Why not treat debt and equity the same?

by on December 31, 2011 at 2:39 am in Economics, Law | Permalink

Varun, a loyal MR reader, asks me:

I do have a fairly simple question on tax policy I’ve never really seen a good answer to:  Why do we treat interest payments differently in terms of taxation? Why are interest payments tax deductible?
Clearly a zero corporate tax rate is best, but why do we offer tax shields for highly levered companies? All of private equity, and much of banking etc. is built on this tax arbitrage. Wouldn’t treating interest payments on par with dividends and corporate profits (hopefully at a lower tax rate) unlock a great deal of value, drive an increase in (stock) investment, while significantly un-levering businesses? Why do we borrow when we can seek investment?
More importantly, isn’t it odd that few advocate such a simple policy change: to treat debt and capital investment identically.
A good question, but there is a problem with treating debt payments any other way.  In general, expenses must be deductible in some manner, if the government is to tax corporations on net rather than gross returns, however roughly or imperfectly.  And it is difficult not to treat interest like an expense of some kind.  For instance de facto interest could be embedded in repurchase agreements, which for the purposes of tax law would look more like “real expenses” and thus would be tax deductible.  The borrowing would still go on, but in a more awkward fashion.
Without tax deductible interest payments, there would be an excess incentive to pay cash up front for assets rather than doing a mix of borrowing and holding cash for option demand.  Corporations would go bankrupt more easily and in general face higher transactions costs.

Contrary to common impression, the tax deductibility of interest payments does not give a tax advantage to borrowing, not if the return to savings is taxed.  What you save by borrowing and writing off interest payments you pay back tax on your more liquid asset holdings; admittedly there are complications and wedges when lending and borrowing rates are not the same.  Therefore tax-deductible interest payments makes tax law roughly neutral in intertemporal terms, with lots of qualifications tacked on to that claim, including the possibility that some corporations can avoid the taxes on liquid asset holdings altogether.

The tax deductibility of interest payments operates in a highly imperfect manner, but at its core it is a piece of what an ideal (roughly) neutral tax system would look like, not a deviation from such neutrality.

1 Leigh Caldwell December 31, 2011 at 4:19 am

In theory, the tax deducted by corporations on interest payments is paid on interest income by the lender.

In practice, I suspect this is often avoided by lending via overseas vehicles or possibly by extracting the interest income in other ways (eg compensation to the employees of lenders – which is of course taxed at the next stage, but avoids the double taxation that would arise if banks made money from dividends instead of interest).

As Tyler says, it’s hard to see how to avoid this without introducing worse distortions – because taxing corporate profits is itself a distortion to an ideal tax system (though politically inevitable).

2 ohwilleke January 1, 2012 at 9:12 am

“taxing corporate profits is itself a distortion to an ideal tax system”

Taxing distributed corporate profits is a distortion, taxing retained corporate profits is not.

3 NAME REDACTED December 31, 2011 at 4:31 am

It still seems though that the system favors debt over equity pretty heavily. This is very bad, as debt is far more fat-tailed than equity.

4 Rahul December 31, 2011 at 6:45 am

Isn’t the average debt-to-equity ratio in the American corporates around 0.5; that doesn’t seem too high; or is it? What’s a “natural” ratio (Bahamas?); or what’s a suitable benchmark?

5 Willitts December 31, 2011 at 8:36 pm

Debt issuers probably exert more influence in favor of corporate well-being than the agency problems associated with equity.

You ask a good question. Of course the optimal debt-equity ratio would vary not only by country, but by industry. There are also strong considerations of corporate control.

An interesting accounting anomaly is that when a corporation’s debt is downgraded, the firm books a gain. This is because they can notionally repurchase their debt at a lower price. Of course, an upgrade has the opposite effect.

6 NAME REDACTED January 1, 2012 at 8:07 pm

And you incidentally point out another problem with debt issuance.

7 Nathan Tankus December 31, 2011 at 4:33 am

“Without tax deductible interest payments, there would be an excess incentive to pay cash up front for assets rather than doing a mix of borrowing and holding cash for option demand. Corporations would go bankrupt more easily and in general face higher transactions costs.” woah where did that jump come from? how would an incentive to be less leveraged result in more bankruptcies. the last time i checked bankruptcy is defined as the “inability to discharge all your debts as they come due”. so having an incentive to be less indebted will somehow cause more bankruptcy because of inability to pay debts? without more explanation I’m highly skeptical.

8 Jay December 31, 2011 at 9:05 am

The cost of equity is always greater than the cost of debt. If you push companies out of debt and into equity you will increase firms’ weighted average cost of capital. When the cost of capital goes up the hurdle rate goes up. When the hurdle rate goes up, companies will be forced into higher risk investments that exceed the hurdle rate. Higher risk investments by nature have more volatile returns. More volatile returns will lead to more bankruptcies.

9 Tom Grey December 31, 2011 at 9:22 am

“Corporations would go bankrupt more easily and in general face higher transactions costs.”

Higher transaction costs are quite debatable, but certainly more difficult expansion leads to less bankruptcy.
Companies will never be “forced into higher risk investments”, there will instead be a different calculation of the after-tax costs of financing the investments.

” the tax deductibility of interest payments does not give a tax advantage to borrowing, not if the return to savings is taxed.” — I flatly do not believe this, nor the explanation.

Besides the “no free lunch”, the other key truth of economics is that “incentives matter”. In two similar countries, one with interest payments tax deductible and one without, the one without will have less, but not no, interest costs. There will likely be less bankruptcies, and perhaps less growth and less experimentation — but less growth is not the argument made.

I don’t believe the neutrality argument. Below the issue of Estonia is interesting. The HELOC issue of borrowing at high rates because of the tax deductibility is a problem of the decision maker wrongly calculating the net costs. But the reality of such folk existing, and making such mistakes, clearly shows “too much borrowing”, not neutrality.

How many mistaken over-borrowers need there be before the reality of too much borrowing, partly based on wrongly calculated deductibility, invalidates the theoretical neutrality?

10 ohwilleke January 1, 2012 at 9:17 am

“The HELOC issue of borrowing at high rates because of the tax deductibility is a problem of the decision maker wrongly calculating the net costs.”

Not so much. The case for this activity being driven by the presence or absence of recourse lending in a particular state is much more compelling. The housing bust and overleveraging phenomena was a regional one strongest in effectively non-recourse states like California and Florida, and much weaker elsewhere. The borrowers weren’t math challenged, they were rational in a moral hazard kind of way and passed risk to finance companies that passed it in turn to mortgage backed security investors who received inaccurate disclosures of risk.

11 NAME REDACTED January 1, 2012 at 8:08 pm

Texas is nonrecourse and didn’t have a housing boom-bust. I don’t agree with your theory.

12 Franco Modigliani December 31, 2011 at 11:18 am

Wrong

13 Commenterlein December 31, 2011 at 11:26 am

^ Thank you.

14 Willitts December 31, 2011 at 8:52 pm

I was about to reply with a discussion of Modigliani-Miller. Your answer was more succinct than mine. 🙂

15 NAME REDACTED January 1, 2012 at 8:14 pm

Modigliani-Miller is simply wrong, because almost none of its assumptions hold, and thats just one of the problems with it. Its really pretty awful.

Here is a more succinct difference. No one ever went bankrupt by using equity financing, but plenty have gone bankrupt from using debt financing. This suggests that there is a categorical difference between the two.

16 Kyle s December 31, 2011 at 9:14 am

The argument is that rather than borrowing to finance a project while simultaneously holding cash, the corporation will not borrow but instead accumulate a chunk of cash and then invest it into a project without the cash cushion that leverage allows. In a downturn, all of the cash it would have had sitting on the sidelines that would have been able to finance a period of losses is now tied up in a factory purchased unlevered, which means that the corporation has fewer resources available to it to withstand a downturn.

17 Tom Grey December 31, 2011 at 9:25 am

BUT — with much less debt, in a downturn with low interest rates (like the last 3 years), the corporation should be able to borrow, when needed. To save good investments.

Or, sell the factory at a loss and eat the loss, if it was a bad investment.

18 D December 31, 2011 at 11:23 am

Heh. The only time a bank will lend you money is when you don’t need. I kid…

19 Tiedemies December 31, 2011 at 4:40 am

From the individual and single firm point of view equity and debt look very different, but is there actually any real difference at a higher, i.e., macro level? OK, so if you hold stock instead of a bond, you treat the expected income differently and make different decisions based on what you expect it to pay, but all in all, both are just claims on future income of the company. I don’t really see any spectacular cause for alarm in the fact that companies are highly leveraged, unless you have built everything on the assumption that the debt is foolproof. The riskprofile of debt just starts to resemble that of equity with some sort of repurchase scheme, when a company is highly leveraged.

20 Nathan Tankus December 31, 2011 at 6:31 am

the problem is your looking at it from an investor’s and not a firm’s point of view. you are guaranteed no amount of money by the holding of a stock. the magic of compound interest doesn’t apply here. the magic of compound interest does apply to loans and (the rolling over of) bonds.

21 derek December 31, 2011 at 12:40 pm

Any matter is tested when it comes to failure. A decrease in equity values is not nice, but a decrease in the ability to pay debt is catastrophic. Debt is an asset to lenders, so failure of debt creates a cascading effect. Equity is a stake as owner, debt on cash flow.

It would make sense to structure incentives for more equity as opposed to debt simply for structural resilience.

22 Björn December 31, 2011 at 10:12 pm

The main diference between equity and debt is the distribution of risks.

In the case of debt, you are entitled to some amount of the company’s >>>revenue<<>>profit<<<. You as an investor will only resveice a return – in the form of a cpaital gain or dividend – if the company has a positive profit.The company is not obligated to pay you a return from the revenue it has. Therefore, the risks mainly lie at the level of the investor (hence the ighigher cost of equity, compared to the cost of debt (see above)).

23 Björn December 31, 2011 at 10:13 pm

Sorry, looks like someting went wrong there…

New try:

The main diference between equity and debt is the distribution of risks.

In the case of debt, you are entitled to some amount of the company’s >>>revenue<<>>profit<<<. You as an investor will only resveice a return – in the form of a cpaital gain or dividend – if the company has a positive profit.The company is not obligated to pay you a return from the revenue it has. Therefore, the risks mainly lie at the level of the investor (hence the ighigher cost of equity, compared to the cost of debt (see above)).

24 Björn December 31, 2011 at 10:14 pm

Hm, I guess it’s the > and < signs? Sorry!

Last try:

The main diference between equity and debt is the distribution of risks.

In the case of debt, you are entitled to some amount of the company's ___revenue____. The only risk you as a creditor have is that the company does not have enough revenue to meet it's interest obligations – in which case it will, ultimately, go bankrupt (and you as a creditor have a senior claim on the company's assets). Therefore, the risks mainly lie at the level of the company and not at the creditor (hence the lower cost of debt, compared to the cost of equity (see below)).

In the case of equity, you are entitled to some amout of the company's ___profit___. You as an investor will only resveice a return – in the form of a cpaital gain or dividend – if the company has a positive profit.The company is not obligated to pay you a return from the revenue it has. Therefore, the risks mainly lie at the level of the investor (hence the ighigher cost of equity, compared to the cost of debt (see above)).

25 Tiedemies January 1, 2012 at 12:34 am

I know what you are saying, and that is all correct, but it was not really my question
.
The structure of capital (equity vs debt) simply distributes the risk a little differently, but my question was, why should we be alarmed by any particular amount of leverage in the whole economy? Is there any reason to believe that more debt increases system risk somehow? As I said, if people at a large scale start to behave as if the debt they hold was as good as cash (as opposed to stock, the risk of which people often do seem to underestimate), that may be a problem. One other thing is if the debt comes from the banks like in some countries, in which case the risks may somehow be offloaded to people who have deposits at banks, but other than that, I do not see any macroeconomic problem here.

26 ohwilleke January 1, 2012 at 9:25 am

@Tiedemies. A tax bias in favor of more debt increases systemic risk in the economy, because debt payments to creditors are less flexibile than dividend payments (even preferred dividend payments which superficially look like debt payments). The policy issue in that case is that more leverage kills marginal businesses more efficiently during recessions which is both good and bad. Bankruptcies of businesses are bad when you are in the recession. But, there are opportunity costs after the recession is over to not having killed off the marginal businesses and thus not having encouraged resources to go instead to new ventures that have better long run prospects.

27 mulp December 31, 2011 at 4:59 am

The real question is “why are dividends not deductible like interest?”

The US tax on corporate profits was based on taxing where the money is. In 1937, undistributed earnings were subject to a surtax to promote distributing profits as dividends. This in part was to collect the unpaid individual income taxes.

Other nations have various means of not taxing dividends, with, for example, the Aussie ACT where the corporation pays taxes on all profits, but distributes the tax paid along with the dividend to shareholders to handle it like withholding tax, even getting a tax rebate if their tax rate is lower than the corporate rate.

The argument against taxing dividends is it “taxes” repayment of principle on debt. However, in the 30s, corporations that paid out most profit in dividends were more than able (before any dividend reinvestment run by computer systems) to raise more capital than dividends by selling new shares.

The primary reason for not paying dividends is the tax dodge of the current US system. Make dividends deductible and the corporate debt that leads way too often to bankruptcy primarily by LBOs but also by the kinds of deals people like Trump do will become much less attractive – buying stock will provide cash returns so buying junk bonds will be a much riskier way to get cash flow.

28 David N December 31, 2011 at 8:58 am

+1

Taxing dividend distributions certainly causes business to use more long-term debt financing than they might otherwise. It also causes public companies to hoard cash in excess of what is needed to fund the business and future growth. Public companies too often support their stock price with buybacks instead of issuing a dividend, and they usually overpay.

29 JWatts December 31, 2011 at 1:39 pm

“It also causes public companies to hoard cash in excess of what is needed to fund the business and future growth. Public companies too often support their stock price with buybacks instead of issuing a dividend, and they usually overpay.”

Precisely, many if not most stock buybacks are a worse deal than dividend payments would be. Furthermore the inevitable cash hoarding encourages companies to invest in other industries and companies they don’t have any expertise in. I’m convinced that stock holders would benefit by larger dividend rates. There are also several studies that indicate that, all factors being equal, stocks with high dividends tend to outperform stocks with low dividends.

30 ohwilleke January 1, 2012 at 9:11 am

FWIW, the moment that the debt-equity distinction really became meaningful was in the General Utilities doctrine was repealed in the 1986 Tax Reform Act. Prior to then, it was possible to retain earnings and do an end run around double taxation when the company was ultimately liquidated in a way that didn’t count as a dividend. The act eliminated the loophole and made double taxation real, and also coincides quite closely with the advent of the leveraged buy-out in which part of the gain from the restructuring transaction comes from reducing the aggregate long term taxation in the deal.

31 James Hass December 31, 2011 at 5:46 am

Corporate taxation includes a bunch of dubious decisions like depreciation treatment, inventory treatment, goodwill amortization etc. Estonia has passed on most of these issues by just taxing payments made to shareholders. The result seems to be an accumulation of cash and equity in firms that limited the number of bad loans in the recent exchange crisis next door in Latvia. Statistically significant and important effects were seern: http://www.eestipank.info/pub/en/dokumendid/publikatsioonid/seeriad/uuringud/_2011/_2_2011/_wp_211.pdf

32 TallDave December 31, 2011 at 6:57 am

Interesting, I’ve wondered how that would work out. You would think it would reduce the deadweight loss, if nothing else. Thanks for sharing.

33 TallDave January 1, 2012 at 1:08 pm

An excerpt from the summary:

“While the major intention of the reform was to increase investment, according to our estimations it also led to an accumulation of liquid assets as cash and equivalents. While higher levels of liquid assets may mean lower productive efficiency, we also found evidence that during the recent crisis in 2009 these high levels have
contributed to firms’ survival and to lower levels of non-performing bank loans. At the same time, also the achieved relatively lower debt financing of
Estonia’s firms has contributed to the firms’ survival during the crisis. The reform also demonstrated positive effects on economic growth through its
positive effect on investment and labour productivity, which was strongest in the services sector and among smaller companies.”

Maybe that’s actually a better tradeoff.

34 Sean Brocklebank December 31, 2011 at 6:52 am

Could someone please point me to a reference with a model of this (textbook or otherwise)? Thanks.

35 TallDave December 31, 2011 at 6:55 am

Contrary to common impression, the tax deductibility of interest payments does not give a tax advantage to borrowing

Yes, as a CPA this drives me nuts. I meet people all the time who carry ridiculous HELOC interest rates of 7%+ because they “like the deduction.” I try to explain they have to net out the expenses as well as the opportunity cost (i.e. borrowing at a tax-adjusted rate of 2.4% (say 4% at 40% marginal tax rate) and buying low-risk bonds that pay 4-5% may be a decent play, borrowing at a tax-adjusted rate of 5% or more is rarely a good idea).

36 Kyle s December 31, 2011 at 9:21 am

One reason to think that disallowing tax deductibility of interest might not change things much is islamic finance, where charging interest is prohibited by the Quran. Well paid attorneys and financiers have invented complex structures involving the sale and repurchase of precious metals at prices that just happen to correspond to LIBOR + 300 bps.

37 Willitts December 31, 2011 at 9:07 pm

Yes. More generally and conventionally, lease transactions of capital equipment can be structured to mimic a sale of equipment under debt financing with interest.

Operating leases should often be considered like a capital lease by an analyst to properly represent the amount of a firm’s debt.

38 Ted Craig December 31, 2011 at 9:41 am

Isn’t this because the government pays interest rather than dividends? Most rules favor the rulemakers.

39 ohwilleke January 1, 2012 at 9:33 am

Governments don’t pay income taxes. And, until the GWB tax cuts, both interest payments and dividend payments were ordinary income when received by the investor (with the exception of muncipal bond interest which is tax free and a defacto subsidy to the municipalities).

Actually, at the federal level, the federal budget deficit would be smaller if federal bond payments were tax free the way that municipal bonds are. Municipal bonds involve indirect taxation because their interest rates are discounted to reflect the tax savings. Federal bonds are risk free returns that pay the actual risk free interest rate (pure time value of money and inflation expection). Domestic federal bond investors pay part of the interest they receive back to the feds as federal income taxes, but foreign federal bond investors (like China) don’t pay federal income taxes back to the federal government on the money they receive from the U.S. government for interest and aren’t indirectly taking an interest rate discount that reflects tax benefits of a tax free return. If federal bonds were tax free, the reduction in the interest rate that the U.S. government pays due to market effects would be greater than the foregone tax revenue. The incidence of the policy change would also fall predominantly on foreign bond investors.

40 K December 31, 2011 at 10:00 am

In fact, the Belgian corporate tax code has a deduction for imputed interest costs on equity, in order to get rid of the unequal tax treatment. It’s called the “notional interest deduction” (http://minfin.fgov.be/portail2/belinvest/downloads/en/publications/bro_notional_interest.pdf).

41 Yossarian December 31, 2011 at 11:25 am

I’m no expert on tax policy- find the whole subject matter quite confusing, actually; but it doesn’t seem prudent for an economy to charge a progressively higher tax rate to companies as their ROE’s rise- why would we want to discourage companies that earn high returns from investing?

42 ohwilleke January 1, 2012 at 9:36 am

The idea is to impose taxes in a way that least alters the economic incentives in the absence of taxation. So long as you succeed in that in a rough justice way, life is good and it doesn’t matter much from an efficiency perspective who actually pays how much. The a slightly imperfect incentive dramatically reduces administration and compliance costs (which are huge in the U.S. system) then is can still make sense.

43 Varun December 31, 2011 at 3:18 pm

K,

Thanks for bringing this to my attention. This is a fairly clever mechanism – I’m surprised it originated in Belgium! Have you seen any studies on the impact it had on investment in Belgium before/after?

Of course setting the “notional” return rate on equity is subject to a lot of potentially dangerous government fiddling, but if nothing else, its yet another mechanism for governments to reduce the corporate tax rate in a clever way.

44 Patrick January 1, 2012 at 3:53 am

Australia’s recent report on tax systems by the Treasury recommended just that (a.notional equity deduction – a straight deduction for dividends is problematic as you want to incentise holding cash!) and (I think) even considered variable ROE taxation. The IMF in a post-Lehman report also recommended notional equity deductions.

In fact something of the sort has already been introduced in Australia, following the ‘Henry Report’ I referred to above, as the Minerals and Resources Rent Tax which, although supplementary to corporate profits tax does provide a deduction for a notional ROE.

There is some support for making this broader. Coincidentally, or not, almost all tax under such a system would be borne by Resources and financial taxpayers.

I will try and provide links later.

45 [Insert here] delenda est January 3, 2012 at 6:34 pm

And the most accessible link to the report recommending an allowance for corporate equity: http://taxreview.treasury.gov.au/content/FinalReport.aspx?doc=html/publications/Papers/Final_Report_Part_2/chapter_b1-3.htm

46 [Insert here] delenda est January 3, 2012 at 6:18 pm

In addition, Italy has now passed such a mechanism, applying to equity from 1 Jan 2012 (i.e. new equity since then not existing equity).

The initial rate is a 3% deduction.

47 ohwilleke January 1, 2012 at 9:43 am

A similar proposal is to impose a property tax on the average market value of publicly held securities in the time period in which those property taxes are collected, rather than a tax on corporate profits per se. Much harder to evade, much easier to determine.

48 byomtov December 31, 2011 at 10:16 am

Contrary to common impression, the tax deductibility of interest payments does not give a tax advantage to borrowing, not if the return to savings is taxed. What you save by borrowing and writing off interest payments you pay back tax on your more liquid asset holdings..

But suppose the company wants to finance real investment, not buy bonds. Then, it seems to me (and to M&M?) that there is a tax advantage to financing with debt, at least from the corporation’s POV, up to a point.

49 Martin Brock December 31, 2011 at 10:58 am

Here’s a related question. Why don’t we pay sales tax on rents (and interest)? If you sell me a car for an indefinite period of time (you transfer the title to me), the transaction is taxed, but if you sell me a car for a definite period of time (you rent the car to me), the transaction is not taxed similarly. Why?

Of course, the short answer is “statesmen want it this way”, but that’s not the answer I’m after. Why do statesmen want it this way?

Couldn’t we avoid all sales taxes by renting everything? Instead of selling me a car, you rent it to me indefinitely. The rent is $200/month for 60 months and a penny per century thereafter. For a gallon of milk, the rent is $3 for the first minute and a penny per century thereafter. I’m not a tax attorney. Presumably, tax law already anticipates this dodge.

50 Lance December 31, 2011 at 4:11 pm

They do. If the IRS views the transaction as a sale, they’ll treat it as an installment sale. If the transaction passes certain tests (economic viability test, bargain purchase option, worthlessness of item at end of lease (I assume this would apply to your milk example!), etc) it will be called a sale despite the form of a lease.

51 James b. December 31, 2011 at 4:13 pm

I pay tax on rentals all the time.

Today I picked up two movies from Redbox: Taxed

I got a car from Avis in Miami earlier this month: Taxed

I rented a pump to clear the water out of my basement: Taxed

52 Martin Brock January 1, 2012 at 8:57 am

You have my sympathies.

53 ohwilleke January 1, 2012 at 9:41 am

False premise. Some places have sales taxes on rents. Some places don’t. Most states do impose sales taxes on automobile lease payments for example and on any other transaction that would be sales taxable if structured as a sale. Most don’t impose sales taxes on rental real estate payments but go after that tax base via property taxes instead for adminstrative convenience and making it harder to evade taxation reasons and for collectibility reasons (property taxes are very easy to collect in the absence of a massive depreciation in property values since they have a first lien on the real estate and the real estate can’t run away).

54 MGM Mirage December 31, 2011 at 10:59 am

…including the possibility that some corporations can avoid the taxes on liquid asset holdings altogether.

I’d say that’s quite a distinct possibility! Maybe even the norm?

I’m fascinated by the ways that the US tax code and taxpaying corporations bend over backward to apply taxes only to net earnings. There should be takes on cash reserves and gross revenue, with no deductions. Then we could get *really* angry at how congress spends money. But seriously, if that were the system, even self-styled libertarian CPA bloggers would demand single payer health care.

55 JWatts December 31, 2011 at 3:14 pm

“There should be takes on cash reserves and gross revenue, with no deductions.”

Only a few seconds of thought would have led to the realization that this is a preposterous proposition. Any taxes on gross revenues would destroy the retail industry. Car dealers and grocery stores are two examples of high gross revenues but small profits. For example: Kroger’s has a Profit Margin of less than 1%.

56 Floccina December 31, 2011 at 11:57 pm

A tax on gross income is a sales tax and would be paid by the purchaser.

57 ed December 31, 2011 at 11:02 am

” the tax deductibility of interest payments does not give a tax advantage to borrowing,”

You may be right in some sense, but that’s not what we’ve been teaching our MBAs, last time I checked. If the people running the companies think there is a tax advantage to borrowing, won’t they borrow excessively?

58 Willitts December 31, 2011 at 9:10 pm

It depends on how many shares and stock options the “people running the company” have and how much influence they have over the Board of Directors. 🙂

59 ohwilleke January 1, 2012 at 9:45 am

If lenders will let them.

60 RobertoT December 31, 2011 at 11:29 am

In Italy, the new Government led by Mr Monti has recently introduced an allowance for new equity injections: basically, corporations can deduct the theoretical “cost” of new equity paid into the company, similarly to what happens for interests. The notional rate at which this cost is calculated is set by the Government (now it’s 3%).

61 Varun December 31, 2011 at 11:31 am

Thanks for addressing my question Tyler.

I like the answer, but I’m not sure how much I agree with it. The point, as someone mentioned in the comments, is not are interest expenses taxed or not, but how differently they are treated from dividend payments. Currently dividends are taxed twice, and by Tyler’s argument interest expenses are taxed once (at best) and at a very different rate.

I don’t think anybody would disagree that the current system is distorted in that borrowing is overly incentivized, and equity overly disincentivized. Since people in general like taxing corporations, can we retain that level of firm disincentive (because that is harder to change) but fix the equity/debt distortion?

If we treat interest payments symmetrically with dividends, it seems hard to believe that there will be an “excess incentive to pay cash up front for assets”, rather it seems the current distortion to avoid up front payment would end.

The point about higher transaction costs seems to be a short term view. Of course when the financial system is so structured towards efficient borrowing, a change in behavior towards less borrowing would have higher transaction cost, but is it hard to believe that the 5-20% of GDP that we spend today on facilitating corporate borrowing would not be diverted towards reducing transaction costs on equity like vehicles?

As has been pointed out in the comments – I’m not sure I buy the higher bankruptcy rate. The typical bankruptcy is driven by long term debt, not by payroll. Making borrowing 30% more expensive doesn’t make it impossible. A good chunk of why WACC is low with borrowing is the effect of the tax shield. I’m puzzled as to why a company financed with debt or only equity in a world in which debt and equity are treated symmetrically would chose projects with different risk levels.

Thanks to K and James Hass on the pointers towards such policies in Estonia and Belgium. I’m looking forward to seeing estimates of the impact of such policies.

Finally – I think Tyler’s best point holds to an extent, which is that as long as the tax code is fairly complex you can come up with other ways to make interest payments tax free. This is a great and correct point. The question is, would fixing one part (interest payment tax deductability) reduce (and by how much) the distorted incentive for borrowing?

So – why no advocates for this? It can’t be as Tall Dave amusingly suggests because the government pays interest.

62 Rahul December 31, 2011 at 12:21 pm

In order to be sure we have a debt/equity distortion do we have any nations to compare to? Is the ratio really very taxation elastic? American corporates seem to be getting debt heavier over the last 5 decades:

http://2.bp.blogspot.com/-Rv1NONloeaY/TouOpFYcx1I/AAAAAAAAEeA/TqRyL9IGf-I/s1600/debt-to-equity-ration-us-corporations-1951-2011.png

Has relative taxation getting more debt friendly over the years?

63 Varun December 31, 2011 at 1:37 pm

A distortion is a marginal incentive no? We should be able to calculate the tax burden on a given (say wholely owned) company given two different balance sheets, 100% equity or 50% equity/50% debt, and account for all of Tyler’s nuances. At the end of the day, the company that is 100% equity financed will have a higher tax burden on its sole shareholder.

I think the story across time in America is that we’ve gotten better at managing to higher debt levels and gotten more comfortable with them. Private equity levels of leveredness was mostly unthinkable in the 60s and 70s (partly at least due to higher interest rates of course). Leveredness in PE has actually come down significantly from its hay day even in the 2000s.

64 ohwilleke January 1, 2012 at 10:26 am

The under the line v. over the line transition corresponds to the 1986 tax reform that strongly favored debt over equity relative to prior law. The dip below the line again lines up with the reduced tax rates on qualified dividends. The gradual rise in debt to equity from 1951-1986 was probably not tax driven; there is nothing in tax law that has changed gradually over that time frame; that is probably real economy driven in some respect or another. It may reflect the absence of funds to invest in the economy after first the Great Depression and then WWII accompanied by very high marginal rates on high incomes, with money gradually leaking out of corporations and being reinvested in corporate bonds gradually over time as income inequality accumulated. A lot of the actual economic growth in that time period came from productivity growth and expansion so it wasn’t necessary that critical to be able to obtain financing for established big businesses in order to have sufficient funds to reinvest.

65 ohwilleke January 1, 2012 at 9:56 am

“The typical bankruptcy is driven by long term debt, not by payroll.”

A footnote on that point. The traditional post-WWII system of employee compensation in Japan is for the employee to get something on the order of 80% of their annual pay in the form of regular pay packets and the balance in an annual bonus. There is strong corporate culture pressure to pay the bonuses if there is an ability to pay them, before paying equity investors very much, even though it isn’t a court enforceable obligation. But, if some surprise economic shock to the firm comes up, bonuses are greatly reduced and the company lives and the workers aren’t laid off. It basically ties lifetime employment expectations for risk sharing on the part of employees.

Also, in actual point of fact, a great many bankruptcies are payroll driven, especially in the “real economy.” For example, in the LTV bankruptcy (the biggest steel industry bankruptcy that was a bellwhether of the industry dying almost entirely), the company bargained with lenders and kept the doors open until the week before it couldn’t make payroll anymore, and that isn’t atypical. Usually rational lenders are going to be willing to renegotiate terms with companies that are sure to be able to make payroll and pay trade creditors because if they don’t, a Chapter 11 will force them to accept those terms anyway. More generally, and univerally, almost all bankruptcies are driven by insufficient ability to meet current cashflow demands, not be balance sheet insolvency, even though a balance sheet insolvent company, in theory, can go bankrupt and shave debts (via a sale of assets to a new shell entity).

66 bill December 31, 2011 at 1:24 pm

Tax corporations like REITs. If they distribute the income, then don’t tax the entity. Tax the recipient on the income.

67 ohwilleke January 1, 2012 at 10:31 am

Closely held companies do that. The trouble is that you get phantom income, taxation without distribution, which can be a problem is the market for the shares isn’t very liquid for people who don’t have the liquidity to pay those taxes. REITs and mutal funds solve that by distributing all of their earnings, but corporations in the real economy need to retain earnings from an economic perspective to a much greater extent. You could have a REIT/mutual fund structure combined with a corporate level withholding tax on undistriubted earnings, but it is tricky to integrate that with diverse state and local taxation rules in enterprises that operate in many different states and shift the mix of their business in each state somewhat over time.

68 Bill December 31, 2011 at 2:18 pm

Re your comment: “Contrary to common impression, the tax deductibility of interest payments does not give a tax advantage to borrowing, not if the return to savings is taxed.”

Pretty big assumption here that that the “returns to savings is taxed”.

Ordinary income dividends are taxed at a 15% rate, and not at the rate of the shareholder recipient (and, before we get into the debate about double taxation, please note that this change in the Bush tax code in 2001 originally had a limitation that if the corporation paid taxes on the income, the taxes would be subtracted from the dividend–but, guess what happened, many companies didn’t pay much taxes, so the dividend would have been taxed at the taxpayers higher marginal rate–so they opted for the 15% number.

Capital gains are taxed at 15%, so a corporation can buy back stock and the stockholder sees capital gains, and is taxed at the cap gains rate.

We can talk about stepped up basis in estates and how that might relate to realizing capital gains income in appreciated property rather than as distributions of income through dividends.

So, is there a preference under the tax code for borrowing. Yes, partly because there is an effect on the shareholder or stockholder’s OI 15% income and realization through capital gains rates.

Now think of this: there was a time when the US was a creditor nation, not a debtor nation: today, as a debtor nation, in effect borrowing money from China to purchase houses via Fanny and freddy, perhaps we should consider how our tax code tilts in favor of borrowing and how it should be neutral, so we will reduce the artificial incentive to borrow we have todayin order to reduce taxes at the corporate or individual level. Favoring borrowing is not a good policy for a debtor nation–and I am using debtor nation to cover not just government, but also business and personal consumption.

69 Bill December 31, 2011 at 7:51 pm

I would also add that tax deductability of interest incents higher income individuals to purchase larger residences because the interest deduction is more powerful at marginal rates. One solution would be to have a fixed housing credit, available to renters and owners, thereby discouraging competitive consumption of high income earners for Veblenesque competitive consumption supported by the tax code. Or, you could cap the interest deduction.

70 ohwilleke January 1, 2012 at 10:00 am

The interest deduction is capped (and even phased out at higher incomes). But, the intuition of your comment is basically right.

The interest interesting thing about the residential mortgage interest deduction is that it and the state and local tax paid deduction account for almost all of the regional variation in income taxation relative to nominal income. These are basically the only regional cost of living factors in the code for ordinary individual taxpayers.

71 Bill January 1, 2012 at 1:24 pm

Just so the readers know, the cap on residential interest deduction: interest is deductible on the first $1 million of debt used for acquiring, constructing, or substantially improving the residence, or the first $100,000 of home equity debt regardless of the purpose or use of the loan. That’s a pretty big cap.

72 ohwilleke January 1, 2012 at 2:33 pm

In Toledo it is a big cap. In Manhattan or San Francisco it is a pretty modest cap.

73 Bill January 1, 2012 at 4:39 pm

Ohwilleke, I think if the interest rate deduction were lower, the prices of high end properties would be lower. There would be a greater sensitivity to prices for the person with the high marginal rate.

74 Bill January 1, 2012 at 5:41 pm

For the reader, here is a list of US median house prices by US cities. http://www.realtor.org/wps/wcm/connect/4cb3128048fdc911a3eeef2e39654e23/REL11Q3T.pdf?MOD=AJPERES&CACHEID=4cb3128048fdc911a3eeef2e39654e23 The median price in SF is in the $500k range.

75 Sean Brown December 31, 2011 at 8:35 pm

I think Tyler is implying that hard-asset intensive companies basically NEED to borrow to pay for power plants, bridges, etc. The alternative is for “capital cos” – think GE Finance with planes – to own assets and “operating cos” to expense the implied interest as a capital lease instead. Do we really want deposit or commercial paper-backed institutions actually owning the hard assets directly and leasing them out to operators? IMO incentives for good LT performance and efficiency improvements – and hence wealth creation – are actually better when the operators own the assets themselves but are levered (think of John Malone’s cable companies over the years).

And think of retail stores. Some chains own their locations while others rent. Would it be fair to disallow deduction of a main occupancy expense (interest on the debt that financed the stores) in one case but not the other?

76 ohwilleke January 1, 2012 at 10:09 am

“Do we really want deposit or commercial paper-backed institutions actually owning the hard assets directly and leasing them out to operators? . . . Some chains own their locations while others rent.”

For tax and asset protection reasons ownership of hard assets is segregated from operating businesses in almost every circumstance when it is possible to do so. For example, retail chains very rarely own their stores (with one of the main exceptions being Kmart-Sears where the companies were bought by current ownership for the purpose of divorcing the two and got stuck with both because the commercial real estate market collapsed).

Lots of non-bank financial institutions, for example, insurance companies and pension funds, do directly own hard assets and lease them to the actual users of those properties. Commercial banking regulations are a bigger factor in this than non-tax, non-regulatory economics. Investment banks that invested on their own accounts routinely purchased long term illiquid assets with commercial paper. Ownership would probably make sense from a pure economics perspective in a lot of these cases (pretty much any case where you have a triple net lease instead of something closer to a residential landlord-tenant relationship), but tax considerations tilt the scales.

77 Sean Brown January 1, 2012 at 3:49 pm

Target owns almost all its locations, Nordstrom owns half, Kmart actually leases almost all. Retail runs the gamut though you’re right, REITs surely dominate asset ownership in the retail sector. Not nearly as true for indistrial cos though.

78 ohwilleke January 2, 2012 at 2:42 pm

Environmental liability risk may explain industrial co real estate ownership.

79 ohwilleke January 1, 2012 at 10:36 am

One other point. All Islamic finance transactions are structured that way. I prefer the Western system, but hundreds of millions of people conduct business that way sometimes on a very large scale, for religious reasons.

80 Sean Brown December 31, 2011 at 8:37 pm

Sorry, meant GE Capital.

81 R Richard Schweitzer December 31, 2011 at 8:55 pm

As Stanley Surrey said 60 years ago (quoting someone else): ” The the tax code functions to garner revenue, not as an exercise in logic.”

However, in underlying “economic logic” (in theory) it is the borrowed capital, not the operations of a business entity that is producing that portion of the returns represented as interest (which may or may not be earned as “net” – but sustains operations). Thus those returns on that capital are taxed to those providers.

Whereas (good lawyers’ term), the overall operations of a business, which may or may not use borrowed capital (that is entitled to separate returns) produces returns that are taxed to the business since they were not (in theory) resultant from “outside owned” capital. The privilege of participating in those operational returns is further taxed (to much grousing).

82 R Richard Schweitzer December 31, 2011 at 9:00 pm

If what I have just written seems silly, consider other ways of supplying “capital” owned by others, such as equipment, land use, or assets other than money or its equivalent in credits.

If the business entity uses those other forms of capital, its providers are separately compensated – even in partnerships, etc.

83 Björn December 31, 2011 at 9:48 pm

Thanks for this post Tyler!

Although your arguments in the first two paragraphs are convincing, you lost me at the third paragraph by stating that debt is not tax-favored.

In Europe many countries are actually considering to change their tax policies as to treat debt and equity more equally (i.e. by alllowing (a form of) dividend-deduction or by limiting the deductibility of interest). In Belgium they already introduced a ‘notional interest deduction’, which, in essence, allows corporations to deduct some percentage of their capital from their taxable income and also in the Netherlands – where I am from – the government is considering changes in this direction (an advisory commission of economists actually advised this to the Dutch government, with the (main) argument that tax-favoring debt distorts capital structure choices by companies). Moreover, there is a capital structure theory, the “trade-off theory”, which states that companies could favor debt because it is tax-favored (but only up to some point since more debt also makes the company more sensitive for financial distress – hence the trade-off).

You write:

“Contrary to common impression, the tax deductibility of interest payments does not give a tax advantage to borrowing, not if the return to savings is taxed. What you save by borrowing and writing off interest payments you pay back tax on your more liquid asset holdings; admittedly there are complications and wedges when lending and borrowing rates are not the same. Therefore tax-deductible interest payments makes tax law roughly neutral in intertemporal terms, with lots of qualifications tacked on to that claim, including the possibility that some corporations can avoid the taxes on liquid asset holdings altogether.”

As I see it, a company can finance it’s investments in three ways: i) debt, ii) equity and iii) retained earnings. And it can invest the attracted financing in two ways i) actual/non-liquid investments or ii) passive/liquid investments.

If I understand your arguments correctly, you argue that a company that uses debt will thereby save money because of the interest deductibility. Moreover, the company would not have saved this money if the interest would not be deductible. On the saved money, the company has to pay taxes, and, assuming interest and tax rates are the same (and no tax avoidance, etc.), this will take the tax-advantage of debt away.

However, why would the company not use the saved money for actual/non-liquid (positive NPV) projects? Or, more generally, wouldn’t a company chose the amount of cash it wants to hold irrespective of it’s capital structure choices? Especially since it’s not like the tax-saving from the interest payments comes as a surprise to the company.

Also, what assumptions are you making exactly? It seems like your argument would only hold if for every dollar of additionally issued debt, a company would increase it’s liquid assets with also one additional dollar (if the savings would increase with less than one dollar, then the additional tax on the savings would be less then the additional tax savings from the use of debt).

This however cannot hold. A simple example: a company borrows for one year $100 at 5% and the corporate tax rate is 30% and interest is tax deductible. By borrowing, the company has a tax saving of $1.5 [i.e. $100 * 5% *30%]. If it would save this money (i.e. not invest it in a real, active, non-luiqid investment), it would have to pay tax on it. However, it would pay the tax only on the $1,5. Leading to an additional tax claim of 30% of $1.5, being $0.45. Taken together, the tax saving is $1.5-$0,45=$1.05. So, debt would still be tax-favored.

One some what related question: what is your opinion on the ‘impliciat tax’ corporations pay on interest? The argument here is basically that because debt is tax-favored (which it, in my opinion, indeed is), the demand for debt increases, thereby also increasing the price for debt (so the interest rate). Theoretically, the ‘implicit tax’ (paid in the form of higher interest rate to the issuer of debt) should, in equilibrium, equal the ‘explicit tax gain’ (a saving in the form of interest-deductibility).

Hoping for a reply ;-),

Björn

84 ohwilleke January 1, 2012 at 9:07 am

Just plain wrong, in multiple ways.

“Without tax deductible interest payments, there would be an excess incentive to pay cash up front for assets rather than doing a mix of borrowing and holding cash for option demand.”

This isn’t what happens in U.S. states that have comprehensive business income taxes that don’t provide for business interest deductions, or in countries that rely heavily on VATs (which treat interest payments, profits, executive compensation, and dividend payments identically).

“Corporations would go bankrupt more easily and in general face higher transactions costs.”

There is a debt-equity bias, that debt-equity bias favors more leverage than would exist otherwise, and greater leverage levels than would exist in the absence of taxation increase the likelihood of bankrutpcy. Entities that don’t have a debt-equity bias or have a reduced bias (e.g. non-profits, individuals not entitled to the home interest deduction, cooperatives, subsidiaries, certain ownership structures for public utilities, trusts (grantor or non-grantor), pass through taxation entities) are less prone to over-leveraging and undercapitalization all other things being equal. For publicly held corporations that main practical limitation on leverage is the one imposed by bondholders and bank lenders in covenants in the loan documents which overwhelms tax issues.

“Contrary to common impression, the tax deductibility of interest payments does not give a tax advantage to borrowing, not if the return to savings is taxed.”

Not even wrong. Tax deductability of interest payments is appropriate in a Platonic ideal of the income tax concept. But, taxing corporate profits at both the corporate level and the shareholder level (either directly through dividend income or indirectly via capital gains income on value that is partially retained taxed corporate earnings) does deviate from that Platonic ideal. There are basically two ways to make the treatments tax neutral. Either you give shareholders a credit for taxes paid when dividends are distributed (which through the wonder’s of mathematics also solves the indirect double taxation of retained post-tax income and is the system used in most of the developed world), or by providing a corporate level deduction of dividends paid (which can lead to indirect double taxation of retained post-tax income in the absence infinite carrybacks and carryforwards of net operating losses, but is easier to administer in a federal system like that of the U.S. where there is no consistent state and federal tax policy coordination and can be managed with dividend policies of corporations). The Platonic income tax concept views taxation of capital gains and dividends as a non-deviation from the ideal and the absence of the dividends paid deduction for corporate income taxes as the preference. (If you don’t tax corporate income that isn’t distributed until it is distributed you get an immense incentive to retain earners which is present even in the current system.)

Incidentally, there is another way to structure the Platonic ideal of your income tax system. You could set it up on a cash flow basis. In this system, loan proceeds would count as income, capital expenditures would be deductible when paid, principal payments on loans would be deductible, and depreciation and amortization deductions wouldn’t exist. It is actually harder to game a cash flow system than our current system and that discrepency between cash flow and accounting depreciation/amortization is at the root of a great many tax loopholes (e.g. a corporate preference for leases over buying on credit that is not present with individuals who can’t deduct consumer interest or tax depreciation deductions on consumer good expenses). Cash flow systems don’t distinguish between dividends and undifferentiated interest-principal payments.

85 Rahul January 1, 2012 at 9:36 am

Interesting. Does any nation follow the cash flow system?

86 ohwilleke January 1, 2012 at 10:18 am

Not that I am aware. So far as I know it is a purely academic concept, and I believe there is a fair bit of scholarsihp on the point often in the context of understanding what depreciation and amortization deductions should be on the basis of some approach that is divorced from the actual useful life of the physical stuff attached to the deduction.

There are some provisions that go part way there. For example, most small businesses that aren’t capital intensive are able to “expense” all of their capital expenditures. And, finance leases can be economically equivalent to this system in proof of Coase’s law.

The loan proceeds/capital expenditure part is pretty subtle, since very few transanctions in real life involve borrowing money for the purpose of sitting on the cash proceeds wihtout making capital expenditures. Basically, the main effect of a system like this in real life would be to substitute a deduction for depreciation deductions for a deduction for loan principal payments. The installment payment system of accounting can also approximate this result.

But, a fully cash flow concept income tax system from the ground up doesn’t exist anywhere as of the last time I did an international income tax survey five or six years ago.

87 [Insert here] delenda est January 3, 2012 at 6:33 pm

Australia considered a variant of it in detail in the early 2000’s (then called ‘tax value method’). It was recommended by a government-commissioned report much of which was implemented but TVM was considered a bridge too far at that point. (the did, in that period, introduce a VAT and rewrite whole swathes of the tax system – it was a more comprehensive reform in 5 years than America has had in the last 25). You can still find it in the ‘Ralph Report’ (Review of Business Taxation, 1999), the relevant discussion is these three chapters: http://www.rbt.treasury.gov.au/publications/paper4/download/part4all_pdf.zip.

88 ohwilleke January 4, 2012 at 4:48 pm

Thanks, I didn’t know that. The Aussies frequently seem to be sources of innovation in Anglo-American law, introducing workable title registration for land (Torrens’ Title), the secret ballot, the authority of courts to honor technically defective wills that represent the true intent of the decedent (e.g. wills that are notarized but have one rather than two witnesses) after considering appopriate evidence.

89 mark January 1, 2012 at 11:49 am

The primary reason interest payments are deductible is that interest income is taxed, so that you need to deduct interest expense to get to the real interest income number, like any other expense. Arguably, interest expense deductions could be limited to the amount of interest income. However, that discriminates, it is asserted, in favor of persons with large income from lending. There was a broader constituency that wanted the broader deductibility (surprise). Such that interest expense became deductible against other income. That this encourages borrowing was apparently not considered a bad thing.

Apologize if others have noted this

Comments on this entry are closed.

Previous post:

Next post: