Analyzing the trend of corporate cash hoarding

Via MR via Ezra Klein Barry Ritholtz points out that the increase in corporate cash is not some new thing.

“1) The average cash-to-assets ratio for corporations more than doubled from 1980 to 2004. The increase was from 10.5% to 24% over that 24 year period. That was the findings of a 2006 study by professors Thomas W. Bates and Kathleen M. Kahle (University of Arizona) and René M. Stulz (Ohio State). When looking for an explanation, the professors found that the biggest was an increase in risk.

Indeed, the phenomena of corporate cash piling up has been going on for a long long time. You can date it back to the beginning of the great bull market in 1982 to 86, went sideways til the end of the 1990 recession. It has been straight up since then, peaking with the Real Estate market in 2006. The financial crisis caused a major drop in the amount of accumulated cash, but it has since resumed its upwards climb.

2) The total cash numbers numbers are somewhat skewed by a handful of companies with a massive cash hoard. Exxon Mobil, GE, Microsoft, Apple, Google, Cisco, Johnson & Johnson, Verizon, Altria, EMC, Disney, Oracle, etc.”

Tyler Cowen believes that because this issue has been around for a while, “there is less to this issue than meets the eye.”  Whether or not corporate cash piling up matters depends on the reasons behind the pile up, of which there are a few.

1. Corporate earnings overseas are not brought back due to tax implications.

2. Corporations are hedging against an increased perception of risk.

3. Lack of attractive investment opportunities.

4. The average new successful company is successful because it knows how to grow, and these types of companies are more likely to hold onto cash than companies run by CEOs who are focused on maximizing shareholder value via more direct means. Over time, the largest companies will be more likely to be the cash hoarders who don’t pay dividends.

5. There have been more very successful IPOs recently and recent IPOs are more likely to hoard cash.

6. The success of companies in #4 and #5 reduces the pressure on other CEOs to keep their cash levels relatively low.

7. Tangible assets have declined in value relative to non-tangible assets, so some companies hold more cash to keep up a safety cushion of real assets.

8. The total amount of cash is increasing in tandem with the total amount of assets.

To address #1, congress passed the American Jobs Creation Act, which reduced the US tax rate on repatriated earnings of foreign subsidiaries from 35% to 5.25% in 2004 and 2005, encouraged corporations to repatriate foreign earnings and invest (many decided to increase their share buy backs) but since that ended it wouldn’t be surprising if the foreign earnings started to pile up.

I decided to go to the flow of funds and look at the data myself, normalizing it for total assets. It turns out that cash and near cash assets, the pink line, hasn’t deviated from its long run mean. The main trend is the decline in tangible assets and the rise of miscellaneous assets.

Source: Fed Flow of Funds. Note: These are only domestic assets, leaving out foreign earnings held overseas. Near cash assets include deposits, currency, money market, security RPs, commercial paper, treasuries, Agency & GSE back securities and municipals.

It is likely that miscellaneous assets include cash-like assets, in order to square these results with those of Bates, Kahle and Stulz in their 2009 paper “Why Do U.S. Firms Hold So Much More Cash than They Used To?” mentioned by Ritholtz in his post. This study, unlike the Flow of Funds data in the chart below, also includes cash held at foreign subsidiaries.

Looking at the data in the paper more closely, it looks as if numbers #2, #4 and #7 are very important. All of these are related to high tech firms replacing old school manufacturing firms which paid more dividends, had relatively stable cash flows and required more tangible assets.


Although the increased hoarding of cash may be a common trend, there is more to this issue than meets the eye.

Links, with commentary

1. Some people might think that Bryan Caplan is ranking how reasonable people are, but he is actually ranking how many of the same priors these people share with him.

 Felix Salmon points us to a fund that is designed to go long tail risk, the kicker is that they expect to lose 12% to 18% in non-tail event years. It is unclear how much the truly provides diversification benefits for hedge fund investors, since its correlation with funds trying to make steady returns via the opposite method may be too close to negative one.  At the very least, it may hedge them against the tail risk in funds that they don't know are producing a majority of their returns via the opposite strategy.

3. Scott Sumner uses ideas similar to Krugman's reply to the ECB's pro-austerity positions to critique Krugman's uncritical examination of the Eichengreen, et al, study on depression multipliers.  Since that is one too many links, the basic summary is that the ECB found time periods when fiscal austerity was associated with recovery. However, Krugman believed that these examples of austerity working occurred in tandem with either a shirt towards trade surpluses or a weaker monetary policy. The ECB didn't control for these positive factors when discussing how austerity might help, since the whole world can't go into a trade surplus nor can they devalue their currency against everyone else.  However, Sumner points out that the same critique of ignoring monetary stimulus applies to many of the depression multipliers calculated in the Eichengreen, et al, study.  

The combination of the ECB view and Krugman's critique makes a good case for a compromise of monetary stimulus and fiscal responsibility. The main problem with this approach is that it doesn't serve the interests of either major party in any way. The democrats would rather find arguments for stimulating the economy with spending, and the republicans would prefer tax cuts. Furthermore, the republicans are very suspicious of any form of inflationary monetary policy, as they correctly recognize that it is a tax on capital while the democrats would worry that pushing for monetary stimulus, which will raise asset prices, will make them look like friends of the rich on Wall Street.

4. Greg Mankiw has an article in the New York Times on the Trilemma of International Finance. The idea is that a country can have two of the following: Openness to flows of international capital, monetary policy as a tool to stabilize the economy and stability in the currency exchange rate. 

Three major players made three different choices:

US: Openness and monetary policy
China: Monetary policy and currency stability
Europe's monetary union: Openness and currency stability (The Bundesbank.. I mean ECB has until recently only been there to preserve price stability, as opposed to the Fed's mandate to "to promote effectively the goals of maximum employment, stable prices, and moderate long-term interest rates.")

Due to these fundamental different economic choices, when there is dialogue between these countries they are often talking past each other because they are assuming that their choice is the optimal one.  However, when the ECB got involved in the Greek bailout and the preemptive bailout of the other PIIGS they moved towards the US's choice. If China starts to float their currency, they'll move towards either Europe or the US, although that seems like it will be a very slow process.

The Jobs Picture

Most people know that private sector jobs have been increasing, although recently at a slower rate than expected.

The question is where new jobs are going to come from. In order to get a clear picture, we need to figure out how jobs have been lost.  The Business Employment Dynamics (BED) survey from the Bureau of Labor Statistics (BLS) has some good data, though it is only quarterly and updated through September 2009. The y axis for the below charts are in thousands of people.

The drop in job creation has occurred due to fewer expanding companies.

The main source of job losses has come from company downsizing as opposed to a spike in bankruptcies.

The BLS’s Job Openings and Labor Turnover Survey gives more recent data (Up to April), and sheds a little more light on where the job losses are coming from. Job gains have occurred because separations have trended downwards, while hiring has only picked up marginally.

Job opening data is a little bit more optimistic.

However, the data for separations is a little less optimistic, the overall job picture is only looking good because people aren’t quitting their jobs at a normal rate. If quits were at normal levels then the private sector would be losing jobs.

All of these charts would look a little bit different if looked at in the rate space, since the population has been increasing the numbers would all look a little less significant.  On a ten or twenty year basis, looking at levels isn’t going to skew the results by more than 10%, but when looking at time frames over a fifty year period, using levels rather than percentages significantly exaggerates the situation.

IQ, patience & risk aversion

The more intelligent a person is, the less risk averse and more patient they are. The study, "Are Risk Aversion and Impatience
Related to Cognitive Ability?" done by Thomas Dohmen, Armin Falk, David Huff man and Uwe Sunde found a significant relationship between these variables when they surveyed a thousand German participants drawn from the general population. 

If this relationship holds across countries and on an aggregate level, it would suggest that countries with low average IQ's have additional barriers to overcome.  It might even suggest that countries with lower average IQ's implement policies of forced savings to increase the focus on the long run, though given how corrupt the governments of these countries are any policy that actually gets implemented is likely to be inefficient at the margin.

(No HT for the study, because I forgot where I found this study.)

The US should do something like this again

Tyler Cowen links to a planned mass medical study in Germany.

German scientists are planning the country’s biggest biomedical study. The National Cohort will be an intensive investigation of the health, lifestyle and genetics of 200,000 people, at an estimated cost of €210m over 10 years.

He notes that they don't mention plans to study the efficacy of health care access, but it could still be studied.  In fact, it could be better if they don't directly study the effect of health care access and instead merely control for it while looking at other things such as the impact of exercise on health.  This way, they will be more likely to throw some variables representing treatment frequency into the regression (At the very least, there is a subgroup of 40,000 receiving more detailed attention) and we can see whether or not the variable is significant or even the right sign.  It is better that they don't study it directly because if they studied the impact of health care access directly, they might be more prone to try and jigger the regression until the variable had a sign and significance that fit with their preconception.

Less Temporary Workers = More Recalculation

The Cleveland Fed's Murat Tasci published a commentary called "Are Jobless Recoveries the New Norm?" in March 2010. This paper adds some weight to those on the side of structural change in the recalculation debate, the idea that a large part of the reason that the economy is slowing down is because resources are improperly allocated.  If this is true, it means that Keynesian stimulus will be less effective.

Another statistic suggests that some of these unemployed workers will not be able to immediately find work: the fraction of unemployed workers who have been laid off temporarily. Traditionally, some employers lay off workers with an implicit (or explicit) understanding that they will be recalled when the economy improves. Temporary layoffs save both parties from having to spend time and effort searching for a job or a worker when the need for labor increases. Usually, the fraction of the unemployed who are on temporary layoffs jumps at the beginning of recessions. Currently, this measure stands at around 10 percent, but it did not increase much during the recession; if anything, it fell slightly. This implies that fi rms do not perceive the employment adjustment they are going through as temporary or that conditions are such that they are willing to incur costly hiring efforts after the recession.
The low levels of temporary layoffs suggest that the skills of some of these unemployed workers are specific to industries that may end up significantly smaller after the recession.

The more people ignore that there is structural economic change, the worse the policy is going to be.  The simple idea of re-inflating aggregate demand doesn't work to the extent that the recovered economy is going to be different than the pre-crash economy, and the efforts of private companies to capture new fiscal spending or figuring out what the new regulatory structure is going to look like is effort not spent towards actually becoming more productive.

Over extrapolating bad news?

There has been quite a few disappointing US economic data point recently - consumer confidence is down and the total amount of job creation in the private sector is disappointing relative to economist's models.  This type of data leads a lot of people to talk about a double dip recession. While this is possible, it may be that this is merely the impact you get when part of the country is hit by an environmental disaster - the job growth is a little more anemic and consumer confidence goes down.  On June 9th, the Atlanta Fed had this to say:

On balance, Sixth District business conditions appear to have improved modestly in April and May.


However, the recent Gulf oil spill and the floods in Tennessee have tempered the outlook in those areas. Contacts indicated that the potential impact on the tourism industry along the coastline of Louisiana, Mississippi, Alabama and western Florida could be substantial. In some cases, vacation lodging cancellations have been replaced by bookings from clean-up crews, laborers, and the National Guard. The Nashville area is expected to see a decline in tourism-related receipts because of damage to several tourist venues there. The near-term outlook among hospitality contacts varied greatly, reflecting the high level of uncertainty.

Since regional labor data lags by a few months, the next Beige book, released on July 28th, might help clear things up a little more.  There is no question that the public at large is over estimating the spill's economic impact and it is also true that the US has some long term structural economic problems that have yet to be resolved, but it would be interesting if the financial markets are over extrapolating by looking at the temporary deviations caused by this disaster and assuming that they are a sign of a systematic problem.

Convenient Keynesianism?

The real time economics blog had a post last week previewing the G20. It had an interesting quote about Obama's views.

Mr. Obama talked of the need to “restore sustainable public finances,” but he pushed that off to the “medium term” — which means three to five years from now. There was no mention of starting by 2011, as the Europeans want.

Waiting for three years means not cutting back on stimulus until 2013 at the earliest. This is good for him because even a short term economic slowdown, no matter how necessary, can ruin a reelection campaign. This isn't to say that Obama doesn't truly believe that this path is best for the economy, but with the health care mandate not occurring until 2014 I am beginning to sense a pattern.

CBO's Long Term Budget Outlook: Revenue will be less than spending

The CBO released their Long-Term Budget Outlook for 2010! This may be a weird thing to get excited about, but I was looking through their 2009 outlook and wishing that I could work with an updated version.

Key quote:

Keeping deficits and debt from growing to unsustainable levels would require raising revenues as a percentage of GDP significantly above past levels, reducing outlays sharply relative to CBO’s projections, or some combination of those approaches. Making such changes while economic activity and employment remain well below their potential levels would probably slow the economic recovery. However, the sooner that long-term changes to spending and revenues are agreed on, and the sooner they are carried out once the economic weakness ends, the smaller will be the damage to the economy from growing federal debt. Earlier action would require more sacrifices by earlier generations to benefit future generations, but it would also permit smaller or more gradual changes and would give people more time to adjust to them.”

Key chart:

In order to make sense of the above chart, the difference between the extended-baseline scenario and the alternative fiscal scenario needs to be understood.  The extended base-line scenario assumes that taxes and revenue change as is expected in the current law.  The alternative fiscal scenario assumes things that are likely but are not scheduled under the current law, such as no Medicare physician payment cuts and the AMT tax relief (See Table 1-1 in the report for more details).

My main complaint about this approach is that the least likely aspect of alternative fiscal scenario is also one of the most important: that it assumes that the tax cuts from 2001 and 2003 are not going to be allowed to sunset. This can be partially corrected for by using the revenue from the extended baseline-scenario and spending from the alternative fiscal scenario.  We find that the primary deficit (Deficit before including interest payments) will be -1.4% in 2020 and -3.1% of GDP in 2035.  Incorporating in the interest payments from the alternative baseline scenario (The actual interest payments will be higher due to a larger debt build up), that means that the total deficit will be -4.5% in 2020 and -7% in 2035.

I’ll end with a CBO table that informs us that the sooner the deficit is brought under control, the better: