« Vernon Smith on Asperger's syndrome | Main | Larry Summers sentence of the day »
Did Paul Krugman commit the Junker fallacy?
It’s possible that sluggish business investment reflects lack of confidence in the economic outlook —... that’s understandable given the bursting of the housing bubble...But...there is a more disturbing possibility. Instead of investing in physical capital, many companies are using profits to buy back their own stock.
Here are longer excerpts. Of course stock buy-backs do not take away resources for subsequent investment. The money used to buy shares can still be funneled into the purchase of capital goods, as no real opportunities have been taken off the table.
That said, cash "in the firm" is more likely to be invested than "cash in the hands of investors," for reasons of credit rationing and other institutional rigidities (for instance borrowing money brings more outside scrutiny). Given the size and profitability of these firms, however, I do not expect that the credit rationing effect is a large one. The causes of the sluggishness of investment are thus to be found elsewhere than through this financial mechanism.
Here is my earlier post on the Junker fallacy. I believe this sort of argument was first criticized by Fritz Machlup in his book on the stock market. Of course if you wish to save the claim through various second best arguments, comments are open...
Posted by Tyler Cowen on April 30, 2007 at 01:06 PM in Economics | Permalink
Comments
It is a myth resulting from bad government accounting that US businesses are underinvesting. Which is not to say there isn't "enough" investment in the US (although there may be enough), whatever "enough" means. But corporate America is in no danger from underinvestment.
Posted by: happyjuggler0 at Apr 30, 2007 1:28:00 PM
By the way, there are only 8 countries in a list of 155 that have a higher corporate tax rate than the US.
Incentives matter....
Posted by: happyjuggler0 at Apr 30, 2007 1:55:00 PM
happy, isn't corp tax thing likely to work both ways, debt shield to save on taxes(buybacks) and capital investment(depreciation shield). assuming a st line method of deprecation an interest rate of 6% should equal a useful life of 16 yrs. it appears to me that companies can choose either route.
Posted by: sa at Apr 30, 2007 2:02:00 PM
I think a lot of the lack of investment in additional capacity is a result of late 90's bubble. In that time period firms spent enormous amounts of money on additional personnel and capacity that turned out to be wasted investments. The investments in expensive staff and high salaries for middling personnel were particularly egregious.
As a long-term investor I am quite happy when cash rich firms return profits to investors via stock buy-backs. It is a much more tax efficient way for companies to reward shareholders than via the doubley taxed dividends.
I am particularly angered when firms use excess capital to make acquisitions or invest in new business ventures where they have no core competency.
Posted by: lannychiu at Apr 30, 2007 2:03:30 PM
Tyler,
Sorry about not posting on the Junker line of thought, I was concentrating on the first part of the thesis that US corporate investment is low, not on the "dissappearing money syndrome" (my phrase) that Krugman is worried about.
Lannychiu is right that there is no economic difference between dividends and stock buybacks, just a substantial tax difference. So if Krugman has a problem with stock buybacks he also ought to have a problem with dividends.
The real reason why companies are buying bakc their stock is because debt is cheaper than the earnings/price ratio, and this is also the reason for the LBO boom (er, I mean private equity boom, sorry to not be pc). Companies to a large extent are not leveraged enough.
The notion that this phenomenon is about disintermediation strikes me as backwards. If anything it is about taking debt capital, which to a decent extent is coming from risk averse foreigners, and using it to invest in higher "yielding" (i.e. e/p ratio, not price/dividend ratio) equities. Thus there is a crowding in of investment from debt financed share buybacks, not a crowding out.
Your analysis that a high chunk of the sellers in buybacks will reinvest somewhere is indeed the most likely scenario. That fact by the way argues for eliminating all investment taxes and replacing them with tax increases elsewhere (or better yet spending cuts) since those non-investment activities are less likely to result in investment than currently taxes investments which would mostly reinvest any tax savings.
Posted by: happyjuggler0 at Apr 30, 2007 3:01:26 PM
Companies are doing share buybacks essetially as a more tax effective way of paying dividends because this is what their institutional shareholders, such as myself, are asking them to do. It isn't a hard sell as it also has the potential benefit of keeping management ESOs above water. But either way it tends to be more value enhancing to minority shareholders than companies spending it themselves on projects that tend to be dilutive to ROIC.
happyjuggler0,
There are two tax rates: Statutory and effective. The ones listed in your link are statutory. Depending on the tax code and what a company can claim as a deduction, there can be a big difference between the two.
Posted by: asiequana at Apr 30, 2007 3:03:04 PM
The real reason why companies are buying bakc their stock is because debt is cheaper than the earnings/price ratio,
Ahem. The real reason companies are buying back stock is because it is the most tax-efficient way to return capital to shareholders.
(I know, I know, I'm dead. But I'm still right!)
Posted by: Merton Miller at Apr 30, 2007 3:04:02 PM
If you go to the BEA capital accounts and look at the data for.
US Direct Investment Abroad:
Capital Outflows
without current cost accounting adjustment:
you get: (m$)
1999.........209,392
2000.........142,627
2001.........124,873
2002.........134,946
2003.........129,352
2004.........224,437
2005.........-12,714
Now the story is more complex then this, and you need to bring in the data
on direct investment abroad of earnings from early foreign investment.
But this first cut at the data clearly does not support the thesis that
the reason for weak domestic investments is firms sending their domestic
earnings abroad in the form of direct investments.
Moreover, on a net basis over the long run the stock market is not a net source of capital for corporate investment. Corporate pay outs of dividends
and in stock repurchases, etc., are a multiple of new capital raised for
capital spending or to capitalize new firms. On a net basis the stock market
is a mechanism for transferring corporate profits to the stock owners.
Posted by: spencer at Apr 30, 2007 3:07:43 PM
spencer,
Interesting numbers. I am curious (and ignorant here), but do those numbers take into account profits earned abroad but aren't repatriated? In other words, if IBM earns $x in India and repatriates $0.1x, reinvesting the other 90% in India, does that count as a $0.1x inflow into the US or a $0.8x net outflow?
Posted by: happyjuggler0 at Apr 30, 2007 3:22:49 PM
Also, 2004 looks severely above trend, perhaps 2005 is a reversion to the mean? Perhaps it was a worry about Kerry winning and getting your money out while the getting was good?
Posted by: happyjuggler0 at Apr 30, 2007 3:26:32 PM
asiequana,
Foreign countries also have deductions. Also countries have been known to give out a 0% tax rate, or a negative one all-in, to corporations to invest there. Just ask Intel.
I am still inclined to think that the US rate, even after adjusting for deductions, is still much higher than almost everyone else.
Also, regardless of zero sum issues, investment is also a positive sum game. Taxing our most growth enhancing activities isn't too smart. We subsidize human capital investment, but we tax investments in physical capital. The latter doesn't make sense.
Posted by: happyjuggler0 at Apr 30, 2007 3:32:35 PM
I don't get it. Isn't this basic "principles" sort of stuff? A firm should retain capital if it can invest it and earn a return that is greater than it could get for that capital if invested in another firm. If not, it should pay out that capital, via a dividend or a stock buyback. The mechanics of returning the capital are irrelevant. Aren't they? Am I missing something?
Posted by: Sameer Parekh at Apr 30, 2007 3:40:30 PM
If not, it should pay out that capital, via a dividend or a stock buyback. The mechanics of returning the capital are irrelevant. Aren't they? Am I missing something?
In a world of no taxes, they are equivalent. But dividends are paid to all shareholders, whether or not they want them, and are taxed at 15%. Buybacks are taxable only to the shareholder who sells his share to the company, and taxed at 5% ,15%, or ordinary rates, depending on his holding period.
So why do companies even pay dividends, rather than buying back shares? I posit that it is a costly signal to shareholders that the firm has fiscal discipline, and is confident in its ability to produce cash as a going concern.
Posted by: Merton Miller at Apr 30, 2007 4:21:21 PM
Oh, and while I am at it, there's nothing wrong with a company holding excess cash on its balance sheet in the hypothetical world with no taxes and no agency costs, as the firm's cost of capital will drop as the riskiness of its assets drops. But cash is double-taxed when held in a corporation, so it's best to get it out of there.
Man, I was a genius!
Posted by: Merton Miller at Apr 30, 2007 4:23:46 PM
happy juggler -- your are right. this data does not include reinvestment of earnings made abroad -- exactly what your example shows.
But remember for multinational corporations standard practice when they undertake a direct investment in a country is to borrow the bulk of the
money for that capital spending -- new facility --in the currency of the
country they are investing in. If the new facility is to have a 20 year life, for example, you have no idea what is going to happen to exchange rates over that period. So it is just a prudent hedging technique to borrow in that currency for the bulk of the fixed investment. The equity in knowledge that the multinational brings to the table is really what is important in such transactions.
Posted by: spencer at Apr 30, 2007 5:16:36 PM
Silly question, but is there any evidence of above average stock buyback
spending?
My cranial data bank has not noted a trend. Need data.
Posted by: save_the-rustbelt at Apr 30, 2007 5:18:30 PM
Happy juggler -- I do a series of the US stock of fixed capital per employee.
The change in this series, lagged one year is a very important determine of productivity growth. It is essentially the Solow one-third of productivity we understand.
But, this variable has been stagnating for the last few years. From 1950 to 1980 the trend growth rate of this series was 1.6%. Since 1980 the trend growth rate has been 0.9%. It stagnated in the 1980s, rebounded strongly in the 1990s and has been flat since 2000. this series leads me to worry very much that the level of capital spending is inadequate. I'm of the old school that the way you increase productivity and standards of living is to increase the capital per employee.
On the direct investment data. it is highly volatile. But on a smoothed basis it consistently showed strong growth prior to 2000 -- it is hard to show data on blogs so I limited the data to the last few years. Except for the one year of 2004 it has been very weak since 2000. There was a clear break in trend in 2000 when you look at a longer data series.
Posted by: spencer at Apr 30, 2007 5:27:45 PM
spencer,
Once again thanks for the info.
My initial post in this thread, namely the first response to Tyler's post, was in part an attempt to show that UC corporations aren't underinvesting from the point of view of the health of US corporations, they are merely doing it overseas. If their capital spending seems to be still "too low", perhaps it is a case of why buy an expensive machine that puts parts together, or a semi-automated one that makes the parts, when it is cheaper to pay someone $1 a day to do it? In other words their labor component is much higher thanks to the availability of cheap labor.
Anyway, what is good for the corporation isn't necessarily good for the US, which is a point I clumsily tried to make in that first post. It is not clear that "we" are investing "enough" in the US, regardless of what happens to the companies that are registered in the US.
Posted by: happyjuggler0 at Apr 30, 2007 6:55:52 PM
By the way, by their labor component is much higher, I am referring to the ratio actual work accomplished (via man or machine), not the ratio of the dollar cost of those two inputs.
Posted by: happyjuggler0 at Apr 30, 2007 6:58:41 PM
I don't see a Junker fallacy in the Neeman-Paserman-Simhon paper you link to: moving money abroad, meaning acquiring financial claims against foreigners, requires shifting real resources abroad to establish these financial claims against foreigners. Keeping the money under the mattress doesn't require shifting resources abroad (in exchange, you get interest on claims against foreigners which you don't get from money under the mattress) and allow others to use domestic resources for whatever they choose, including capital accumulation.
As for Krugman's suspicion that the US corporate sector now has, for a given set of prices and expectations, a higher propensity to pay out dividends: this would make a lot of sense, since two puzzles in asset pricing are that historically a)the firms that issue seasoned equity underperform firms that do not, b) firms that buy back their stock outperform firms that do not. The corporate sector may now be moving close to firm value maximization.
Posted by: stefan at Apr 30, 2007 11:12:00 PM
Hmmm...Krugman committed yet ANOTHER fallacy. How many does he get before we stop considering him an economist anymore?
Posted by: Methinks at May 1, 2007 2:12:28 AM
Generating growth in a business, especially a 'large' one isn't quite as easy as just investing more capital - especially in hi-tech. The firm I work at had tremendous growth in the 90's, but growth has slowed and we're buying back stock. The reason is that there is not a sound way to use that capital to grow our business. We are semi capital intensive, but most of our growth is driven by ideas. If your option is a bad capital expenditure or buying back stock, what do you do?
In the 90's you had tech firms that were growing quickly, Intel, Cisco, Microsoft, and not paying dividends. Don't all 3 now pay dividends? Hasn't Microsoft given back tens of billions of profits to shareholders? It's not that Bill Gates doesn't have ideas to spend it, just that those ideas are not in the core Microsoft business and wouldn't generate the returns on capital that many firms expect.
Krugman should also remember that all capital expenditures are not created equal - there was a lot of ridiculous spending on tech equipment by dot-coms and other techs (with investor $) that had very little chance of ever generating any positive return.
Posted by: NC Engineer at May 1, 2007 6:33:59 AM
After reading more of Krugman's article, another reason for rising corporate profits is that many unprofitable businesses from 2000 (or ones that were 'manufacturing' profits in their accounting department) are no longer in business, or have been reorganized.
Where Krugman might have had a point is that the US is borrowing heavily, and not significantly increasing its capital stock. Using a business analogy, you can borrow to build a factory, but don't want to borrow to put TV's in the break room and throw lavish parties.
US companies will continue their profitable ways, but a lot of those profits will head overseas to cover debts incurred.
Posted by: NC Engineer at May 1, 2007 6:50:57 AM
Don't forget we had a change in tax law, I'm not sure in what year off the top
of my head -- that encouraged multinationals to repatriate foreign earnings.
Posted by: spencer at May 1, 2007 8:39:55 AM
Spencer - the reduction in rates on repatriated money was a temporary change in effect for 2005 (with an extension into 2006 for select firms).
Krugman's initial statement ("It’s possible that sluggish business investment reflects lack of confidence in the economic outlook") is illogical and just wrong.
Posted by: jult52 at May 1, 2007 11:16:12 AM
