheet.

Note that the buyback can be funded entirely with cash, partly with cash and partly with new debt or even entirely with debt. (I am going to leave out the perverse but not uncommon scenario of a company that funds a buyback with a new stock issue, since the only party that is enriched by that transaction is the investment banker who manages both the issuance and the buyback). The value of the operating assets can change, if the net debt ratio of the company changes after the buyback (thus affecting the cost of capital) or if the buyback reduces the amount that the company was planning to invest in its operating assets (thus changing the cash flows, growth and risk in these assets).  This framework is a useful vehicle to look at the conditions under which buybacks have no effect on value, a positive one and a negative one.

The indifferent: For buybacks to have no effect on value, they should have no effect on the value of the operating assets. That must effectively mean that the buyback is entirely funded with cash off the balance sheet or that even if funded with debt, there is no net value effect (tax benefits cancel out with default cost) and that the buyback has no effect on how much the company invests back into its operating assets. As an example, consider the $13.2 billion in stock buybacks at Exxon Mobil in 2013. The company funded the buybacks entirely with cash surpluses and it not only had more than enough cash to cover reinvestment needs but continues to generate billions of dollars in excess cash (over and above its reinvestment needs).

The good: There are three pathways through with which a buyback can have a positive effect on value:
  1. Discounted cash holdings: There are some companies with significant cash balances, where investors do not trust the management of the company with their cash (given the track record of the company). Consequently, they discount the cash in the hands of the company, on the assumption that they will do something stupid, and this stupidity discount can be substantial. This is one of the few scenarios where a stock buyback, funded with cash, is an unalloyed plus for stockholders, since it eliminates the cash discount on the cash paid out to stockholders.  
  2. Financial leverage effect: A firm that finances a buyback with debt, increasing its debt ratio, may end up with a lower cost of capital, if the tax benefits of debt are larger than the expected bankruptcy costs of that debt. That will occur only if the firm has debt capacity to begin with, but that lower cost of capital adds to the value of the operating assets, though it can be argued that it is less value enhancement and more of a value transfer (from taxpayers to stockholders). 
  3. Poor investment choices: There is also the scenario where a firm that has been actively investing in a bad business or businesses (earning less than the cost of capital) redirects the cash towards buybacks. Here, less investment is value increasing and I will let you be the judge on how many firms on the top fifteen list in 2013 fall into that scenario. (I can think of quite a few...)
The bad: There are two ways in which a buyback can have a negative effect on value. The first is if the firm is correctly or over levered and chooses to finance the buyback with even more debt, since that would push the cost of capital higher after the buyback (as the expected bankruptcy costs overwhelm the tax benefits of debt). The second is if the firm takes cash that would have been directed to superior investment opportunities (where the return on capital > cost of capital) and uses it to buy back stock; this requires that the company also face a capital constraint, imposed either internally (because the company does not like to raise new financing) or externally (because the company is prevented from raising new financing). 

Buybacks: The Pricing Effect
If buybacks have no effect on value, can they still affect stock prices? Sure, and there are three possible factors that may cause the effect. The first is if there is a market mistake at play, where the stock is priced above or below its intrinsic value and the buyback occurs at a price that is not equal to the value. The second is that markets extrapolate from corporate actions and may view the buyback as a signal about what managers of the company think about its fair value. The third is that a buyback, especially if large and/or on a lightly traded stock, can have liquidity effects, tilting the demand side of the pricing equation. All of the effects are captured in the picture below:


  1. Market mispricing: If the stock is mispriced before the buyback, the buyback can create a value transfer between those who tender their shares back in the buyback and those who remain as stockholders, with the direction of the transfer depending on whether the shares were over or under valued to begin with. If the price is less than the value, i.e., the stock is under priced, a buyback at the prevailing price will benefit the remaining shareholders, by letting them capture the difference but at the expense of the stockholders who chose to sell their shares back at the “low price”. In fact, it is likely that the market will view the announcement of the buyback as a signal that the stock is under valued and push the price impact in what is commonly categorized as a signaling effect. If the price is greater than the value, i.e., the stock is over priced, a buyback will benefit those who sell their shares back, again at the expense of those who hold on to their shares. In either case, there is no value creation but only a value transfer, from one group of stockholders in the company to another. Lest you feel qualms of sympathy for the losing group in either scenario, remember that most stockholders get a choice (to tender or hold on to the shares) and if they make the wrong choice, they have to live with the consequences. 
  2. Signaling: For better or worse, markets read messages into actions and then translate them into price effects. Thus, when companies buy back stock, investors may consider this to be a signal that these companies view their stock to be under valued. If there is a signaling effect, you should expect to see the stock price jump on the announcement of the buyback and not the actual execution. The problem with this signaling story is that it attributes information and valuation skills to the management of the company that is buying back stock, that they do not possess. The evidence on whether companies time stock buybacks well, i.e., buy back their stock when it is cheap, is weak. While there is some evidence that companies that buy back their own stock outperform the market in the months after the buyback, there is also evidence that buybacks peak when markets are booming and lag in bear markets. 
  3. Liquidity effects: A stock buyback, especially if it is of a large percentage of the outstanding shares, does create a liquidity effect, with the buy orders from the company pushing up the stock price. For this to occur, though, the shares bought back have to be a high percentage of the shares traded (not the shares outstanding). If there is a liquidity effect, you should expect to see the stock price rise around the actual buyback (and not the announcement) but that price effect should fade in the weeks after. While the Wall Street Journal makes legitimate points about how buybacks can sometimes tilt the liquidity playing field, looking across companies that buy back stock and scaling the buyback to the daily trading volume on the stock, the median value of buybacks as a percent of annual trading volume was 0.79% and the 75th percentile across all firms is 2.17%. It is true that there are firms like IBM and Pfizer that rank among the biggest buyback firms, where buybacks are a significant percentage of annual trading volume and there will be a liquidity effect at these companies, albeit short lived:

The Sum of the Effects
In summary, buybacks can increase value, if they lower the cost of capital and create a tax benefit that exceeds expected bankruptcy costs, and can increase stock prices for non-tendering stockholders, if the stock is under valued. Buybacks can destroy value if they put a company’s survival at risk, by either eliminating a cash buffer or pushing debt to dangerously high levels. They can also result in wealth transfer to the stockholders who sell back over those who remain in the firm, if the buyback price exceeds the value per share. 

What about the share count effect? This is the red herring of buyback analysis, a number that looks profoundly meaningful at first sight but is useless in assessing the effect of a buyback, on deeper analysis. Let’s start with the obvious. A stock buyback will always reduce share count. For those lazy enough to believe that dilution is the bogeyman, and that less shares is always better than more, buybacks are always good news. However, lower share count often does not signify higher value per share and it may not even signify higher earnings per share (or whatever per share metric you use). For those slightly less lazy, focused on earnings per share, the assessment of whether a buyback is good news boils down to estimating how much earnings per share goes up after it happens. In a world where PE ratios stay constant, come out of sector averages, or are just made up, this will translate into higher price per share. The problem is that a buyback alters the risk profile of a firm and should also change its PE ratio (usually to a lower number).

To assess the effect of a buyback, you have to consider the full picture. You have to look at how it is financed (and the effect it has on debt ratio and cost of capital) and how the stock price relates to its fair value (under priced, correctly priced or over priced) to make a judgment on whether stockholders will benefit or be hurt by the stock buyback. I have a simple spreadsheet that tries to do this assessment that you are welcome to take for a spin.

Back to the Market
Now that we have the tools to assess how and why stock buybacks affect stockholders in the companies involved, let’s use them to look at whether the buyback “binge” in the market is good news, neutral news or bad news, at least in the aggregate.  The article in the Economist provides the perspective of those who believe that stock buybacks are the most destructive trend in corporate America. Looking at the value destruction pathways described in the last section, this group believes that the stock buybacks at US companies are increasing leverage to dangerously high levels and/or reducing investment in good projects. But are these contentions true? Let’s check the facts:

1. The leverage story: The notion that US companies are dangerously over levered seems to be built on two arguments: the aggregate debt levels of businesses as reported in the US national accounts and on anecdotal evidence (Apple borrowed money to do buybacks, so every one must...). To examine this argument,  I have estimated debt levels at US companies from 1980 to 2013 in the graph below, both as a percentage of capital (book and market) and as a multiple of EBITDA.
Debt as % of capital & multiple of EBITDA: All US companies (Source: Compustat)
It is true that overall financial leverage, at least as measured relative to book value and EBITDA has increased over time (though it has remained relatively stable, as a percent of market value). While this increase can be partially explained by decreasing interest rates over the period, it is worth asking whether buybacks were the driving force in the increased leverage. To answer this question, I compared the debt ratios of companies that bought back stock in 2013 to those that did not and there is nothing in the data that suggests that companies that do buybacks are funding them disproportionately with debt or becoming dangerously over levered.
Data from 2013: Debt burdens at buyback versus no-buyback companies
Companies that buy back stock had debt ratios that were roughly similar to those that don't buy back stock and much less debt, scaled to cash flows (EBITDA), and these debt ratios/multiples were computed after the buybacks.

2. The under investment story: The belief that US companies in sectors other than technology have been reinvesting less back into their businesses is widespread, but let’s check the facts again. In the table below, I look at capital expenditures at US firms collectively, as a percent of revenues and invested capital, from 1980 to 2013: 
Capital Expenditures, Revenues and Invested Capital: US companies (Source - Compustat)
The trend line (on everything other than cap ex as a percent of sales) does back the conventional wisdom, and since buybacks went up over the same period, the bad news bears seem to win this round, right? Well, not so fast! What if investment opportunities in the US, in sectors other than technology, are drying up, either because of global competition or due to industry maturation? If this is the case, not only should you expect exactly what you observe in the data (less reinvestment, more cash returned) but it is a good thing, not a bad one. Before you get too heated under the collar, there are three things to remember in this debate.
  1. The first is that there is little evidence that companies that buy back stock reduce their capital expenditures as a consequence. The table reports on the capital expenditures and net capital expenditures, as a percent of enterprise value and invested capital, at companies that buy back stock and contrasts them with those that do not, and finds that at least in 2013, companies that bought back stock had more capital expenditures, as a percent of invested capital and enterprise value. When you net depreciation from capital expenditures (net cap ex), the two groups reinvested similar amounts, as a percent of enterprise value), but the buyback group reinvested more as a percent of invested capital.
    Capital Expenditures & Net Capital Expenditures = Capital Expenditures - Depreciation; US firms in 2013
  2. The second is that the cash that is paid out in buybacks does not disappear from the economy. It is true that some of it is used on conspicuous consumption, but that is good for the for the economy in the short term, and a great deal of it is redirected elsewhere in the market. In other words, much of the cash paid out by Exxon Mobil, Cisco and 3M was reinvested back into Tesla, Facebook and Netflix, a testimonial to the creative destruction that characterizes a healthy, capitalist economy. 
  3. The third is the notion that more reinvestment by a company is always better than less is absurd (unless you are a politician), especially if that reinvestment is in bad businesses. In the table below, I have listed the ten companies that were the biggest buyers of their own stock over the last decade (using the Economist's ill advised heading for those who buy back stocks):

As a stockholder in any of these companies, can you honestly tell me that you would rather have had these companies reinvest back in their own businesses? Put differently, how many of you wish that Microsoft had not bought back $100 billion worth of shares over the last decade and instead pumped that money into more Zunes and Surfaces? Or that Hewlett Packard instead of paying out $60 billion to stockholders had bought three more companies like Autonomy (and written them off soon after)? Or that Cisco had spent the $70 billion in buyback money on a hundred small acquisitions? If, as the Economist labels them, these companies are cannibals for buying back their own stock, investors in these companies wish they had more voracious appetites and eaten themselves faster.

There are two other issues brought up by critics of stock buybacks. One is that there is firms may buy back stock ahead of positive information announcements, and those investors who tender their shares in the buy back will lose out to those who do not. The other is that there is a tie to management compensation, where managers who are compensated with options may find it in their best interests to buy back stock rather than pay dividends; the former pushes up stock prices while the latter lowers them. Note that doing a buyback ahead of material information releases is already illegal, and any firm that does it is already breaking the law. As for management compensation, I agree that there is a problem, but buybacks are again a symptom, not a cause of the problem. In my view, it is poor corporate governance practice on the part of boards of directors to grant huge option packages to managers and then vote for buybacks designed to make managers even better off. Again, fixing buybacks does nothing to solve the underlying problem.

Wrapping up
I think that both ends of the spectrum on buybacks are making too much of a simple cash-return phenomenon. To the boosters of buybacks as value creators, it is time for a reality check. Barring the one scenario where companies that buy back stock stop making value-destructive investments, almost every other positive story about buybacks is one about value transfers: from taxpayers to equity investors (when debt is used by an under levered firm to finance buybacks) and from one set of stockholders to another (when a company buys back under valued stock).

To those who argue that buybacks are destroying the US economy, I would suggest that you are using them as a vehicle for real concerns you have about the evolution of the US economy. Thus, if you are worried about insider trading, executive compensation, tax-motivated transactions and or under investment by the manufacturing sector, your fears may be well placed, but buybacks did not cause of these problems, and banning or regulating buybacks fall squarely in the feel-good but do-bad economic policy realm.

Is it possible that some companies that should not be buying back stock are doing so and potentially hurting investors? Of course! As someone who believes that corporate finance at many companies is governed by inertia (we buy back stock because that is what we have always done...) and me-too-ism (we buy back stock because every one else is doing it...), I agree that there are value destroying buybacks, but I also believe that collectively, buybacks make far more sense than dividends as a way of returning cash to equities. In the Economist article, I am quoted as saying that dividends are a throwback to the nineteenth century (not the twentieth), when stocks were offered as investment choices to investors who were more used to bonds and that fixed, regular dividends were designed to imitate coupons. Since equity is a residual claim, it is not only inconsistent to offer a fixed cash flow claim to its owners, but can lead (and has led) to unhealthy consequences for firms. In fact, I think firms are far more likely to become over levered and cut back on reinvestment, with regular dividends that they cannot afford to pay out, than with stock buybacks.

Attachments:
  1. Stock buybacks, dividends, stock issuances - Aggregate for US companies (1980-2013)
  2. Debt ratios/multiples - Aggregate for US companies (1980-2013)
  3. Buyback effect calculator

41 comments:

  1. Nice analysis. But, a couple of questions:

    1. The data regarding share buybacks does not indicate whether it is a gross or net figure. For example, many companies issuing stock options repurchase shares in the market and then deliver those shares to the person exercising an employee stock option. The net effect of this on outstanding shares is zero, but I suspect that this might show up in your data as a "share repurchase". Is that correct? By the way, if this is correct about the use of gross rather than net, it would partially explain why "share repurchases" appear to increase in the 1990's. The incidence of stock options as a form of incentive compensation increased dramatically about that time.

    2. Your analysis does not appear to offer any direct help on evaluating whether share buybacks are superior to the obvious alternative. True, you initially list some differences between these alternatives, but there is little attempt at making any qualitative judgement, other than that share buybacks are pretty much a mixed bag. Further thoughts on that would be interesting.

    ReplyDelete
  2. Vivian,
    1. The buyback is a gross figure. While option exercises may be the motive, the sheer value of the buybacks overwhelms the option exercise portion. I will try to see if I can tease out from the dat a net effect.
    2. As I see it, there is little effect on the firm of a buyback vs a special dividend. The choice therefore will depend on the investors you have as a firm, and whether they like dividends or stock buybacks. That choice will be driven by cash needs and tax considerations.

    ReplyDelete
  3. Thanks.

    I think an important issue is in analysts using a constant P/E multiple. If you use a constant EV/EBITDA, you would immediately find there is no theoretical reason for a higher share price after a buyback.

    Buybacks vs dividends mainly boil down to what is price versus intrinsic value. What are your thoughts on price-to-book as an indicator of whether buying back stock makes sense? Berkshire for example uses a 120% of book value maximum threshold.

    ReplyDelete
  4. Thanks for all of your posts! I don't know if it's relevant to valuation, but have you thought about writing a post on the call for universities to divest certain stock holdings from their endowments? I've read many opinions about the issue and whether or not divesting will actually have an effect on the companies "bottom line," but I would appreciate an analysis and some facts from a financial expert to help clarify this issue.

    Cheers,
    Ian

    ReplyDelete
  5. @Thijs, you are correct that simple theory suggests the share price should not change when there is a buyback. But theory (and evidence) also suggest that price DOES fall when a dividend is paid. So if you return cash with buybacks instead of dividends, that effectively leads to higher share prices over time, and thus the returns to the investor come in the form of capital gains rather than dividends.

    ReplyDelete
  6. 3rd Moment,
    The theory says nothing of the sort. In fact, I state the conditions that have to hold for a buyback to not affect price: it has be funded entirely with cash, the stock has to be correctly priced and there can be no feedback effects on reinvestment. If any of these conditions don't hold, the price will be affected. And it is obvious (and not meaningful) to state that the returns to stockholders, with buybacks, will take the form of price appreciation, not dividends. It is only if you bring investor taxes into the process that this will make a difference.

    ReplyDelete
  7. Ian,
    When you add a constraint to an equation (as you are when you restrict investments in any sector), the best you can hope for is that you are not worse off. Any arguments that a constrained optimal is better than an unconstrained optimal are on quicksand.

    ReplyDelete
  8. Prof, you have beautifully captured various aspects related to stock buybacks.

    However, I note that value of the firm increasing or decreasing based on changes to the cost of capital can be attributed not only to stock buybacks, but also to special dividends. I consider this is as a case of recapitalization process rather than stock buyback phenomena. Here buybacks are only one of the tools, special dividends being another, and borrowing more (no dividends or buybacks) is yet another.

    Therefore, I am not going to mention about stock buybacks funded by debt.

    At a rational level, stock buybacks make sense only on two grounds:
    1) The firm should have excess cash, i.e. it has more cash than its reinvestment needs. By reinvestment I mean naturally positive npv projects. If it does not have any good projects, reinvestment need is zero and all cash is excess cash. And
    2) The stock price is lower than its intrinsic value. If not, why should the managers use cash (belonging to all shareholders) to give an opportunity to the tendering shareholders to transfer wealth from the remaining shareholders to themselves?

    Assuming that the firm does have excess cash, the value of total equity should change post buyback.

    Value of equity before buyback = Value of operating business + (excess) Cash

    Since cash goes out of the firm, the value of total equity should reduce post buyback.

    Value of equity after buyback = Value of operating business

    Now, the wealth of the shareholders (on a per share basis) is assessed depending upon the price at which buybacks are carried out. If price paid is lower than value, the transfer of wealth is from the tendering to the remaining shareholders. But then, as you rights put they chose it. If price is more than value, it is the other way around. But then, it is poor management. If price paid is equal to value, there is no change in wealth of the either group.

    To do buybacks with debt is nonsense and defeats the purpose of long term value creation.

    ReplyDelete
  9. This is an important topic and a well written piece. You have given short shrift to the link to exec comp - briefly mentioning it at the end. I think this deserves a deeper exploration given the short-termism prevailing in exec suites.
    1. Increased leverage always look benign in good times - then turns burdensome as profit margins revert to historical levels.
    2. The high level of buybacks recently has coincided with a high level of insider selling.
    3. It sort of looks like managements are levering up companies ( to levels that will look dangerous in the future), cashing out and using share-buybacks to protect the share price as they cash out. In other words , sucking the enterprise dry, before they quit to "spend more time with family".

    is that too cynical?

    ReplyDelete
  10. Keep it simple: Buybacks are a return of cash to stockholders

    This is fundamentally untrue.

    Shareholders that keep their shares do not get cash, they hopefully get an increase in the value of their shares if management buys back shares for less than fair value, but that is not the same as cash.

    Buybacks are a return of cash to former-shareholders.

    "As a stockholder in any of these companies, can you honestly tell me that you would rather have had these companies reinvest back in their own businesses?.....Or that Hewlett Packard instead of paying out $60 billion to stockholders had bought three more companies like Autonomy (and written them off soon after)?"

    Hewlett Packard is a great example of what I mean. They did not pay out $60 billion to stockholders. They bought shares at inflated prices from exiting stockholders at the expense of current stockholders. No cash from the 60 billion went to current shareholders.

    I would have rather $60 billion went to existing shareholders as dividends.

    ReplyDelete
  11. Professor D,

    Do you have any thoughts on the rules governing the disclosure of buybacks? From reading lots of disclosures about buybacks, many boards seem willing to approve buybacks that lack specific criteria and are fairly open ended in terms of their discretion about when and at what price share repurchases will take place. As far as I know firms do not needs to file interim 8-Ks to update shareholders on meaningful repurchases but given the speed and increased volume of markets, perhaps that should be revisited. Also, as with many disclosures, it seems that perhaps the SEC could issue some guidance to help streamline disclosures about stock buybacks in some kind of standardized format that gives investors a clearer picture of the success or failure of programs. (As an aside, does the SEC provide model disclosures for common items, and if not why not (maybe a future blog post topic?))

    Thank you,

    ReplyDelete
  12. Continuing on the link between share-buybacks and exec comp:
    - Anticipating the retort that the $volume of share-buybacks is orders of magnitude larger than $volume of exec share sales:
    - I dont find that to be a convincing argument. If an exec has the opportunity to increase his net worth from $10 million to $30 million - but it would require $10 Billion of share buybacks to make it happen - why not? As long as the corporate debt markets are squacking like geese for junk bonds yielding 4% - sell as much as you can - and buyback as many shares as you can.
    Best to retire before the next downturn hits- when EBIT margins will probably decline from the current 14% ( large cap cos) to a more normal 6 or 7% and the increased leverage will start causing distress.

    This is not a very original concept. After all - the LBO/Private equity groups do much the same. Its just that perhaps in the last 3 years or so the technique has been adopted on a wider scale by public companies.

    ReplyDelete
  13. "The other is that there is a tie to management compensation, where managers who are compensated with options may find it in their best interests to buy back stock rather than pay dividends; the former pushes up stock prices while the latter lowers them".

    Actually, perversely, in many markets such as in Australia, the payment of more dividends or a higher payout ratio would tend to push stock prices up. When buybacks are preferred over high and consistent dividend payouts, prices do not seem to be pushed up to the same degree.

    ReplyDelete
  14. Dan,
    Your post does not seem to contest my point that buybacks are a return of cash to stockholders. What you do seem to be suggesting is that those who tender back their shares are the beneficiaries of the cash return and that those who remain get punished. I disagree with this, and the evidence on how stock prices do at companies that do buybacks backs me up, but if you do believe that non-tendering stockholders get the short end of the stick, I have a simple investment recommendation. If you own shares in a company (say Krispy Kreme) and it announces a buyback, tender your shares back and become part of the winning team (in your story).
    As for those long-suffering stockholders in HP, I would suggest that it is not the buybacks that sealed your fate but the fact that you chose to tie yourself to an incompetent management and an inept board of directors.

    ReplyDelete
  15. Executive compensation is a big topic and deserves its own post, but here is the reason why I side swiped the issue in this post. Does the form of executive compensation affect dividend policy? Of course. If you pay managers with options, you will encourage them to buy back stock, rather than pay dividends. As a stockholder, the question is whether this hurts you. It will only if the buyback is funded with debt that the company cannot afford to carry or if it results in under investment. In my post, I argue that there is no evidence of either. So, as a stockholder, it matters little to me that executive compensation affects the form of cash return.
    If your point is that executive compensation is too large and too tied to equity, that is a different issue and it is tied to corporate governance. The buyback then is a symptom of a corporate governance problem and you have to deal with the problem, not the symptom.

    ReplyDelete
  16. prof Damodaran:
    1. I am surprised that Debt/EBITDA seems fairly stable in 2008-2009, (when earnings plunged) - according to your chart.
    If the data is correct, then I have to admit my concerns are probably overstated.
    Still, personally, I much prefer dividends. I dont like companies paying out cash disproportionately to insiders. I dont like executives overwhelmingly selling, while institutional passive money is buying.

    ReplyDelete
  17. The 2008 recession was devastating for financial service firms but the earnings at other firms were not as affected. Also, I sympathize with your view about dividends, but what do insiders have to do with buybacks? Buybacks, or at least most of them, are tender offers where everyone has the option of tendering. I am not aware of any evidence that insiders are disproportionately part of the tendering group.

    ReplyDelete
  18. " where everyone has the option of tendering"
    Sure, but the evidence is that insider sales are running high while institutional ownership of equities has been climbing.
    I would love to see the buyback numbers at a more granular level. Would be interesting to compare two classes of companies: a) Companies that have large exec stock programs , and b) Companies that have modest or minimal exec stock programs.
    That might give some clues as to the motivation behind buybacks.

    Also, Buybacks accelerated in 2006-2007 and then plunged. Again, accelerated in 2012-2014. It would be interesting to look over a decade or more of history to test if volume-weighted buybacks are happening at much higher than the average prices over a long period.

    ReplyDelete
  19. Thank you for this clear and helpful post. What is the logic for netting buybacks against stock issuance? If stock issuance is not conditioned upon expected buybacks, and therefore present in the dividend counterfactual then it seems like a sunk cost.


    ReplyDelete
  20. Thanks for your post. Share buybacks are clearly an interesting topic of debate especially amongst people looking for reasons to predict the next market meltdown.

    An additional confusion that sometimes arises is when people use LFCF/share projections for their DCF calculations. For the LFCF/share projections they assume that the company uses the free cash to buyback shares but do not deduct that amount from the numerator. This may seem an obvious error but I have been shocked how many times I have seen people run their analyses with LFCF/share projections assuming share buybacks but not adjusting the numerator (LFCF) for the same.

    From a DCF perspective using projections for LFCF/share do you think the two approaches yield similar results:

    a) (Cash from Ops) Less (Capex) / (Current Sharecount)

    b) (Cash from Ops) Less (Capex) Less (Cash used to buyback shares) / (Current Sharecount - Cum. Shares Repurchased)

    thanks.
    yani

    ReplyDelete
  21. Thanks for your post. Share buybacks are clearly an interesting topic of debate especially amongst people looking for reasons to predict the next market meltdown.

    An additional confusion that sometimes arises is when people use LFCF/share projections for their DCF calculations. For the LFCF/share projections they assume that the company uses the free cash to buyback shares but do not deduct that amount from the numerator. This may seem an obvious error but I have been shocked how many times I have seen people run their analyses with LFCF/share projections assuming share buybacks but not adjusting the numerator (LFCF) for the same.

    From a DCF perspective using projections for LFCF/share do you think the two approaches yield similar results:

    a) (Cash from Ops) Less (Capex) / (Current Sharecount)

    b) (Cash from Ops) Less (Capex) Less (Cash used to buyback shares) / (Current Sharecount - Cum. Shares Repurchased)

    thanks.
    yani

    ReplyDelete
  22. Dr. Damodaran, one quick question - in ERP calculatuion, aren't we counting the effect of share buybacks twice, once in the yield/carry and second in growth (per share growth)? Sorry, I am new to your blog and recently read you piece on ERP.

    Thanks,
    Anoop

    ReplyDelete
  23. Some great points from an investor's point of view.
    However, for the real economy I would think that when companies use a larger share of their profits on buybacks rather than investing in their business this does not bode well for the economy at large because the companies are saying: We see less business opportunities going forward, i.e. we have a negative view on our market. This can spiral into an evil cycle potentially hurting suppliers to those companies and in the end consumers as well.

    ReplyDelete
  24. Perhaps these examples are helpful:

    http://hvass-labs.blogspot.com/2014/03/the-evils-of-debt-funded-share-buybacks.html

    http://hvass-labs.blogspot.com/2014/02/how-to-value-apples-share-buybacks.html

    ReplyDelete
  25. Anoop,
    Excellent point about double counting and I go out of my way to avoid it. That is why I don't use analyst bottom up estimates of growth in earnings per share for individual companies and instead use growth rates in aggregate earnings in the S&P 500.

    ReplyDelete
  26. Michael,
    You are absolutely right. An economy where all that companies can do is buy back stock is in deep trouble, but I think may be a little extreme for the US economy, where cash from buybacks can go into new businesses (many of them in technology).

    ReplyDelete
  27. Dear Prof,
    Great write-up. One note: funding a buyback with a new issue sounds perverse indeed. However, I recall Easterbrook's argument that "New investors do not suffer under the
    collective choice disabilities of existing investors.[...] New investors are better than old ones at chiseling down agency costs" (Easterbrook'84). In other words, the market gets more information/insights about the company when it uses the new issue proceeds for the buyback. This translates into better monitoring, presumably better pricing and mitigation of agency costs. Given this hypothesis, new issue might not be a perverse buyback funding source.
    Cheers,
    Evgeny

    ReplyDelete
  28. Hello,

    My question does not directly concern this blog post, but I'll post my question here anyway.

    I checked Your latest ERP spreadsheet for the S&P 500 index and I noticed that You have used unit adjuster for year 2012 (instead of unit adjuster for the latest period) in calculating dividends and buybacks for the trailing 12 month period. I refer to the sheet called "Buyback & Dividend computation" in the latest ERP spreadsheet.

    Is this just a mistake, or if it's not, what is the explanation for this?

    Thank you in advance,

    ReplyDelete
  29. Jarmo,
    You are right. I screwed up. Fixed now. Changes ERP to 5.43%.

    ReplyDelete
  30. I'm coming back to my initial comment in this thread. You wrote, in response:

    "1. The buyback is a gross figure. While option exercises may be the motive, the sheer value of the buybacks overwhelms the option exercise portion. I will try to see if I can tease out from the dat a net effect."

    I suspect that the effect of stock option exercises and other employee stock grants has a much more significant effect on the net versus gross repurchase total than you are admitting here. Just out of curiosity, I picked one company, Microsoft, to see what the effect might be. Here's what I found based on their annual report:

    The number of outstanding common MSFT shares in 2010 was 8,668 billion. At end of FY 2013 it was 8.328 billion. The represents a net decrease of 340 million shares. Based on a share repurchase program authorized in late 2008, MSFT repurchased 747 million shares during that period of time. The difference between the shares repurchased and the net balance difference is thus 407 million shares. It appears that most, if not all that difference is due to various employee stock and option programs. While MFST states in its annual report that it issues "new shares" to fund their various employee stock programs, this is of no matter, because the shares are fungible.

    In order to understand why there has been such a large increase in share repurchases, one needs to have a thorough understanding of the role employee compensation plays here. It stands to reason that in a rising stock market (to wit, the last 5 years), one would see an increasing number of stock repurchases to fund these programs. You've indicated that "the sheer value of buybacks overwhelms the option exercise program" (which I take to mean all forms of employee stock comp), but you give no evidence for that. You may be right, but unless there is a more thorough study, I'm inclined to be more skeptical. Another indication is that in the year 2013 alone, MFST *authorized* 425 million shares under their various employee stock schemes.

    Another effect could be the issuance of stock to finance corporate acquisitions (allowing more favorable tax treatment to the takeover target's shareholders). Companies may wish to counteract this dilution by share repurchases (allowing those who don't mind paying tax to sell). In essence, this is an indirect way of paying cash for a corporate acquisition (Acquisition Co Shareholder A gets cash and indirectly delivers the shares to Target Company Shareholder B).

    In sum, there is likely a lot more going on here than companies repurchasing their own shares in order to return cash to shareholders in lieu of dividends. That's why I think it is perhaps necessary to look at *net* repurchases rather than gross repurchases as announced to the public as such.

    ReplyDelete
  31. Vivian,
    I don't quite get your point. I do look at net buybacks in this figure. That is the number that you see in the net cash yield column. It is net of the shares issued to managers.

    ReplyDelete
  32. Thanks for the response. I don't recall seeing that graph in the original post. Has it been updated? And, how do you explain this response to me on that very question?

    "1. The buyback is a gross figure. While option exercises may be the motive, the sheer value of the buybacks overwhelms the option exercise portion. I will try to see if I can tease out from the dat a net effect."

    If you were originally and consistently using a "net effect", there would be nothing left to "tease out". Does that net figure carry through to the other graphs on "cash returned"? It's not at all clear.

    Viv

    ReplyDelete
  33. This is an interesting topic for discussion. However I must say that the author's arguments lose credibility when he cannot clearly explain himself or does not accurately record changes to his narrative.

    First: "Your post does not seem to contest my point that buybacks are a return of cash to stockholders. What you do seem to be suggesting is that those who tender back their shares are the beneficiaries of the cash return and that those who remain get punished. I disagree with this, and the evidence on how stock prices do at companies that do buybacks backs me up, but if you do believe that non-tendering stockholders get the short end of the stick, I have a simple investment recommendation. If you own shares in a company (say Krispy Kreme) and it announces a buyback, tender your shares back and become part of the winning team (in your story).
    As for those long-suffering stockholders in HP, I would suggest that it is not the buybacks that sealed your fate but the fact that you chose to tie yourself to an incompetent management and an inept board of directors."

    This makes absolutely no sense. The (a) management team effectuated the (b) buybacks. You are saying 100% of the blame for underperformance should go to (a) and 0% of the blame should go to (b). Why? If management improperly allocated capital, one must blame both (a) the management AND (b) the method of capital allocation, aka the buyback. BTW the commenter is spot on, had HP's incompetent management merely paid the cash out as dividends, shareholders would've been better off as they would have received more value in cash than through reduction in the share count. Your response to his comment was at best poorly worded and at worst spurious.

    Secondly Vivian made an excellent point, namely that in modern corporate finance buybacks are used to make dilution through stock option issuance. Many contemporary commentators gloss over or are incapable of giving this subject the attention it deserves (BTW for an excellent example of this in the case of Adobe, see Michael Burry's Scion Capital letter). The way the author just glosses over and minimizes this point is highly frustrating; I would expect much more from a finance professor, with all due respect.

    ReplyDelete
  34. Grim Reaper,
    I don't quite understand your point about buybacks being to blame for bad capital allocation. How can cash that leaves the company be blamed for bad investments made by cash that stays in the company? That would be just as absurd as saying that dividends are to blame for bad investments made by a company. I think that our disagreement is in the characterization of buybacks. You are classifying it as an investment by the company (a capital allocation) and I am arguing that it is a cash return.

    As for Vivian's point about compensation, I agreed with her that some of the stock issuances may have been to cover management compensation (and it is impossible to tell how much since that is not publicly disclosed) but I do report the net cash yield (net of all stock issuances).

    ReplyDelete
  35. “Grim Reaper,
    I don't quite understand your point about buybacks being to blame for bad capital allocation. How can cash that leaves the company be blamed for bad investments made by cash that stays in the company? That would be just as absurd as saying that dividends are to blame for bad investments made by a company. I think that our disagreement is in the characterization of buybacks. You are classifying it as an investment by the company (a capital allocation) and I am arguing that it is a cash return.”

    Hmm. How can a buyback be an example of bad capital allocation? Hint: does the price at which the shares are repurchased have any importance? What if Microsoft repurchases shares at $1/share? What if, instead, it pays $2000 per share? I never made the distinction of cash staying in or outside the company. My point was limited to *capital allocation* (“capital” meaning cash or other resources, and “allocation” meaning use of).

    As for Vivian's point about compensation, I agreed with her that some of the stock issuances may have been to cover management compensation (and it is impossible to tell how much since that is not publicly disclosed) but I do report the net cash yield (net of all stock issuances).
    “Net cash yield”? Again, that’s entirely missing the point. As long as cash paid for buybacks exceeds cash received from stock options, you’ll have a positive net cash yield. HOWEVER, this does not account for the fact that many companies dilute when compensating with stock. Some companies, like Ebay, pay billions repurchases shares (high net cash yield) yet their share count never actually decreases…almost like magic.

    ReplyDelete
  36. Grim Reaper,
    Let me clarify why a buyback is not a capital investment. Let me take your example of Microsoft, where you ask whether it makes a difference as to whether they buy back at $1 a share or $2000/share.

    Here is what you should do if Microsoft offers to buy its shares back at $2000. Sell your shares. That is the option you don’t have if Microsoft overpays on an acquisition or on an investment. You are not a helpless bystander in an buyback. You get to decide whether the offered price is too high or low. If you think the buyback price is too high, you can cash out and those who don’t implicitly don’t agree with you. If Microsoft tries to buy shares at $1, no one will tender. So, it will be pointless. If you are not convinced, let's agree to disagree and you should find someone who does agree with you & continue this conversation with them.

    And if a company issues enough shares to its employees to overwhelm its buybacks and cause dilution, its net cash yield will reflect it since stock issuances will exceed stock buybacks.

    ReplyDelete
  37. Professor,

    Thank you for the article. I had a couple of questions/comments.

    1) Some investors like share buy backs because they perceive it as a more tax efficient way of returning cash to shareholders, but to realize the benefits of a buyback, the shareholder has to sell his shares and when he sells his shares he incurs a tax liability, which could be short term capital gains which are taxed at a higher rate than dividends. Wouldn't this effect make buybacks less tax efficient on average? (now that dividends are taxed at 15%)
    2) You note that buybacks could be have a positive, neutral or negative effect on equity valuations. Ins't a share buyback that has a neutral effect on the share price actually value destroying for shareholders due to the opportunity cost of missed dividends?
    3) You list the Cap Ex/Enterprise Value of no buyback firms at 3.82%, and of buyback firms at 4.27% so how can the Cap Ex/Enterprise value of all firms be 5.53%?

    ReplyDelete
  38. Aswath- this analysis is exceptional! Much better than the two McKinsey articles I came across, and honesty better than anything else I've come across on the internet so far. I see good things come out of NYU MBA. Thank you very much!! PS. If you want a good laugh, read through these two McKinsey articles on the topic and it will become obvious why they're McKinsey consultants and not equity analysts/portfolio managers/investors:

    http://www.mckinsey.com/insights/corporate_finance/the_value_of_share_buybacks

    http://www.mckinsey.com/insights/corporate_finance/paying_back_your_shareholders

    Thanks again!!

    ReplyDelete
  39. Hello!
    I can't believe I'm so late in this. I found it terrific. Just have a couple of questions: Where does the data of the last three tables regarding Capex and Debt come from?
    Thanks again.
    Best,

    Horacio

    ReplyDelete
  40. Hi,

    Can you please recheck the following formula in buyback.xls spreadsheet :

    Tab After tax cost of debt - Post buyback (C35) - the formula is

    =IF(B22=0,B12*(1-B27),B23*(1-B27))

    In the scenario when there partial funding of buyback with New debt, the formula only seems to consider the "cost of new debt".

    I think this should be the weighted average cost of all debt (old + new)i.e. new formula (in my opinion)

    =IF(B22=0,B12*(1-B27),((B10*B12)+(B22*B23))/(B10+B22)*(1-B27))

    Regards
    Prithvi

    ReplyDelete

Given the amount of spam that I seem to be attracting, I have turned on comment moderation. I have to okay your comment for it to appear. I apologize for this intermediate oversight, but the legitimate comments are being drowned out by the sales pitches and spam.